Review Article

Fracture Resistance of Teeth Restored with Post-retained Restorations: An Overview

Mahmoud Khaled AL-Omiri, BDS, PhD, FDS RCS,* Ahmad Abdelaziz Mahmoud, BDS, PhD,* Mohammad Ramadan Rayyan, BDS, MDSc, and Osama Abu-Hammad, BDS, MSc, PhD*
Abstract
Introduction: Posts have been used efciently to retain restorations for badly destructed teeth. This article critically analyzes the concerned topics related to the fracture resistance of teeth restored with dowelretained restorations. Methods: A systematic review of PubMed/MEDLINE, Cochrane, and Scopus databases was completed (from 1960 to 2010). Single or combined key words were used to obtain the most possible comprehensive list of articles. Checking the references of the relevant obtained sources completed the review along with a manual search to locate related articles on the topic. In vivo and ex vivo (laboratory, computer-based nite element, and photoelastic stress analysis studies) investigations related to the topic were included. Results: Many factors have been proposed to inuence the fracture resistance of postrestored teeth. Recognizing the signicance of these factors on the fracture resistance of teeth would aid in choosing the suitable treatment modality for every individual case. Fracture resistance was improved if tooth structure loss was limited, a ferrule was obtained, a post with similar physical properties to natural dentine was used, and adhesive techniques for post luting and coronal restoration were used. Adhesively luted resin/ ber posts with composite cores appear to be the best currently available option in terms of tooth fracture and biomechanical behaviour. Conclusions: Most guidelines were based mainly on ex vivo studies and to a lesser extent on limited in vivo studies. The lack of long-term controlled randomized clinical studies was the main hindrance to reaching a conclusive and undisputable opinion regarding endodontic posts in terms of tooth fracture and biomechanical behaviour. (J Endod 2010;36:14391449)

Key Words
Endodontic post, failure modes, fracture resistance, review

ndodontically treated teeth were claimed to be weaker and more prone to fracture than vital teeth (1). Fennis et al (2) investigated 46,000 insurance claims and reported a higher incidence of tooth fracture among endodontically treated teeth. The loss of water and collagen cross-linking might underlie the brittleness and weakness of enododontically treated teeth (3, 4). On the other hand, some studies reported that tooth substance of endodontically treated teeth had comparable biomechanical and physical properties to vital teeth (57). The loss of structural integrity is the main reason behind the vulnerability of endodontically treated teeth and their reduced resistance to fracture (5, 8). Most endodontically treated teeth suffer massive reduction in their structural stability because of the great loss of coronal dental structure caused by caries, fractures, and access preparations. Tang et al (9) summarized the risks that increased the potential of tooth fracture after endodontic treatment. The risks included loss of tooth structure, stresses attributed to endodontic and restorative procedures, access cavity preparation, instrumentation and irrigation of the root canal, obturation of the root canal, post canal preparation, post selection, coronal restoration, and inappropriate selection of tooth abutments for prostheses. Vertical root fractures of endodontically treated teeth prepared to receive endodontic posts were more frequent in the teeth of older patients and when dentine thickness was reduced (10). In their review, Dietschi et al (11) concluded that changes in tooth biomechanical behavior, tissue composition, and moisture after the loss of tooth vitality and proper endodontic treatment were limited and negligible. However, they found that teeth became weaker as they lost more coronal tissue because of caries or restorative procedures. Another possible reason behind their inferior resistance to fracture is the reduced proprioception of endodontically treated teeth (12). Consequently, they will be subjected to more harmful forces without a protective reex. Because of their inherent weakness, endodontically treated teeth need to be restored in a manner that would provide protection for the remaining tooth structure but would also allow the restoration of esthetic and functional demands (13). The restoration of endodontically treated teeth should aim at increasing tooth fracture resistance especially in cases with extensive tooth destruction (13). Some researchers recommended the use of posts for support and reinforcement of remaining tooth structure. This claim was supported by the ability of posts to distribute stress in a favorable way that would improve the fracture resistance of restored teeth (1419). Salameh et al (17, 20) showed that endodontically treated teeth restored with ber posts and ceramic crowns were more resistant to fracture and had less catastrophic fracture patterns than the ones restored with ceramic crowns and no posts. In another study, Salameh et al (21) used porcelain fused to metal, Empress II (Ivoclar

From the *Department of Prosthodontics, University of Jordan, Amman, Jordan; and Riyadh Colleges of Dentistry and Pharmacy, Riyadh, Saudi Arabia. Address requests for reprints to Dr Mahmoud AL-Omiri, BDS, PhD, FDS RCS, Department of Prosthodontics, Faculty of Dentistry, The University of Jordan, Amman 11942, Jordan. E-mail address: alomirim@yahoo.co.uk. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.005

JOE Volume 36, Number 9, September 2010

Fracture Resistance of Teeth Restored with Post-retained Restorations

1439

Review Article
Vivadent, Schaan, Liechtenstein), SR Adoro (Ivoclar Vivadent, Noble Park North, Victoria, Australia), and Cercon (Dentsply Ceramco, York, PA) crowns to restore endodontically treated maxillary incisors and reported similar conclusions as described previously regardless the type of used crown. Also, Cagidiaco et al (18) and Ferrari et al (19) showed that the placement of bre posts did improve the survival rate of endodontically treated premolars. Furthermore, Nam et al (22) found that the fracture resistance of endodontically treated premolars with one to four remaining coronal walls was signicantly increased when they were restored with ber posts. Moreover, teeth showed better stress distribution and fracture patterns when restored with ber posts. However, fracture resistance of teeth with no remaining coronal walls was not improved when ber posts were used. Figure 1A through D presents some fracture modes that associate metal, glass ber, and carbon ber posts. When compared with no post treatment, Nothdurft et al (23) reported better fracture resistance of premolars with class II cavity preparations after they were restored with zirconia, ber, or titanium posts. They concluded that posts in premolars with class II cavities would improve tooth resistance to the extra-axial forces. In another study, Nothdurft et al (24) reported no difference in fracture resistance of premolars with class II cavities restored with crowns alone or crowns and posts (titanium, zirconium dioxide, glass ber, and quartz ber posts). From these two studies, it can be concluded that the use of crowns might cancel the effect of posts on fracture resistance of restored teeth. On the other hand, many studies challenged the use of posts for support and reinforcement of remaining tooth structure and even considered post placement as a risk factor that weakened the remaining tooth structure and predisposed tooth fracture. These studies showed that restoring endodontically treated teeth using cast metal, prefabricated metal, or ber posts had negative effects on teeth fracture resistance (2530). Unlike other modes of failure, root fracture of post restored teeth is the most catastrophic and almost always results in extraction of the involved tooth (27, 28). A higher incidence of vertical root fractures was reported among teeth restored with titanium, zirconia, and prefabricated/cast metal endodontic posts (26, 28, 31, 32). This was greatly attributed to stress concentration within the radicular dentine during post placement and, consequently, the altered pattern of stress distribution upon loading (3234). Fokkinga et al (35) reported that the presence or absence of metal/ber posts did not affect the fracture resistance and failure modes of endodontically treated premolar teeth with resin composite crowns and no retained coronal tooth structure. Therefore, they suggested that posts are not necessary for the restoration of such teeth. Also, Mohammadi et al (36) found no difference in fracture resistance of premolars restored with direct resin composite in the presence or absence of ber post and cusp coverage. Furthermore, Soares et al (29) found that the loss of dentinal structures and the presence of bre posts caused more stress concentration in tooth and restoration and decreased the fracture resistance of teeth. However, they found that ber posts were associated with less catastrophic fracture modes when there was an extensive loss of tooth tissues. An in vitro study by Pilo et al (30) showed that endodontic therapy for upper bifurcated premolars caused loss of more dentine at bifurcation area of both roots in comparison to outer areas. Furthermore, the preparation of post canals undermined root strength because it left less than the recommended 1-mm dentine thickness around the post canal. Buccal roots were more affected by this pitfall. Therefore, they recommended limiting the use of posts in upper bifurcated rst premolars, and when necessary the posts should be used in lingual roots rather than the buccal ones. 1440 The literature is full of controversial conclusions regarding the best post to use for the restoration of endodontically treated teeth. This article critically analyzes the concerned topics and controversy related to the fracture resistance of teeth restored with dowelretained restorations.

Methods
A systematic review of PubMed/MEDLINE (from 1960 to 2010), Cochrane, and Scopus databases (to 2010) was completed. Single or combined key words (fracture resistance, endodontic post and core, ber posts, adhesive luting, and endodontically treated teeth) were used to obtain the most possible comprehensive list of articles. Checking the references of the relevant obtained sources completed the review along with a manual search to locate the most relevant articles on the topic. In vivo and ex vivo (laboratory, computer-based nite element, and photoelastic stress analysis studies) investigations related to the topic were included in this review. Because the number of long-term randomized controlled clinical trials (RCTs) was limited in this eld, retrospective, prospective, descriptive, review, and RCT studies were included. Studies describing post and core systems to restore endodontically treated teeth and their mechanical and physical properties were included. Also, articles investigated ber posts, ceramic posts, cast posts, and prefabricated metal posts, and different core systems were included. Furthermore, articles studied failure modes and fracture resistance of teeth restored with different post and core systems were included.

Results
Methods Used to Assess Stress Distribution and Fracture Resistance of Post-Restored Teeth Few RCT studies have investigated the fracture resistance of teeth restored with post and core restorations. This might be attributed to the difculties encountered in controlling related factors clinically such as force magnitude and direction, teeth geometry, and remaining tooth structure (13). In vitro studies, on the other hand, are easier to control and conduct, but their recommendations should be interpreted with caution because of their limitations and conicting results. It is impossible to accurately simulate intraoral conditions by in vitro studies (13). However, attempts were made to investigate fracture resistance of endodontically treated teeth restored with posts and cores experimentally (37). Three methods have been frequently used for this purpose including laboratory experiments, photoelastic analysis, and nite element analysis (13, 37). Although most mechanical laboratory studies were aimed at investigating the failure loads and modes of restored teeth, photoelastic and nite element analysis studies were used to investigate stresses within restored teeth upon loading and the effect of post placement on stress values and distribution (13). Laboratory Experiments Many mechanical studies were conducted to investigate the effect of post placement and related factors on the fracture resistance of endodontically treated teeth (38-43). Extracted teeth, especially incisors and premolars, were used in these studies. Static loading at a constant angle was applied to restored teeth in some studies (41, 44). However, actual masticatory forces are multidirectional and repeatedly applied on larger areas (13). In order to mimic such conditions, cyclic loading was also applied in some studies (42, 43, 45). Hayashi et al (46) applied simultaneous static and cyclic loading to restored teeth in both horizontal and vertical directions and reported that teeth restored with ber posts and composite cores were more
JOE Volume 36, Number 9, September 2010

Khaled AL-Omiri et al.

http://endodontic.ws

Review Article

Figure 1. A catastrophic coronal dentine-core-root vertical fracture. (A) A catastrophic horizontal root fractureassociated prefabricated metal post. (B) A favorable core fracture-associated ber post. (C) A catastrophic core and root fractureassociated carbon ber post. (D) A prepared post canal for a glass ber post. (E) Note the oval shape of the root canal cross-section and the rounded cross-section of the post canal. There was no need for further preparation of the canal to improve post tting.

resistant to fracture than those restored with metallic posts (46). Also, Hu et al (47) applied static and cyclic loading to teeth restored with cast metal posts, resin composite posts, and carbon ber posts and found that teeth restored with carbon ber posts were resistant to more load cycles, whereas cast metal posts required the highest fracture loads. However, resin composite posts were associated with favorable root fractures, whereas all cast metal posts and carbon ber posts were associated with unfavorable root fractures (47). Strain gauges were used to calculate stresses within examined teeth models. However, this method may need complex mathematical calculation and is limited by the number of gauges that can be placed on the

model (48). Experimental studies in general have many limitations. The strength of dentine varies according to age, pulpal condition before extraction, and the storing media, which may affect the fracture patterns (13). Resins and stones are commonly used for mounting teeth during testing (4951). They set by the exothermic reaction, which may affect the dental structures (52). Moreover, despite researchers attempts to reproduce the resiliency of periodontal ligament and bone, they could not provide the correct actual resiliency of these structures. Some experimental studies applied forces directly on the post head or the core, which is commonly not the case clinically (49, 50, 53). This may produce misleading and contradictory results. Naumann et al 1441

JOE Volume 36, Number 9, September 2010

Fracture Resistance of Teeth Restored with Post-retained Restorations

Review Article
(54) highlighted the lack of standardization of test parameters applied for in vitro studies of fracture resistance of teeth restored with endodontic posts. They found that 95% of studies used static loading, and only 15% of the studies used thermocycling and mechanical loading to test fracture resistance of teeth restored with endodontic posts. Furthermore, different studies used different teeth, and 57% of the studies used upper incisors only. Also, most studies used specimens without crowning. Therefore, the different test parameters and standards might be the reason behind the controversy surrounding the issue of fracture resistance of teeth restored with endodontic posts. within the root. These microcracks grow and propagate to cause fatigue failure and unrestorable root fractures (32, 63). The coronal third of the root has been reported as the main site of stress concentration (69, 70). The inner dentine of the root is usually less mineralized and possesses more water content than the outer dentine (71). Therefore, the inner dentine has a higher potential for plastic deformation and crack formation. Using experimental and clinical investigations, Kishen et al (72) examined fractured post and core restored teeth using laser scanning confocal microscopy and scanning electron microscopy and observed numerous microcracks within the inner dentine material adjacent to the endodontic post. They also used FEMs of dentine to relate crack formation and root fractures to tensile stresses generated within dentine. High strains were generated within the inner dentine substance upon tensile loading. They, among other researchers, concluded that crack formation and fracture progression in postrestored teeth were initiated from the inner region of dentine (72 74). Thickness of the inner dentine and factors related to the postdentine interface seem to play a major role in stress distribution and fracture resistance of endodontically treated teeth. Therefore, the removal of inner dentine during post placement should be minimized as much as possible to maintain adequate fracture resistance of dowel-restored teeth (72).

Photoelastic Stress Analysis Photoelastic models were used to study the effect of post placement and related factors on the patterns of stress distribution within endodontically treated teeth (5557). Photoelasticity is the property of transparent materials to exhibit colorful patterns known as fringes when stressed under polarized light (55). A transparent double-refraction plastic sheet is used to fabricate specimens for two-dimensional photoelastic stress analysis (57). Stress concentration areas can be identied according to the sequence of color bands of the fringes. Three-dimensional photoelastic models are less frequently used because of the difculty of their construction and their high cost (55). Photoelastic methods provide visual evidence of stress patterns within tested models. However, the properties of photoelastic materials are different from those of tooth structure, and the modeling of objects made of more than one material is technically difcult (13). Finite Element Method The nite element method (FEM) has been frequently used for stress analysis in many aspects of dentistry. Many studies used the FEM to investigate the effect of post placement and related parameters on the stress picture within dowel-restored teeth (34, 5865). Models are simulated using a computer and subdivided into a nite number of smaller divisions termed elements. Material properties, boundary conditions, and loads are then assigned for the elements. Stress analysis is then performed by solving differential equations of elements to quantify stresses generated within these elements. FEM offers the advantages of easy simulation of nonhomogenous models and easy changing of parameters like material properties and loading conditions (58). However, material properties, loading conditions, and boundaries simulated do not represent the absolute clinical situation that constitutes a limitation for the application of this method (13). Stress Distribution Within Post-Restored Teeth and Its Relation to Fracture Resistance Viscoelastic properties of a tooth structure affect stress distribution within the tooth. If viscoelastic properties are undermined, the tooth will be mechanically compromised in terms of stress distribution, values, and concentration. Root canal treatment and post and core restorations are examples of conditions in which tooth viscoelasticity is reduced and this might be the reason why teeth are more liable to fracture (66). Post insertion alters the pattern of stress distribution within root dentine (37). Different posts affect the stress picture in post-restored teeth differently, which will be discussed in the following sections. Upon loading, teeth restored without post insertion show a concentration of stresses at the circumference of the tooth with uniform stress distribution within the root canal (67). Post insertion results in a nonuniform distribution of dentinal stresses within the root (68). Upon functional loading, critical stress concentration areas are produced at the post-dentine interface that precipitates microcracks
1442
Khaled AL-Omiri et al.

Factors That Affect the Fracture Resistance of PostRestored Teeth Many factors inuence the fracture resistance of post-restored teeth. Some factors are directly related to the post-core system including post length, post diameter, post design, post material, post tting, core material, ferrule effect, and luting cement (13). Other factors are related to the restored tooth and include cuspal coverage, remaining coronal tooth structure, loading conditions, and alveolar bone support (13, 15). The effect of these factors on the fracture resistance of dowelrestored teeth will be discussed in details throughout the following sections. Post Length Stress analysis studies showed better stress distribution within dentine when longer metal, ber, or zirconia posts were used (65, 75). Similarly, increased fracture resistance was associated with increased post length (15, 76). Longer posts provide greater rigidity and less root bending than short posts (13). An endodontic post should extend beyond the level of alveolar bone to provide better root support (77). Davy et al (75) reported a decrease in cervical stresses when the post length was increased up to two thirds of the root length. However, increasing the post length beyond two thirds of the root may cause stress concentration at apical area of the root meanwhile provide no additional support for the cervical region (78). Burns et al (79) reported minimal effects of the post length on stress distribution within dentine. Other studies supported this nding and found no effect of the post length on fracture resistance of restored teeth (45, 80, 81). Giovani et al (82) found no effect of post length on the fracture resistance of teeth restored with metal cast post and core. However, they showed signicantly higher fracture resistance for teeth restored with longer glass ber posts (10-mm long) when compared with shorter ones (6-mm long). On the other hand, Cecchin et al (83) found that longer ber posts (12- or 8-mm long) were associated with higher fracture resistance of teeth when compared with shorter ones (4-mm long). However, they concluded that too much preparation to get the longest post space was not essential for better fracture resistance of postJOE Volume 36, Number 9, September 2010

http://endodontic.ws

Review Article
restored teeth. In their study, posts that were just above half of root length (8-mm long) were enough to improve the root fracture resistance and were similar to posts that were two thirds of the root length (12-mm long). Adhesive xation of the post and ferrule incorporation might decrease the effect of post length on the fracture resistance of dowel-restored teeth (45, 81, 84). In conclusion, the optimum post length depends on several factors including root length, crown height, level of bone support, and technique of cementation. Adhesive cements, ferrule effect, and full coronal restoration may reduce the effect of post length on the tooth fracture resistance (77, 84). was more than that recorded for tapered ones over up to 8 years of function.

Post Diameter A smaller post diameter is recommended to retain more dentine during preparation of post channel, which enhances the fracture resistance of dowel-restored teeth (39, 85). The ability of a tooth to resist fracture is directly related to the amount of remaining dentine around the post (39, 86). Increased radicular dentinal stresses were observed when a post diameter was increased (58, 87). The larger the post diameter the less the fracture resistance of a dowel-restored tooth (39). Recommendations regarding the adequate amount of radicular dentine around a post vary among researchers. One third of the root width was recommended as the higher limit for the post diameter (52). Some researchers suggested the preservation of 1 mm of sound dentine around the post channel (25). Halle et al (88) recommended the preservation of more radicular dentine and suggested that 1.75 mm retained dentine around the post was sufcient to resist fracture of the tooth. In order to reduce failures and fractures, Mou et al (89) recommended that the optimum post to root diameter ratio should be approximately 1:4. Post Design Tapered metal posts cause greater cervical stress concentration than parallel-sided posts (75, 90). This was attributed to the wedging effect introduced by tapered posts. Apical stresses, on the other hand, tend to be higher when parallel-sided posts are used (91). This was attributed to the sharp angles and reduced tooth structure at the apical area. A higher incidence of root fracture was reported when tapered posts were used (50). Threaded metal posts were associated with stress concentration at the dentine-thread interface. Such areas can predispose crack formation and jeopardize the fracture resistance of dowel-restored teeth (92). Decreasing the number of threads and increasing the spaces between them produces less harmful stresses (93). Prethreading the post cavity and slight counter rotation of the post after engagement may also reduce the harmful stresses produced by threads (48). There are no clinically available threaded ber posts. However, Uddanwadiker et al (94) proposed a nite model of threaded ber post and found increased stress concentration because of this post, which reduced the fracture resistance. Among various post designs, tapered threaded metal posts were reported as the most hazardous to the fracture resistance of dowelrestored teeth. Likewise, parallel serrated posts were the most favorable in this regard (14). Silva et al (95) compared four different metal posts to glass ber posts and found better stress distribution within the teeth restored with glass ber post. Furthermore, they found more stress concentration at the coronary portion of metal posts. They concluded that post material was more important for stress distribution than the external characteristics of the post. On the other hand, Signore et al (96) found that the survival rate of parallel-sided glass ber posts
JOE Volume 36, Number 9, September 2010

Post Material The mechanical properties of materials used for post construction as well as their biocompatibility may inuence the fracture resistance of restored teeth. Posts with higher modulus of elasticity like metals are associated with higher failure loads (52, 97103). However, they tend to cause catastrophic and irreparable root fractures when they fail (97, 98, 100, 101, 104106). Being more rigid than the tooth, high modulus elasticity posts produce stress concentration at critical areas of the root and cause more fractures (58, 63). Unlike rigid posts, posts with a similar modulus of elasticity to dentine (eg, ber posts) can distribute stresses more evenly along the post-dentine interface and cause less root fractures (56, 58, 63, 107). Materials with a low modulus of elasticity bend more under load and tend to fail before causing root fracture (108110). This constitutes a protective mechanism for the tooth structure. Many studies showed high clinical survival and success rates for teeth restored with ber posts (111113). The vast majority of failures were attributed to causes other than catastrophic tooth fracture. Cagidiaco et al (110) concluded that ber posts outperform metal posts in treatment of root canal treated teeth. However, in a randomized controlled clinical pilot study; Naumann et al (114) compared the clinical performance of titanium and ber posts for 2 years and reported similar successful clinical outcomes for both treatments. Many studies showed better fracture resistance of teeth restored with ber-reinforced resin posts (which had a similar rigidity to dentine) when compared with metal or zirconia posts (which had a much higher modulus of elasticity than dentine) (45, 82, 98, 106, 115117). Cast posts and cores were frequently associated with deep catastrophic root fractures (27, 28, 97, 118). Nevertheless, Newman et al (119) and Toksavul et al (120) reported that less fracture resistance and more catastrophic root fractures were associated with glass ber posts when compared to zirconium posts. Stockton and Williams (121) suggested that ber post exibility might cause stress redirection toward the post-tooth interface and thus increase the failure rate. Furthermore, some studies reported no significant difference in fracture resistance of restored teeth when berreinforced resin or metal posts were used (118, 122, 123). Moreover, Fokkinga et al (123) found similar fracture patterns and behaviors of premolars restored with metal crowns when either metal or ber posts were used. Also, Nothdurft et al (23) reported no difference in fracture modes and patterns when either titanium, quartz ber, glass ber, or zirconium dioxide posts were used to restore premolars with class II cavity. Furthermore, Toman et al (124) found that teeth restored with resin cemented silica-coated titanium posts and composite cores had higher fracture resistance than teeth restored with resin-cemented zirconia or glass ber posts (with or without silica coating) and composite cores. Corrosion resistance of post material may inuence the fracture resistance of restored teeth. Metal posts were found to corrode overtime (125). It was proposed that corrosion products could migrate through the dentinal tubules and build up intratubular pressure, which predisposed root fractures (13). Titanium posts have high corrosion resistance when compared with other metal alloys. Brass alloys, on the other hand, have very low corrosion resistance (13). The storage of ber posts in saline water reduces their fracture strength and causes voids between the resin matrix and bers (126). However, this does not occur if ber posts are stored at in a condition in mineral oil or in human teeth in saline water. This nding might
Fracture Resistance of Teeth Restored with Post-retained Restorations

1443

Review Article
explain some of the controversy found in the literature because different storage conditions used in different studies might affect ber post fracture strength differently and thus affect the results. composite cores regardless of the pretreatment of the post surface (141). Radovic et al (142) found that sandblasting increased the microtensile bond strength between composite cores and ber posts. Furthermore, they reported no benet from further chairside treatment (ie, application of silane or bonding agent) of the sandblasted post surface because this was associated with a reduction rather than an improvement of bond strength. The use of ber posts improved the exural properties of core composite resin regardless of the ber direction in the ber post (143). When the post surface was treated with bonding agents before the addition of the composite core, Artopoulou et al (144) recorded less adhesive failures between ber posts and composite cores than that between composite cores and a titanium or stainless steel post. However, the lack of pretreatment of the post surface was associated with better retention between metal posts and composite cores than between ber posts and composite cores. Core materials on titanium posts had better resistance to torsional forces when the post surface is treated with chemical surface conditioning techniques such as silica coating (145). The bond between core material and the ber post is chemical, and this increases the retention of cores when ber posts are used. For stability, a reliable bond between the core material and the post should be generated. Schmitter et al (81) concluded that centrally positioned berreinforced posts did not contribute to load transfer as long as the bond between the tooth and composite core was intact.

Post Fitting Goracci et al (127) concluded that sliding friction was the main factor that affected resistance to dislocation of resin-bonded ber posts. Also, the use of dentine adhesive did not improve dislocation resistance when compared with the use of resin cement without dentine adhesive. The presence of interfacial gaps and the incomplete removal of smear layer might be the reason for these ndings. Poorly tted posts might create levers within the root canal, making the tooth more liable to fracture (128). Close adaptation of posts to the canal walls was found to increase the fracture resistance of restored teeth signicantly (50). Santos et al (129) showed that lacking effective bonding between the root and posts with different elastic modulus was associated with a higher risk of vertical root fracture in upper premolars. Schmitter et al (81) concluded that when ber post-restored teeth were crowned, centrally positioned berreinforced posts did not contribute to load transfer as long as the bond between the tooth and composite core was intact and resin cement was used to bond the ber post. Buttel et al (76) found that the fracture resistance of teeth restored with ber posts and composite crowns without ferrules was not affected by post t within the root canal. Therefore, excessive post canal preparation to achieve optimal circumferential post t is unjustiable because it will not increase fracture resistance of teeth. Figure 1E presents adequate post canal preparation when the cross-section of the root is oval. Core Material Less stiff cores are expected to deform under occlusal loads and thus reduce the stress concentration within the dentine (108). Composite resins were reported to fracture under loads lower than those necessary to fracture the tooth (130). This is considered as a protective mechanism for tooth structure. Cast metal posts and cores were associated with more root fractures than prefabricated posts and amalgam or composite cores (131). However, crown placement with adequate ferrule can mask the effect of core build ups on the fracture resistance of restored teeth (132135). This was attributed to the fact that a crown restoration could favourably distribute stresses and redirect them toward the root (90,134). Coating zirconia posts with tribochemical silica coating and silanization increases their fracture resistance and the bond strength to composite resin (136). When compared with no airborne particle abrasion of the post surface, better long time bond strength to composite cores was recorded when zirconia posts were abraded using airborne particle abrasion and received primer and silane coupling agent to their surface (137). Fiber post surface can be treated by silane coupling agents or bonding agents to improve their bond to composite resins (138). Better bond strength was recorded when silane coupling agents were used. Treatment of the ber post surface with hydrogen peroxide before silanization increases the bond strength to composite cores more than using silane coupling agent alone (139). This might be caused by the dissolution of epoxy resin matrix of the post, which enhances better surface characteristics. Monticelli et al (140) concluded that surface conditioning enhanced ber post bonding properties. Also, the bond strength of pretreated ber posts to composite cores was satisfactory. However, thermocycling reduces the bond strength between ber posts and
1444
Khaled AL-Omiri et al.

Ferrule Effect The ferrule concept was proposed by Rosen (146) in 1961. He recommended the use of a metal collar extension beyond the gingival margin of the core to encircle the tooth. The ferrule effect in association with post and core treatment was investigated by many researchers (47, 147-150). Most of the previous studies were performed in vitro and generally have accepted that ferrules incorporated within cores or nal crowns might increase the fracture resistance of restored teeth by reinforcing their external surfaces to resist stresses accompanied by functional lever forces. Ferrules also help to maintain the integrity of cement seal around the restoration (151). Opinions vary regarding the optimum height and design of adequate ferrule. However, most researchers recommended a minimum 1- to 2-mm of ferrule height of almost parallel dentine walls at the whole circumference of the tooth (47, 152). The higher the ferrule the greater the fracture resistance (151). A uniform height of the ferrule at the whole tooth circumference was recommended because it was found more effective in supporting the tooth than a nonuniform height (150). Al-Omiri and Al-Wahadni (99) reported that retaining coronal dentine did increase the fracture resistance of teeth. However, they found that increasing the amount of retained dentine more than 2 mm did not improve the tooth fracture resistance any further. Schmitter et al (81) concluded that increased ferrule height and resin bonding of the crown resulted in higher fracture loads of teeth. They recommended the use of resin-bonding agents with crowns that had a small ferrule height. Also, Dorriz et al (153) recommended the use of ferrule or bonding with an opaque porcelain layer (if cast metal post was used) to improve the fracture resistance of grossly destroyed teeth. Despite the large number of studies that supported the use of ferrules, some researchers questioned the benet of ferrules because they did not provide additional support for restored teeth (50, 104, 135, 154, 155). Saupe et al (104) and Al-Hazaimeh and Gutteridge (154) concluded that the use of resin cements in their studies might
JOE Volume 36, Number 9, September 2010

Review Article
have cancelled the effect of ferrule on the fracture resistance of restored teeth. These ndings were also supported by Mezzomo et al (155) who reported no signicant difference in fracture resistance of nonferruled specimens restored with resin-cemented posts and ferruled ones. When zinc phosphate cement was used, ferruled specimens showed signicantly higher fracture resistance than nonferruled ones. Moreover, Ng et al (156) reported a higher incidence of root fracture among teeth restored with bonded posts and cores when ferrules were incorporated. Also, Naumann et al (157) found that incomplete ferrules (that does not encircle 360 of the tooth) were associated with better tooth fracture resistance when compared with ferrules that totally encircle the tooth. They concluded that tooth structure preservation is more important for the fracture resistance of post-treated teeth. The ndings of these studies underlie that the incorporation of ferrules in conjunction with resincemented posts for the sake of tooth reinforcement might constitute an unjustiable insult to the remaining tooth structure. The invasion of the biological width during tooth preparation should be avoided (5). This creates a dilemma in cases in which no adequate height of tooth structure is remained above the crestal bone for the incorporation of a ferrule. Surgical crown lengthening and orthodontic extrusion of the tooth may help in the establishment of adequate ferrules in such cases (151). Besides patient discomfort, extra cost and time are required for such procedures. Moreover, surgical crown lengthening is accompanied by the reduction of the effective root length, which affects the crown:root ratio negatively. Gegauff (158) investigated the effect of crown lengthening for ferrule purposes on the failure loads of simulated analog teeth restored with post- and core-retained crowns impeded within simulated periodontal ligament and alveolar bone. He reported signicantly lower failure loads of teeth that received crown lengthening and ferrules. The reduction of supporting tissues combined with the altered crown:root ratio seemed to weaken the restored teeth even with the incorporation of ferrules. The literature lacks retrospective and prospective clinical studies that investigated the ferrule effect. Torbjorner et al (27) reviewed records of 72 failed metal posts and observed more post fractures in cases in which a ferrule was not incorporated. However, Cagidiaco et al (18) and Ferrari et al (19) found no role for ferrule effect on the survival of premolar teeth restored with ber posts and composite cores. This could be caused by the effect of resin cements used for the cementation of ber posts and the close similarity between the values of the elastic modulus of ber posts/resin cements and the elastic modulus of dentine and thus obtain favorable stress distribution. Crown coverage might also explain these results because it directs the load to tooth nish line and bypasses the post-core assembly and thus directs stress towards the outer surface of the tooth and prevent tooth fracture. More controlled clinical follow-ups are required to reach sound conclusions regarding the ferrule effect and its proposed benets. Nevertheless, the available literature favors the incorporation of a ferrule for nal restorations. However, this should be in balance with the remaining tooth structure and crown:root ratio (11, 151). modulus caused higher stress concentration within the cement layer. Also, cement thickness did not affect stress distribution within the post, dentine, or cement layer. Previous studies concluded that resin-based cements could increase the fracture resistance of dowel-restored teeth (49, 104, 154, 155, 160). Teeth restored with posts luted with resin cements were found more resistant to fracture than teeth restored with posts luted with zinc phosphate or glass ionomer cements (49, 155, 161). According to these studies, the favorable behaviour of berreinforced resin posts might be related to the resin cements frequently used to lute them rather than to the post material itself. In order to obtain favorable stress distribution, the monoblock type of restoration was recommended for the restoration of pulpless teeth (162). This treatment involves restoring the tooth with post, core, and crown using biomechanically homogenous bonded materials and adhesive techniques. However, it proved difcult to predict or achieve this because it is difcult to clean the smear layer from the root, to remove the water droplets and moisture from the root, and to use post and cement materials that do not shrink and have moduli of elasticity that match the dentine (163). The technique of cementation might also affect the fracture resistance of restored teeth (13). Residues within a post space, bubbles within the cement layer, and excessive seating pressure can cause stress concentration within the root and predispose fracture (13). Dietschi et al (164) recommended the use of specic combinations of adhesives and cements to overcome the problems of ovoid canal shape and dentine moisture that might reduce the efcacy of adhesion between the tooth and the post. Also, Kivanc and Gorgul (117) concluded that self-etching adhesives were better to use than etch and rinse adhesives for luting endodontic posts. An increased ferrule height and resin bonding of a crown resulted in higher fracture loads of post-restored teeth (81). Finally, Hammad et al (165) concluded that the obturation of roots with resin-based obturation materials increased the resistance of teeth to vertical root fracture.

Coronal Coverage Crowning endodontically treated posterior teeth and badly damaged anterior teeth increases their resistance to fracture, whereas crowning endodontically treated anterior teeth with intact coronal structure does not improve their fracture resistance (13, 166, 167). Articial crowns alter the distribution and transmission of stresses into a post-root complex (90, 134, 168). Providing a crown with adequate ferrule has more inuence on the fracture resistance of dowel-restored teeth than factors related to post and core materials and designs (134, 169). DArcangelo et al (170, 171) suggested the use of ber posts when veneer restorations were used to restore endodontically treated teeth because they found that ber posts increased the fracture resistance of endodontically treated teeth prepared for or restored with composite or porcelain veneers. Remaining Coronal Tooth Structure Some studies suggested that fracture resistance of post-restored teeth would be reduced if more coronal dental structures were lost (13, 73, 172, 173). In a retrospective long-term clinical study, Ferrari et al (112) concluded that the mechanical failure of teeth restored with ber posts was related to the remaining coronal tissues. Similar results were reported in a follow-up clinical trial by Cagidiaco et al (110). Nam et al (22) found that the fracture resistance of endodontically treated premolars with one to four remaining coronal walls was significantly increased when they were restored with ber posts. Moreover, teeth showed better stress distribution and fracture patterns when
Fracture Resistance of Teeth Restored with Post-retained Restorations

Luting Cement The luting cement provides a buffer zone between the post and the dentine, which might affect stress distribution upon loading (1). Brittle conventional cements like zinc phosphate may disintegrate upon functional loading and cause levers that concentrate stresses at the apical root portion and cause root fracture (159). Using adhesive cements allows even stress distribution over the entire bonded surfaces. Consequently, a post can absorb functional stresses and then direct them toward the long axis of root and thus make them more favourable (118). Spazzin et al (107) found that cements with a higher elastic
JOE Volume 36, Number 9, September 2010

1445

Review Article
restored with ber posts. However, fracture resistance of teeth with no remaining coronal walls was not improved when ber posts were used. However, the relation between remaining coronal dentine and fracture resistance has been questioned by other researchers especially when coronal coverage was provided because this might mask the effect of the retained tooth structure (104, 154, 174). The last 2 decades have witnessed a shift toward studying the effect of different post and core systems on the fracture resistance of dowelrestored teeth. Variety of available posts and introduction of new non-metal post systems have greatly contributed to the large number of available studies that handled this topic. Recently, a new shift occurs again in favour of using such posts as they are believed to even improve the fracture resistance of endodontically treated teeth when compared with no post treatment. However, this new tendency lacks enough evidence from long term randomized controlled clinical studies to set this trend as the standard of future treatment of endodontically treated teeth. In order to avoid weakening of teeth by post placement, many post features were thoroughly investigated. Because of the lack of long-term controlled clinical trials and the contradictory results of in vitro studies, the optimum post features are not yet clearly determined. It is important to keep in mind that the restoration of teeth using posts does involve a system that consists of post, core material, and luting cement. The overlying crown and functional occlusal loads are added to this system because they all inuence the success of the whole treatment. Therefore, it is less realistic to test fracture resistance of postrestored teeth without including the effect of coronal coverage, the direction and magnitudes of masticatory forces, and the cyclic nature of functional loads. Failing to consider all this might be the reason behind the contradiction regarding the best post core luting agent to use in order to reduce tooth fracture or even strengthen the endodontically treated tooth. The treatment of endodontically treated teeth using posts might be more successful if tooth structure loss is limited, a ferrule is obtained, a post with similar physical properties to natural dentine is used, and adhesive techniques for post luting and coronal restoration are used. Therefore, when the advantages and disadvantages of different luting agent-post-core systems and materials are considered, the adhesively luted resin/ber posts with composite cores appear to be the best offered luting agent-post-core system currently available in terms of tooth fracture and biomechanical behaviour. It is essential to keep in mind that these guidelines are based mainly on ex vivo studies and to a lesser extent on limited in vivo studies. A lack of long-term controlled randomized clinical studies is the main hindrance to reaching a conclusive and undisputable opinion regarding endodontic posts in terms of tooth fracture and biomechanical behavior.

Loading Conditions The magnitude and direction of functional loads play a major role in stress concentration within dowel-restored teeth. Higher occlusal forces, like in cases of parafunctional habits, were associated with higher failure rates of such teeth (175177). Horizontal loads cause a signicantly higher stress concentration within dentine than loads more parallel to the long axis of the tooth (64, 90, 91, 148, 168, 178181). The effect of load direction on maximum stresses generated within restored teeth was found to be more signicant than the effect of post design and dimensions (64, 90). Because of their inclination, anterior teeth are most likely subjected to more horizontally directed loads (13, 177). Post insertion would magnify stresses produced within dentine upon such loads (64). This is the reason behind the conception that the preservation of the coronal tooth structure of endodontically treated anterior teeth is more effective for tooth support than post insertion (13). Meira et al (182) found that the effect of the elastic modulus of a post on concentration, magnitude, and direction of dentinal stress was dependent on load direction. They showed that when highmodulus posts were modeled, horizontal loads caused more stress on the apical area of roots and such a load suggested a vertical root fracture. On the other hand, they found that forces at 45 and 90 caused more stress on the cervical area when low-modulus posts were modeled; such loads suggested debonding of the post. Alveolar Bone Support A lower fracture resistance was reported among roots of periodontally compromised teeth reconstructed with posts and cores (183). Naumann et al (184) concluded that the reduction of the level of bone support would reduce the fracture resistance of ber posts restored teeth. Alveolar bone level is considered a critical factor for stress concentration and tooth fracture (185). Finite element studies reported massive increase in dentinal stresses as the alveolar bone level was diminished (91, 180). The loss of alveolar support will lower the level of the mechanical fulcrum, which in turn will jeopardize the fracture resistance of post restored teeth (158, 186). In order to obtain adequate fracture resistance, at least 1:1 crown:root ratio should be ensured and the post should extend beyond the level of alveolar bone (77). Surgical crown lengthening can reduce the crown:root ratio and predispose tooth fracture. Gegauff (158) reported signicantly lower failure loads of teeth that had crown lengthening even with the presence of a ferrule.

Acknowledgment
The authors thank Mrs AbdelAziz for all her help during the preparation of this manuscript.

References
1. Schwartz RS, Robbins JW. Post placement and restoration of endodontically treated teeth: a literature review. J Endod 2004;30:289301. 2. Fennis WM, Kuijs RH, Kreulen CM, et al. A survey of cusp fractures in a population of general dental practices. Int J Prosthodont 2002;15:55963. 3. Helfer AR, Melnick S, Schilder H. Determination of moisture content of vital and pulpless teeth. Oral Surg Oral Med Oral Pathol 1972;34:66170. 4. Rivera EM, Yamauchi M. Site comparisons of dentine collagen cross-links from extracted human teeth. Arch Oral Biol 1993;38:5416. 5. Peroz I, Blankenstein F, Lange K, et al. Restoring endodontically treated teeth with posts and cores: a review. Quintessence Int 2005;36:73746. 6. Huang TJ, Schilder H, Nathanson D. Effects of moisture content and endodontic treatment on some mechanical properties of human dentin. J Endod 1991;18: 20915. 7. Sedgley CM, Messer HH. Are endodontically treated teeth more brittle? J Endod 1992;18:3325. 8. Reeh ES. Reduction in tooth stiffness as a result of endodontic restorative procedures. J Endod 1989;15:5126.

Conclusions
The fracture resistance of post-restored teeth has been investigated thoroughly in the literature. Previously, the argument was in favor of reinforcing the teeth by endodontic posts. Nowadays, there is a general agreement among researchers that posts per se do not offer reinforcement for restored teeth; on the contrary, their insertion involves procedures that usually sacrice tooth structure and reduce the fracture resistance of teeth. Unrepairable root fractures have been frequently reported as the most catastrophic mode of failure that was associated with post placement, especially when rigid posts were used. 1446

Khaled AL-Omiri et al.

JOE Volume 36, Number 9, September 2010

Review Article
9. Tang W, Wu Y, Smales RJ. Identifying and reducing risks for potential fractures in endodontically treated teeth. J Endod 2010;36:60917. 10. Mireku AS, Romberg E, Fouad AF, et al. Vertical fracture of root lled teeth restored with posts: the effects of patient age and dentine thickness. Int Endod J 2010;43:21825. 11. Dietschi D, Duc O, Krejci I, et al. Biomechanical considerations for the restoration of endodontically treated teeth: a systematic review of the literaturePart 1. Composition and micro- and macrostructure alterations. Quintessence Int 2007;38:73343. 12. Randow K, Glantz P. On cantilever loading of vital and non-vital teeth. Acta Odontol Scand 1986;44:2717. 13. Fernandes A, Dessai G. Factors affecting the fracture resistance of post-core reconstructed teeth: a review. Int J Prosthod 2001;14:35563. 14. Standlee JP, Caputo AA, Holcomb JP. The dentatus screw: comparative stress analysis with endodontic dowels designs. J Oral Rehabil 1982;9:2333. 15. Sokol DJ. Effective use of core and post concepts. J Prosthet Dent 1984;52:2314. 16. Silverstein WH. The reinforcement of endodontically treated teeth. Dent Clin North Am 1964;14:37281. 17. Salameh Z, Ounsi HF, Aboushelib MN, et al. Fracture resistance and failure patterns of endodontically treated mandibular molars with and without glass ber post in combination with a zirconia-ceramic crown. J Dent 2008;36:5139. 18. Cagidiaco MC, Garca-Godoy F, Vichi A, et al. Placement of ber prefabricated or custom made posts affects the 3-year survival of endodontically treated premolars. Am J Dent 2008;21:17984. 19. Ferrari M, Cagidiaco MC, Grandini S, et al. Post placement affects survival of endodontically treated premolars. J Dent Res 2007;86:72934. 20. Salameh Z, Sorrentino R, Ounsi HF, et al. Effect of different all-ceramic crown system on fracture resistance and failure pattern of endodontically treated maxillary premolars restored with and without glass ber posts. J Endod 2007;33: 84851. 21. Salameh Z, Sorrentino R, Ounsi HF, et al. The effect of different full-coverage crown systems on fracture resistance and failure pattern of endodontically treated maxillary incisors restored with and without glass ber posts. J Endod 2008;34: 8426. 22. Nam SH, Chang HS, Min KS, et al. Effect of the number of residual walls on fracture resistances, failure patterns, and photoelasticity of simulated premolars restored with or without ber-reinforced composite posts. J Endod 2010;36:297301. 23. Nothdurft FP, Seidel E, Gebhart F, et al. The fracture behavior of premolar teeth with class II cavities restored by both direct composite restorations and endodontic post systems. J Dent 2008;36:4449. 24. Nothdurft FP, Seidel E, Gebhart F, et al. Inuence of endodontic posts on the fracture behavior of crowned premolars with Class II cavities. J Dent 2008;36:28793. 25. Sorensen JA, Martinoff JT. Intra-coronal reinforcement and coronal coverage: a study of endodontically treated teeth. J Prosthet Dent 1984;51:7804. 26. Fuss Z, Lustig J, Katz A, et al. An evaluation of endodontically treated vertical root fractured teeth: impact of operative procedures. J Endod 2001;27:468. 27. Torbjorner A, Karlsson S, Odman PA. Survival rate and failure characteristics for two post designs. J Prosthet Dent 1995;73:43944. 28. Ferrari M, Vichi A, Garca-Godoy F. Clinical evaluation of ber-reinforced epoxy resin posts and cast post and cores. Am J Dent 2000;13:15B8. 29. Soares CJ, Soares PV, de Freitas Santos-Filho PC, et al. The inuence of cavity design and glass ber posts on biomechanical behavior of endodontically treated premolars. J Endod 2008;34:10159. 30. Pilo R, Shapenco E, Lewinstein I. Residual dentin thickness in bifurcated maxillary rst premolars after root canal and post space preparation with parallel-sided drills. J Prosthet Dent 2008;99:26773. 31. Butz F, Lennon A, Heydecke G, et al. Survival rate and fracture strength of endodontically treated maxillary incisors with moderate defects restored with different post and core systems, an in-vitro study. Int J Prosthodont 2001;14:5864. 32. Lertchirakarn V, Palamara JE, Messer HH. Patterns of vertical root fracture: factors affecting stress distribution in the root canal. J Endod 2003;29:5238. 33. Peters MC, Poort HW, Farah JW, et al. Stress analysis of a tooth restored with a post and core. J Dent Res 1983;62:7603. 34. Toparli M. Stress analysis in a post-restored tooth utilizing the nite element method. J Oral Rehabil 2003;30:4706. 35. Fokkinga WA, Le Bell AM, Kreulen CM, et al. Ex vivo fracture resistance of direct resin composite complete crowns with and without posts on maxillary premolars. Int Endod J 2005;38:2307. 36. Mohammadi N, Kahnamoii MA, Yeganeh PK, et al. Effect of ber post and cusp coverage on fracture resistance of endodontically treated maxillary premolars directly restored with composite resin. J Endod 2009;35:142832. 37. Torbjorner A, Fransson B. A literature review on the prosthetic treatment of structurally compromised teeth. Int J Prosthodont 2004;17:36976. 38. Sidoli GE, King PA, Setchell DJ. An in vitro evaluation of a carbon bre-based post and core system. J Prosthet Dent 1997;78:59. 39. Tjan AH, Whang SB. Resistance to root fracture of dowel channels with various thicknesses of buccal dentin walls. J Prosthet Dent 1985;53:496500. 40. Costa LC, Pegoraro LF, Bonfante G. Inuence of different metal restorations bonded with resin of fracture resistance of endodontically treated maxillary premolars. J Prosthet Dent 1997;77:3659. 41. Drummond JL, Bapna MS. Static and cyclic loading of bre-reinforced dental resin. Dent Mater 2003;19:22631. 42. Cohen BI, Pagnillo MK, Newman I, et al. Pilot study of the cyclic fatigue characteristics of ve endodontic posts with four core materials. J Oral Rehabil 2000;27: 8392. 43. Mannocci F, Ferrari M, Watson TF. Microleakage of endodontically treated teeth restored with bre posts and composite cores after cyclic loading: a confocal microscopic study. J Prosthet Dent 2001;85:28491. 44. Martinez-Gonzalez A, Amigo-Borras V, Fons-Font A, et al. Response of three types of cast posts and cores to static loading. Quintessence Int 2001;32:55260. 45. Isidor F, Brondum K, Ravnholt G. The inuence of post length and crown ferrule length on the resistance to cyclic loading of bovine teeth with prefabricated titanium posts. Int J Prosthodont 1999;12:7882. 46. Hayashi M, Sugeta A, Takahashi Y, et al. Static and fatigue fracture resistances of pulpless teeth restored with post-cores. Dent Mater 2008;24:117886. 47. Hu S, Osada T, Shimizu T, et al. Resistance to cyclic fatigue and fracture of structurally compromised root restored with different post and core restorations. Dent Mater J 2005;24:22531. 48. Ross RS, Nicholls JI, Harrington GW. A comparison of strains generated during placement of ve endodontic posts. J Endod 1991;17:4506. 49. Mendoza D, Eakle W, Kahl E, et al. Root reinforcement with a resin bonded preformed post. J Prosthet Dent 1997;78:104. 50. Sorensen JA, Engelman MJ. Effect of post adaptation on fracture resistance of endodontically treated teeth. J Prosthet Dent 1990;64:41924. 51. King P, Setchell D. An in vitro evaluation of a prototype CFRC prefabricated post developed for the restoration of pulpless teeth. J Oral Rehabil 1990;17:599609. 52. Sirimai S, Riis DN, Morgano SM. An in vitro study of the fracture resistance and the incidence of vertical root fracture of pulpless teeth restored with six post-and-core systems. J Prosthet Dent 1999;81:2629. 53. Eskitascioglu G, Belli S, Kalkan M. Evaluation of two post core systems using two different methods: fracture strength test and a nite elemental stress analysis. J Endod 2002;28:62933. 54. Naumann M, Metzdorf G, Fokkinga W, et al. Inuence of test parameters on in vitro fracture resistance of post-endodontic restorations: a structured review. J Oral Rehabil 2009;36:299312. 55. Loney W. Three-dimensional photoelastic stress analysis of the ferrule effect in cast post and core. J Prosthet Dent 1990;63:50612. 56. Assif D, Oren E, Marshak BL, et al. Photoelastic analysis of stress transfer by endodontically treated teeth to the supporting structure using different restorative techniques. J Prosthet Dent 1989;61:53543. 57. Cohen BI, Pagnillo M, Musikant BL, et al. Comparison of the retentive and photoelastic properties of two prefabricated endodontic post systems. J Oral Rehabil 1999;26:48894. 58. Albuquerque Rde C, Polleto LT, Fontana RH, et al. Stress analysis of an upper central incisor restored with different posts. J Oral Rehabil 2003;30:93643. 59. Pegoretti A, Fambri L, Zappini G, et al. Finite element analysis of a glass bre reinforced composite endodontic post. Biomaterials 2002;23:266782. 60. Pierrisnard L, Bohin F, Renault P, et al. Corono-radicular reconstruction of pulpless teeth: a mechanical study using nite element analysis. J Prosthet Dent 2002; 88:4428. 61. Ho MH, Lee SY, Chen HH, et al. Three-dimensional nite element analysis of the effects of posts on stress distribution in dentin. J Prosthet Dent 1994;72:36772. 62. Lewgoy HR, Youssef MN, Matson MR, et al. Finite elements study of the Flexi Post and Flexi Flange post systems in a maxillary central incisor. Pesqui Odontol Bras 2003;17:1326. 63. Lanza A, Aversa R, Rengo S, et al. 3D FEA of cemented steel, glass and carbon posts in a maxillary incisor. Dent Mater 2005;21:70915. 64. Yang HS, Lang LA, Molina A, et al. The effects of dowel design and load direction on dowel-and-core restorations. J Prosthet Dent 2001;85:55867. 65. Asmussen E, Peutzfeldt A, Saha A. Finite element analysis of stresses in endodontically treated, dowel-restored teeth. J Prosthet Dent 2005;94:3219. 66. Khani MM, Tafazzoli-Shadpour M, Aghajani F, et al. Mechanical vulnerability of lower second premolar utilising visco-elastic dynamic stress analysis. Comput Methods Biomech Biomed Engin 2009;12:55361. 67. Guzy GE, Nicholls JI. In vitro comparison of intact endodontically treated teeth with and without endo-post reinforcement. J Prosthet Dent 1979;42:3944. 68. Cailleteau JG, Rieger MR, Akin JE. A comparison of intracanal stresses in a postrestored tooth utilizing the nite element method. J Endod 1992;18:5404. 69. Holmes DC, Diaz-Arnold AM, Leary JM. Inuence of post dimension on stress distribution in dentin. J Prosthet Dent 1996;75:1407.

JOE Volume 36, Number 9, September 2010

Fracture Resistance of Teeth Restored with Post-retained Restorations

1447

Review Article
70. Darendeliler S, Darendeliler H, Kinoglu T. Analysis of a central maxillary incisor by using a three-dimensional nite element method. J Oral Rehabil 1992;19:37183. 71. Kishen A, Ramamurty U, Asundi A. Experimental studies on the nature of property gradients in the human dentine. J Biomed Mater Res 2000;51:6509. 72. Kishen A, Kumar GV, Chen NN. Stress-strain response in human dentine: rethinking fracture predilection in postcore restored teeth. Dent Traumatol 2004;20:90100. 73. Fan P, Nicholls JI, Kois JC. Load fatigue of ve restoration modalities in structurally compromised premolars. Int J Prosthodont 1995;8:21320. 74. Thorsteinsson TS, Yaman P, Craig RG. Stress analyses of four prefabricated posts. J Prosthet Dent 1992;67:303. 75. Davy DT, Dilley GL, Krejci RF. Determination of stress patterns in root-lled teeth incorporating various dowel designs. J Dent Res 1981;60:130110. 76. Buttel L, Krastl G, Lorch H, et al. Inuence of post t and post length on fracture resistance. Int Endod J 2009;42:4753. 77. Leary JM, Aquilino SA, Svare CW. An evaluation of post length within the elastic limits of dentine. J Prosthet Dent 1987;57:27781. 78. Hunter AJ, Feiglin B, Williams JF. Effects of post placement on endodontically treated teeth. J Prosthet Dent 1989;62:16672. 79. Burns DA, Krause WR, Douglas HB, et al. Stress distribution surrounding endodontic posts. J Prosthet Dent 1990;64:4128. 80. Lambjerg-Hansen H, Asmussen E. Mechanical properties of endodontic posts. J Oral Rehabil 1997;24:8827. 81. Schmitter M, Rammelsberg P, Lenz J, et al. Teeth restored using ber-reinforced posts: in vitro fracture tests and nite element analysis. Acta Biomater 2010;6: 374754. 82. Giovani AR, Vansan LP, de Sousa Neto MD, et al. In vitro fracture resistance of glass-ber and cast metal posts with different lengths. J Prosthet Dent 2009; 101:1838. 83. Cecchin D, Farina AP, Guerreiro CA, et al. Fracture resistance of roots prosthetically restored with intra-radicular posts of different lengths. J Oral Rehabil 2010; 37:11622. 84. Nissan J, Dmitry Y, Assif D. The use of reinforced composite resin cement as compensation for reduced post length. J Prosthet Dent 2001;86:3048. 85. Sorensen JA, Engelman MJ. Effect of post adaptation on fracture resistance of endodontically treated teeth. J Prosthet Dent 1990;64:41924. 86. Assif D, Gorl C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994;71:5657. 87. Mattison GD. Photoelastic stress analysis of cast-gold endodontic posts. J Prosthet Dent 1982;48:40711. 88. Halle E, Nicholls J, Hassel V. An in vitro comparison of hollow post and core and a custom hollow post and core. J Endod 1984;10:96100. 89. Mou YB, Chen YM, Smales RJ, et al. Optimum post and tooth root diameters for a cast post-core system. Am J Dent 2009;22:3114. 90. McAndrew R, Jacobsen PH. The relationship between crown and post design on root stressa nite element study. Eur J Prosthodont Restor Dent 2002;10:913. 91. Reinhardt RA, Krejci RF, Pao YC, et al. Dentin stresses in post reconstructed teeth with diminishing bone support. J Dent Res 1983;62:10028. 92. Mentink AG, Creugers NH, Hoppenbrouwers PM, et al. Qualitative assessment of stress distribution during insertion of endodontic posts in photoelastic material. J Dent 1998;26:12531. 93. Deutsch AS, Musikant BL, Cavallari J. Retentive properties of a new post and core system. J Prosthet Dent 1985;53:124. 94. Uddanwadiker RV, Padole PM, Arya H. Effect of variation of root post in different layers of tooth: linear vs nonlinear nite element stress analysis. J Biosci Bioeng 2007;104:36370. 95. Silva NR, Castro CG, Santos-Filho PC, et al. Inuence of different post design and composition on stress distribution in maxillary central incisor: nite element analysis. Indian J Dent Res 2009;20:1538. 96. Signore A, Benedicenti S, Kaitsas V, et al. Long-term survival of endodontically treated, maxillary anterior teeth restored with either tapered or parallel-sided glass-ber posts and full-ceramic crown coverage. J Dent 2009;37:11521. 97. Martinez-Insua A, da Silva L, Rilo B, et al. Comparison of the fracture resistances of pulpless teeth restored with a cast post and core or carbon-bre post with a composite core. J Prosthet Dent 1998;80:52732. 98. Fokkinga WA, Kreulen CM, Vallittu PK, et al. A structured analysis of in vitro failure loads and failure modes of ber, metal, and ceramic post-and-core systems. Int J Prosthodont 2004;17:47682. 99. Al-Omiri MK, Al-Wahadni AM. An ex vivo study of the effects of retained coronal dentine on the strength of teeth restored with composite core and different post and core systems. Int Endod J 2006;39:8909. 100. Bonfante G, Kaizer OB, Pegoraro LF, et al. Fracture strength of teeth with ared root canals restored with glass bre posts. Int Dent J 2007;57:15360. 101. Al-Wahadni AM, Hamdan S, Al-Omiri M, et al. Fracture resistance of teeth restored with different post systems: in vitro study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:e7783. 102. Ozcan M, Valandro LF. Fracture strength of endodontically-treated teeth restored with post and cores and composite cores only. Oper Dent 2009;34:42936. 103. McLaren JD, McLaren CI, Yaman P, et al. Dennison JD, McDonald NJ. The effect of post type and length on the fracture resistance of endodontically treated teeth. J Prosthet Dent 2009;101:17482. 104. Saupe WA, Gluskin AH, Radke RA Jr. A comparative study of fracture resistance between morphologic dowel and cores and a resin-reinforced dowel system in the intraradicular restoration of structurally compromised roots. Quintessence Int 1996;27:48391. 105. Akkayan B, Caniklioglu B. Resistance to fracture of crowned teeth restored with different post systems. Eur J Prosthodont Rest Dent 1998;6:138. 106. Hajizadeh H, Namazikhah MS, Moghaddas MJ, et al. Effect of posts on the fracture resistance of load-cycled endodontically-treated premolars restored with direct composite resin. J Contemp Dent Pract 2009;10:107. 107. Spazzin AO, Galafassi D, de Meira-Junior AD, et al. Inuence of post and resin cement on stress distribution of maxillary central incisors restored with direct resin composite. Oper Dent 2009;34:2239. 108. Caputo AA, Standlee JP. Biomechamics in Clinical Dentistry. Chicago: Quintessence Publishing Co; 1987:185203. 109. Bitter K, Kielbassa AM. Post-endodontic restorations with adhesively luted berreinforced composite post systems: a review. Am J Dent 2007;20:35360. 110. Cagidiaco MC, Goracci C, Garcia-Godoy F, et al. Clinical studies of ber posts: a literature review. Int J Prosthodont 2008;21:32836. 111. Grandini S, Goracci C, Tay FR, et al. Clinical evaluation of the use of ber posts and direct resin restorations for endodontically treated teeth. Int J Prosthodont 2005; 18:399404. 112. Ferrari M, Cagidiaco MC, Goracci C, et al. Long-term retrospective study of the clinical performance of ber posts. Am J Dent 2007;20:28791. 113. Piovesan EM, Demarco FF, Cenci MS, et al. Survival rates of endodontically treated teeth restored with ber-reinforced custom posts and cores: a 97-month study. Int J Prosthodont 2007;20:6339. 114. Naumann M, Sterzenbac G, Alexandra F, et al. Randomized controlled clinical pilot trial of titanium vs. glass ber prefabricated posts: preliminary results after up to 3 years. Int J Prosthodont 2007;20:499503. 115. Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. J Prosthet Dent 2002;87:4317. 116. Hayashi M, Takahashi Y, Imazato S, et al. Fracture resistance of pulpless teeth restored with post-cores and crowns. Dent Mater 2006;22:47785. 117. Kivanc BH, Gorgul G. Fracture resistance of teeth restored with different post systems using new-generation adhesives. J Contemp Dent Pract 2008;9:3340. 118. Cormier CJ, Burns DR, Moon P. In vitro comparison of the fracture resistance and failure mode of bre, ceramic, and conventional post systems at various stages of restoration. J Prosthodont 2001;10:2636. 119. Newman MP, Yaman P, Dennison J, et al. Fracture resistance of endodontically treated teeth restored with composite posts. J Prosthet Dent 2003;89:3607. 120. Toksavul S, Toman M, Uyulgan B, et al. Effect of luting agents and reconstruction techniques on the fracture resistance of pre-fabricated post systems. J Oral Rehabil 2005;32:43340. 121. Stockton LW, Williams PT. Retention and shear bond strength of two post systems. Oper Dent 1999;24:2106. 122. Hu YH, Pang LC, Hsu CC, et al. Fracture resistance of endodontically treated anterior teeth restored with four post-and-core systems. Quintessence Int 2003;34: 34953. 123. Fokkinga WA, Kreulen CM, Le Bell-Ronnlof AM, et al. In vitro fracture behavior of maxillary premolars with metal crowns and several post-and-core systems. Eur J Oral Sci 2006;114:2506. 124. Toman M, Toksavul S, Sarikanat M, et al. Fracture resistance of endodontically treated teeth: effect of tooth coloured post material and surface conditioning. Eur J Prosthodont Restor Dent 2010;18:2330. 125. Pameijer CH, Glantz P, Mobasherat MA. On clinical corrosion of pins. Swed Dent J 1983;7:1617. 126. Vano M, Carvalho C, Sedda M, et al. The inuence of storage condition and duration on the resistance to fracture of different ber post systems. Am J Dent 2009;22: 36670. 127. Goracci C, Fabianelli A, Sadek FT, et al. The contribution of friction to the dislocation resistance of bonded ber posts. J Endod 2005;31:60812. 128. Turner CH. Post retained crown failure: a survey. Dent Update 1982;9: 193202. 129. Santos AF, Tanaka CB, Lima RG, et al. Vertical root fracture in upper premolars with endodontic posts: nite element analysis. J Endod 2009;35:11720. 130. Bex RT, Parker MW, Judkins JT, et al. Effect of dentinal bonded resin post-core preparations on resistance to vertical root fracture. J Prosthet Dent 1992;67: 76872. 131. Chan RW, Bryant RW. Post-core foundations for endodontically treated posterior teeth. J Prosthet Dent 1982;48:4016.

1448

Khaled AL-Omiri et al.

JOE Volume 36, Number 9, September 2010

Review Article
132. Assif D, Bitenski A, Pilo R, et al. Effect of post design on resistance to fracture of endodontically treated teeth with complete crowns. J Prosthet Dent 1993;69:3640. 133. Gelfand M, Goldman M, Sunderman EJ. Effect of complete veneer crowns on the compressive strength of endodontically treated posterior teeth. J Prosthet Dent 1984;52:6358. 134. Pilo R, Cardash HS, Levin E, et al. Effect of core stiffness on the in vitro fracture of crowned, endodontically treated teeth. J Prothet Dent 2002;88:3026. 135. Sorensen JA, Engelman MJ. Ferrule design and fracture resistance of endodontically treated teeth. J Prosthet Dent 1990;63:52936. 136. Xible AA, de Jesus Tavarez RR, de Araujo Cdos R, et al. Effect of silica coating and silanization on exural and composite-resin bond strengths of zirconia posts: An in vitro study. J Prosthet Dent 2006;95:2249. 137. Akgungor G, Sen D, Aydin M. Inuence of different surface treatments on the shortterm bond strength and durability between a zirconia post and a composite resin core material. J Prosthet Dent 2008;99:38899. 138. Aksornmuang J, Nakajima M, Foxton RM, et al. Regional bond strengths of a dualcure resin core material to translucent quartz ber post. Am J Dent 2006;19:515. 139. Yenisey M, Kulunk S. Effects of chemical surface treatments of quartz and glass ber posts on the retention of a composite resin. J Prosthet Dent 2008;99:3845. 140. Monticelli F, Osorio R, Sadek FT, et al. Surface treatments for improving bond strength to prefabricated ber posts: a literature review. Oper Dent 2008;33: 34655. 141. Bitter K, Neumann K, Kielbassa AM. Effects of pretreatment and thermocycling on bond strength of resin core materials to various ber-reinforced composite posts. J Adhes Dent 2008;10:4819. 142. Radovic I, Monticelli F, Goracci C, et al. The effect of sandblasting on adhesion of a dual-cured resin composite to methacrylic ber posts: microtensile bond strength and SEM evaluation. J Dent 2007;35:496502. 143. Hattori M, Takemoto S, Yoshinari M, et al. Durability of ber-post and resin core build-up systems. Dent Mater J 2010;29(2):2248. 144. Artopoulou II, OKeefe KL, Powers JM. Effect of core diameter and surface treatment on the retention of resin composite cores to prefabricated endodontic posts. J Prosthodont 2006;15:1729. 145. Akis I, Ozcan M, Nergiz I. Effect of surface conditioning techniques on the resisxli tance of resin composite core materials on titanium posts. Quintessence Int 2003; 34:76671. 146. Rosen H. Operative procedure in mutilated endodontically treated teeth. J Prosthet Dent 1961;11:97386. 147. Barkhordar RA, Radke R, Abbasi J. Effect of metal collars on resistance of endodontically treated teeth to root fracture. J Prosthet Dent 1989;61:6768. 148. Loney RW, Moulding MB, Ritsco RG. The effect of load angulation on fracture resistance of teeth restored with cast posts and cores and crowns. Int J Prosthodont 1995;8:24751. 149. Akkayan B. An in vitro study evaluating the effect of ferrule length on fracture resistance of endodontically treated teeth restored with bre reinforced and zirconia dowel systems. J Prosthet Dent 2004;92:15562. 150. Tan PL, Aquilino SA, Gratton DG, et al. In vitro fracture resistance of endodontically treated central incisors with varying ferrule heights and congurations. J Prosthet Dent 2005;93:3316. 151. Stankiewicz NR, Wilson PR. The ferrule effect: a literature review. Int Endod J 2002; 35:57581. 152. Libman WJ, Nicholls JI. Load fatigue of teeth restored with cast posts and cores and complete crowns. Int J Prosthodont 1995;8:15561. 153. Dorriz H, Alikhasi M, Mirfazaelian A, et al. Effect of ferrule and bonding on the compressive fracture resistance of post and core restorations. J Contemp Dent Pract 2009;10:18. 154. Al-Hazaimeh N, Gutteridge DL. An in vitro study into the effect of the ferrule preparation on the fracture resistance of crowned teeth incorporating prefabricated post and composite core restorations. Int Endod J 2001;34:406. 155. Mezzomo E, Massa F, Libera SD. Fracture resistance of teeth restored with two different post-and-core designs cemented with two different cements: an in vitro study. Part I. Quintessence Int 2003;34:3016. 156. Ng CC, al-Bayat MI, Dumbrigue HB, et al. Effect of no ferrule on failure of teeth restored with bonded posts and cores. Gen Dent 2004;52:1436. 157. Naumann M, Preuss A, Rosentritt M. Effect of incomplete crown ferrules on load capacity of endodontically treated maxillary incisors restored with ber posts, composite build-ups, and all-ceramic crowns: an in vitro evaluation after chewing simulation. Acta Odontol Scand 2006;64:316. 158. Gegauff AG. Effect of crown lengthening and ferrule placement on static load failure of cemented cast post-cores and crowns. J Prosthet Dent 2000;84:16979. 159. Mitchell CA. Selection of materials for post cementation. Dent Update 2000;27: 3504. 160. Junge T, Nicholls JI, Phillips KM, et al. Load fatigue of compromised teeth: a comparison of three luting cements. Int J Prosthodont 1998;11:55864. 161. Naumann M, Sterzenbach G, Rosentritt M, et al. Is adhesive cementation of endodontic posts necessary? J Endod 2008;34:100610. 162. Dallari A, Rovatti L. Six years of in vitro/in vivo experience with Composipost. Compend Contin Educ Dent 1996;17:S5763. 163. Tay FR, Pashley DH. Monoblocks in root canals: a hypothetical or a tangible goal. J Endod 2007;33:3918. 164. Dietschi D, Duc O, Krejci I, et al. Biomechanical considerations for the restoration of endodontically treated teeth: a systematic review of the literature, Part II (Evaluation of fatigue behavior, interfaces, and in vivo studies). Quintessence Int 2008; 39:11729. 165. Hammad M, Qualtrough A, Silikas N. Effect of new obturating materials on vertical root fracture resistance of endodontically treated teeth. J Endod 2007;33:7326. 166. Vire DE. Failure of endodontically treated teeth: classication and evaluation. J Endod 1991;17:33842. 167. Aquilino SA, Caplan DJ. Relationship between crown placement and the survival of endodontically treated teeth. J Prosthet Dent 2002;87:25663. 168. Joshi S, Mukherjee A, Kheur M, et al. Mechanical performance of endodontically treated teeth. Finite Elem Anal Des 2001;37:587601. 169. Heydecke G, Butz F, Strub JR. Fracture strength and survival rate of endodontically treated maxillary incisors with approximal cavities after restoration with different post and core systems: an in-vitro study. J Dent 2001;29:42733. 170. DArcangelo C, De Angelis F, Vadini M, et al. In vitro fracture resistance and deection of pulpless teeth restored with ber posts and prepared for veneers. J Endod 2008;34:83841. 171. DArcangelo C, De Angelis F, Vadini M, DAmario M, Caputi S. Fracture resistance and deection of pulpless anterior teeth restored with composite or porcelain veneers. J Endod 2010;36(1):1536. 172. Al-Wahadni A, Gutteridge DL. An in vitro investigation into the effects of retained coronal dentine on the strength of a tooth restored with a cemented post and partial core restoration. Int Endod J 2002;35:9138. 173. Sorrentino R, Monticelli F, Goracci C, et al. Effect of post-retained composite restorations and amount of coronal residual structure on the fracture resistance of endodontically-treated teeth. Am J Dent 2007;20:26974. 174. Patel A, Gutteridge DL. An in vitro investigation of cast post and partial core design. J Dent 1996;24:2817. 175. Bergman B, Lundquist P, Sjogren U, et al. Restorative and endodontic results after treatment with cast posts and cores. J Prosthet Dent 1989;61:105. 176. Bender IB, Freedland JB. Adult root fracture. J Am Dent Assoc 1983;107:4139. 177. Mehta SB, Millar BJ. A comparison of the survival of bre posts cemented with two different composite resin systems. Br Dent J 2008;205:E23. 178. Yaman SD, Karacaer O, Sahin M. Stress distribution of post-core applications in maxillary central incisors. J Biomater Appl 2004;18:16377. 179. Abu-Hammad O, Dar-Odeh N. Factors affecting the stress picture in a molar tooth restored with an endodontic post retained lling. Part 1: the stress picture with the system under various loading conditions. Cairo Dent J 2004;20:1214. 180. Pao YC, Reinhardt RA, Krejci RF. Root stresses with tapered end post design in periodontally compromised teeth. J Prosthet Dent 1987;57:2816. 181. Kalachev YS. Analysis of stresses in hard dental tissues generated by clinical application of retentive root canal posts. Folia Med (Plovdiv) 2004;46:426. 182. Meira JB, Esposito CO, Quitero MF, et al. Elastic modulus of posts and the risk of root fracture. Dent Traumatol 2009;25:3948. 183. Nyman S, Lindhe J. A longitudinal study of combined periodontal and prosthetic treatment of patients with advanced periodontal disease. J Periodontol 1979;50: 1639. 184. Naumann M, Rosentritt M, Preuss A, et al. The effect of alveolar bone loss on the load capability of restored endodontically treated teeth: a comparative in vitro study. J Dent 2006;34:7905. 185. Hunter AJ, Flood AM. The restoration of endodontically treated teeth. Part 2. Posts. Aust Dent J 1989;34:512. 186. Chen XM, Wu XH, Niu L, et al. Effects of the loss of alveolar bone on the areas of periodontal ligament, mechanical fulcrum and fracture resistance of root and postcore system. Sichuan Da Xue Xue Bao Yi Xue Ban 2005;36:8537.

JOE Volume 36, Number 9, September 2010

Fracture Resistance of Teeth Restored with Post-retained Restorations

1449

CONSORT Randomized Clinical Trial

The Effect of Premedication with Ibuprofen and Indomethacin on the Success of Inferior Alveolar Nerve Block for Teeth with Irreversible Pulpitis
Masoud Parirokh, DDS, MSc,* Rezvan Ashouri, DMD, Ali Reza Rekabi, DMD, Nouzar Nakhaee, MD, Abbas Pardakhti, PhD, Sara Askarifard, DMD, MS,* and Paul V. Abbott, DDS, PhDk
Abstract
Introduction: Achieving pulp anesthesia with irreversible pulpitis is difcult. This study evaluated whether nonsteroidal anti-inammatory drugs assist local anesthesia. Methods: In a randomized double-blinded clinical trial, 150 patients (50 per group) with irreversible pulpitis were given placebo, 600 mg ibuprofen, or 75mg indomethacin 1 hour before local anesthesia. Each patient recorded their pain score on a visual analog scale before taking the medication, 15 minutes after anesthesia in response to a cold test, during access cavity preparation and during root canal instrumentation. No or mild pain at any stage was considered a success. Data were analyzed by the chi-square and analysis of variance tests. Results: Overall success rates for placebo, ibuprofen, and indomethacin were 32%, 78%, and 62%, respectively (p < 0.001). Ibuprofen and indomethacin were signicantly better than placebo (p < 0.01). There was no difference between ibuprofen and indomethacin (p = 0.24). Conclusions: Premedication with ibuprofen and indomethacin signicantly increased the success rates of inferior alveolar nerve block anesthesia for teeth with irreversible pulpitis. (J Endod 2010;36:14501454)

Key Words
Anesthesia, ibuprofen, indomethacin, inferior alveolar nerve block, irreversible pulpitis, nonsteroidal antiinammatory drugs, placebo

ain control particularly during the early phases of endodontic treatment is of paramount importance and makes both the dentist and the patient condent and comfortable for the remainder of the treatment (1). The inferior alveolar nerve block (IANB) is the conventional method for anesthetizing mandibular molar teeth (2, 3). Research has shown that gaining anesthesia in mandibular molars with irreversible pulpitis is much more difcult in comparison to the teeth with normal healthy pulps (46). Some investigations have been performed to overcome pulp pain that remains despite having had an IANB injection (2, 3, 714). Numerous investigations have been performed to increase the success rate of anesthesia during dental, and particularly endodontic, procedures such as the use of various anesthetic techniques and solutions as well as pretreatment with analgesics (725). The concept of using preoperative analgesic drugs to increase the effectiveness of IANB is based on reports of their benecial effects on reducing postoperative pain (23). Previous investigations using analgesics before administering IANB have reported conicting results (10, 2325). For example, Modaresi et al (23) reported signicant improvements in the success rate of IANB in teeth with inamed pulps after the use of analgesics, and Ianiro et al (10) reported higher success rates although they were not signicantly different. In contrast, two separate studies reported no signicant difference in IANB success rates when the patients were premedicated with analgesics (24, 25). Several reasons could explain these promising but not completely different results such as an insufcient number of subjects (10) and a lack of similarity of methods and clinical conditions (23). The aim of this study was to compare two types of nonsteroidal anti-inammatory (NSAID) medication (ibuprofen and indomethacin) with a placebo regarding their effects on the success rates of IANB for endodontic treatment of mandibular molar teeth with irreversible pulpitis.

Materials and Methods

This study was approved by the Ethics Committee of Kerman University of Medical Sciences in Iran (no. KA/ 88-05 and Iranian Registry of Clinical Trials ID no. 138803192016N1). The calculation of sample size was based on a medium effect size of 0.5, an a error of 0.05, and power at 0.8. The sample size was calculated with PASS software version 6.0 (NCSS, Kaysville, UT). Based on sample size calculations, up to 50 patients were required in each group. The following inclusion and exclusion criteria were used for this study. The exclusion criteria were the presence of systemic disorders, a sensitivity to lidocaine with 1:80,000 epinephrine, a sensitivity to NSAIDs, nasal polyp, a history of gastrointestinal ulcers, the presence of widening of the periodontal ligament space, the presence of a periapical radiolucency, lactation, pregnancy, using any type of analgesic medication in the preceding 12 hours before the treatment, having a tooth not suitable for restoration, and having a full crown. Inclusion criteria included healthy patients having a rst or second mandibular molar tooth with irreversible pulpitis and normal periapical radiographic appearance. The clinical diagnosis of irreversible pulpitis was conrmed by a response to an electric pulp test (The Element Diagnostic Unit; SybronEndo, Glendora, CA) and a prolonged

From the *Oral and Dental Research Center, Endodontic Department, Neuroscience Research Center, and Pharmaceutics Research Center, Kerman University of Medical Sciences, Kerman, Iran; and kSchool of Dentistry, University of Western Australia, Perth, Australia. Supported by the Neuroscience Research Center, Kerman University of Medical Sciences. Address requests for reprints to Dr Masoud Parirokh, Oral and Dental Research Center, Kerman University of Medical Sciences, Kerman, Iran. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.007

1450

Parirokh et al.

JOE Volume 36, Number 9, September 2010

CONSORT Randomized Clinical Trial

exaggerated response (10 seconds) with moderate to severe pain to a cold test (Roeko Endo-Frost; Roeko, Langenau, Germany) after the size 2 cotton pellet was removed. One hundred fty patients were eligible to participate in this prospective, randomized double-blind study. All patients were adults over 18 years of age, and they were treated in the postgraduate clinic of the Endodontic Department of Kerman Dental School in Iran from September 2009 to February 2010. Informed consent of all subjects who participated in this study was obtained after the nature of the procedure and the possible discomforts and risks had been fully explained. All patients who agreed to participate in the study were randomly divided into three groups of 50 patients each. In order to randomize the patients, the numbers of patients in each group were written on paper, and each one was kept in a sealed opaque envelope. Each patient was asked to choose one of the envelopes and based on the number was assigned to one of the groups. A pharmacist (AP) was responsible for preparing the capsules and encoding them to ensure that neither the clinicians nor the patients were aware of the placebo or NSAIDs medication being taken. All patients were given a capsule of the same color and size 1 hour before starting the treatment. The capsules contained either a placebo of lactose powder (placebo group), 600 mg ibuprofen (ibuprofen group) (Hakim Pharmaceutical Co, Tehran, Iran), or 75mg indomethacin (indomethacin group) (Hakim Pharmaceutical Co). Before taking the capsules, the patients were asked to rate their pain using a HeftParker visual analog pain scale (VAS) after the cold test. The VAS scores were divided into four categories. No pain corresponded to 0 mm, mild pain was dened as being >0 mm and $54 mm, moderate pain was dened as being >54 mm and <114 mm, and severe pain was dened as being $114 mm. A side-loading cartridge syringe (Dena Instruments; Forgeman Instruments Co, Sialkot, Pakistan) was used to administer the injections. The syringe was equipped with a blood aspiration device and a thumb ring. A 27-G 38-mm needle (Carpule; Heraeus Kulzer GmbH, Hanau, Germany) was tted to the syringe. A blood aspiration test was performed before the IANB was administered. In all patients, after needle insertion and based on a standard IANB method (26), when bone contact was established, the needle was withdrawn 1 to 2 mm and aspiration was performed. If the aspiration was negative for blood, then the anesthetic solution (2% lidocaine with 1/80000 epinephrine; Darupakhsh, Tehran, Iran) was injected. One cartridge (1.8 mL) was used in each patient. Fifteen minutes after administering anesthesia, each patient was asked about any signs of soft-tissue anesthesia. If the patient did not report profound lip numbness, then the IANB was considered to be a failure, and the patient was excluded from the study. In IANB successful patients, the teeth were again tested with the same cold pulp sensibility test, and the patients were asked to rate their pain using the Heft-Parker VAS. Then, the teeth were isolated with a rubber dam, and endodontic access cavity preparation was commenced. Access cavity preparation was only started in patients who reported lip numbness after administration of the anesthetic.
TABLE 1. Comparison of Baseline Variables between Groups Characteristic
Age Sex (%) Male Female Before-anesthesia pain score
*Signicant difference between groups 1 and 3.

The patients were instructed to rate any pain experienced during each step of access cavity preparation. This was recorded as pain within dentin, when entering the pulp chamber, or when a le was inserted into the root canals. No or mild pain (faint, weak, and mild pain) were classied as success, whereas moderate and severe pain were classied as the failure of anesthesia. If a patient reported sensitivity to the cold test before starting the access cavity preparation or at any time during subsequent treatment, another method of anesthesia was then used in order to be able to continue with the treatment. All patients were monitored for 48 hours after the procedure. Chi-square and ANOVA tests were used to compare qualitative and quantitative data between the three groups. Because there were differences in VAS scores before medication among the three groups, a repeated-measure ANCOVA was used to compare variations in VAS scores. The comparisons were considered signicant if p < 0.05.

Results
None of the patients reported any side effects for up to 48 hours after the procedures. Table 1 shows the distribution of the baseline characteristics of the participants in the present study. The youngest and the oldest participants were 18 and 64 years old, respectively. No signicant differences were found between sex or age among the patients in the three groups (p > 0.05). The only signicant difference was the degree of preoperative pain before using medications (Table 1) with the patients in the indomethacin group showing signicantly higher pain scores when they initially presented for treatment (p < 0.03). The average pain score 1 hour after the medication was taken based on the Heft-Parker VAS were not signicantly different among the three groups (p = 0.540) (Fig. 1). However, after adjusting for baseline VAS scores, there was a signicant difference in the amount of pain relief between patients who were given placebo and those who took either indometacin or ibuprofen (p < 0.001). All patients reported lip numbness, which was assumed to be a sign of IANB success. The overall success rates for the placebo, ibuprofen, and indomethacin groups were 32%, 78%, and 62%, respectively (p < 0.001) (Table 2). The ibuprofen and indomethacin groups showed signicantly higher success rates compared with the placebo group (p < 0.01). There was no signicant difference between ibuprofen and indomethacin (p = 0.24). Overall, 64 of the 150 patients failed to show complete success of anesthesia throughout the present study (placebo group = 34, ibuprofen group = 11, and indomethacin group = 19). Thirty of these patients showed sensitivity to the cold pulp test 15 minutes after anesthesia. Distribution of the failure and success of anesthesia during different stages of the endodontic procedures is shown in Figure 2 for all groups.

Discussion
The results of the present study have shown that premedication with two types of NSAIDs (ibuprofen and indomethacin) signicantly

Group 1 (n = 50)
25.9 (6.0) 22 (44.0) 28 (56.0) 106.3 (24.9)

Group 2 (n = 50)
26.2 (9.5) 26 (52.0) 24 (48.0) 114.4 (27.0)

Group 3 (n = 50)
26.2 (5.7) 23 (46.0) 27 (54.0) 120.2 (26.5)

p value
0.962 0.706 0.030*

JOE Volume 36, Number 9, September 2010

Ibuprofen vs Indomethacin for Inferior Alveolar Nerve Block

1451

CONSORT Randomized Clinical Trial

TABLE 2. Comparison of Success Rates between the Three Groups End result Group
1 2 3

Success (%)
16 (32) 39 (78) 31 (62)

Failure (%)
34 (68) 11 (22) 19 (38)

p value*
0.001

*Chi-square test. There was a signicant difference between groups 1 and 2 and between groups 1 and 3 (p < 0.01) using pair-wise comparisons with the Bonferroni correction.

Figure 1. Severity of pain scores at different stages: before and 1 hour after medication and 15 minutes after the administration of local anesthesia. This shows a higher pain rating before starting treatment in the indomethacin group (sample size in each group = 50).

increased the success rate of IANB anesthesia for mandibular molar teeth with irreversible pulpitis. The demographic characteristics of the patients, such as sex and age, were not signicantly different in all three groups (Table 1). Therefore, these factors did not inuence the results. In contrast, the premedication pain scores were signicantly higher in the indomethacin group than in the other groups (p < 0.05). No signicant difference in pain score among the three groups before anesthesia conrms the potential of indomethacin as an effective adjunctive analgesic for pain during endodontic treatment of teeth with irreversible pulpitis.

In the present study, of the 120 patients who had no response to the cold pulp test 15 minutes after anesthesia, 34 reported pain during access cavity preparation or during instrumentation of the root canals. Similar ndings have been reported in previous studies, which conrms that no response to a cold test is not a guarantee that complete pulp anesthesia has been achieved and that no pain will be felt during root canal treatment (5, 10, 15). Previous studies that have assessed the effectiveness of anesthesia after premedication have used either an electric pulp test or a VAS (10, 2325) to rate pain. In the present study, the Heft-Parker VAS was used to assess patient pain before and after local anesthetic injection. Most of the previous investigations on the efcacy of anesthetic techniques and solutions have used the same method of evaluation (9, 12, 13, 24, 25, 27, 28). In the present study, 2% lidocaine with 1:80000 epinephrine was used because this is a common anesthetic solution in dental practice plus most of the previous investigations have used the same solution. The latter allows comparison between studies of the effects of different techniques and/or medications on pulp anesthesia (2, 4, 7, 9, 10, 12, 13, 18, 22, 23). Signicantly higher amounts of prostaglandins in inamed pulps compared with normal pulps have been reported (29). Prostaglandins can affect tetrodotoxin-resistant receptors and decrease nerve responses to anesthetic agents (30, 31). Previous investigations have described the anti-inammatory effects of ibuprofen and indomethacin

Figure 2. Frequency distribution of the stages of failure and success among the three groups (placebo, ibuprofen, and indomethacin) based on patients reported pain after anesthesia and during access preparation (sample size in each group = 50).

1452

Parirokh et al.

JOE Volume 36, Number 9, September 2010

CONSORT Randomized Clinical Trial

(32, 33). Several subtypes of sodium channels play important roles in mediating inammatory pain, such as Nan 1.7, Nan 1.8, and Nan 1.9 (34). Gould et al (33) in their animal study showed that prostaglandins play an important role in sodium channel augmentation during inammation and that pretreatment with ibuprofen prevented up-regulation of Nan1.7 and Nan1.8 sodium channels. In particular, the effects of Nan1.7 were greater. Ibuprofen has been used in previous investigations for pretreatment or posttreatment analgesia (10, 23). Seymour and Ward (35) evaluated various doses of ibuprofen (200 mg, 400 mg, and 600 mg) for the management of postoperative dental pain, and they reported a trend of higher pain relief in patients who had taken 600-mg doses. Indomethacin has not been commonly used or recommended in endodontic therapy. However, it has strong anti-inammatory effects, and, therefore, it was used as one of the analgesics in the present study. Indomethacin is an NSAID that is used for the management of moderate to severe muscular and joint pain (36). However, indomethacin has several side effects that should be considered before prescribing this medication (36). Shrestha et al (37), in a randomized clinical trial, reported that both indomethacin and ketorolac provided similar pain relief for patients with gouty arthritis. Ketorolac has been successfully used to manage endodontic pain (38, 39), and, therefore, it could be extrapolated that indomethacin may have similar effects for endodontic pain. In the present study, only one single dose of indomethacin was prescribed in order to minimize the possibility of any side effects for the patients. In addition, the patients were monitored for 48 hours, and there were no reports of any side effects. This may either show a reasonable case selection or that a single dose of either medication is unlikely to cause signicant problems for the patients. Despite the previously mentioned efcacy of indomethacin, ibuprofen may be a better choice because it not only showed signicant effects on the success rate of anesthesia, but it also has fewer side effects than indomethacin. Several other investigations have evaluated the efcacy of premedication on IANB success rates for teeth with irreversible pulpitis (10, 2325). Modaresi et al (23) reported signicantly lower sensitivity to electric pulp tests after premedication with either 400 mg ibuprofen or a combination of 600 mg acetaminophen with 40 mg codeine compared with the placebo group. In other studies, the root canal treatment was commenced, and the patients responses during treatment were evaluated, which is more similar to clinical situations (10, 24, 25). Hence, in the present study, endodontic treatment was performed, and the patients pain experiences during treatment were evaluated in addition to using a cold pulp sensibility test. Ianiro et al (10) studied 40 patients with irreversible pulpitis, and they reported a trend toward greater success with an IANB after pretreatment with analgesic drugs although they could not show a significant difference between the placebo and analgesic groups. In contrast, in the present study of 150 patients, both types of NSAIDs showed significantly higher success rates compared with the placebo group. The difference in statistical signicance between these two studies is likely to be a result of the difference in the number of patients, and this emphasizes the need to include as many subjects as possible in such studies. In contrast to the previously mentioned studies, two other investigations have reported no signicant differences between ketorolac, 400 mg ibuprofen, and 800 mg ibuprofen for patients with irreversible pulpitis (24, 25). The differences between the inclusion criteria for the various studies may be a further reason for conicting results. In the current study, as well as the study by Ianiro et al (10), only patients with irreversible pulpitis conrmed by a prolonged pain in response to a cold pulp test were selected for treatment. In contrast, in both the Aggarwal et al (24) and Oleson et al. (25) studies, patients were selected on the basis of having spontaneous pain, and they had been referred to an emergency clinic for endodontic treatment. The degree and duration of pulp inammation when a patient presents for treatment may be another factor contributing to conicting results regarding the efcacy of the pretreatment use of analgesics (10, 2325). In an animal study, Modaresi et al (40) showed that inamed dental pulp is resistant to local analgesic. Based on the results of the present study and previous investigations (10, 23-25), it can be hypothesized that premedication with an analgesic in emergency patients is not effective because the previously released prostaglandins have already resulted in the formation of TTx-resistant receptors, and, therefore, the analgesic has no signicant adjunctive effect on traditional anesthetic techniques such as an IANB. In contrast, the use of premedication in patients who have had prolonged pain to cold without spontaneous pain may improve the effectiveness of IANB anesthesia. In the current investigation, the overall success rates for the placebo, ibuprofen, and indomethacin groups were 32%, 78%, and 62%, respectively. These gures are similar to the study by Ianiro et al (10) who reported 46.2%, 71.4%, and 76.9% success rates for placebo, acetaminophen, and a combination of acetaminophen with ibuprofen, respectively. The current study suggests that combinations of analgesics are not required and that the use of only one antiinammatory drug is just as effective as when combined with acetaminophen. Although more effective anesthesia was obtained when an NSAID analgesic was prescribed before treatment, neither drug provided complete anesthesia during endodontic treatment of teeth with irreversible pulpitis. Therefore, dentists should also be prepared to use other supplementary anesthetic techniques to overcome patients discomfort whenever encountered pain in treating teeth with irreversible pulpitis (16, 21). In conclusion, the results of the present study support the use of NSAID premedication before administering local anesthesia in patients with irreversible pulpitis if there is no spontaneous pain reported by the patient.

References
1. Walton RE, Reader A, Nusstein JM. Local anesthesia. In: Torabinejad M, Walton RE, eds. Endodontics, Principles and Practice. 4th ed. St Louis, MO: Saunders Elsevier; 2008:12947. 2. McLean C, Reader A, Beck M, et al. An evaluation of 4% prilocaine and 3% mepivacaine compared with 2% lidocaine (1:100,000 epinephrine) for inferior alveolar nerve block. J Endod 1993;19:14650. 3. Yared GM, Dagher FB. Evaluation of lidocaine in human inferior alveolar nerve block. J Endod 1997;23:5758. 4. Tortamano IP, Siviero M, Costa CG, et al. A comparison of the anesthetic efcacy of articaine and lidocaine in patients with irreversible pulpitis. J Endod 2009;35: 1658. 5. Reisman D, Reader A, Nist R, et al. Anesthetic efcacy of the supplemental intraosseous injection of 3% mepivacaine in irreversible pulpitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;84:67682. 6. Hargreaves KM, Keiser K. Local anesthetic failure in endodontics: mechanisms and management (a review). Endod Topics 2002;1:2639. 7. Robertson D, Nusstein J, Reader A, et al. The anesthetic efcacy of articaine in buccal inltration of mandibular posterior teeth. J Am Dent Assoc 2007;138: 110412. 8. Haase A, Reader A, Nusstein J, et al. Comparing anesthetic efcacy of articaine versus lidocaine as a supplemental buccal inltration of the mandibular rst molar after an inferior alveolar nerve block. J Am Dent Assoc 2008;139:122835. 9. Bigby J, Reader A, Nusstein J, et al. Anesthetic efcacy of lidocaine/meperidine for inferior alveolar nerve blocks in patients with irreversible pulpitis. J Endod 2007;33: 710. 10. Ianiro SR, Jeansonne BG, McNeal SF, et al. The effect of preoperative acetaminophen or a combination of acetaminophen and Ibuprofen on the success of inferior alveolar nerve block for teeth with irreversible pulpitis. J Endod 2007;33:114. 11. Corbett IP, Kanaa MD, Whitworth JM, et al. Articaine inltration for anesthesia of mandibular rst molars. J Endod 2008;34:5148.

JOE Volume 36, Number 9, September 2010

Ibuprofen vs Indomethacin for Inferior Alveolar Nerve Block

1453

CONSORT Randomized Clinical Trial

12. Mikesell P, Nusstein J, Reader A, et al. A comparison of articaine and lidocaine for inferior alveolar nerve blocks. J Endod 2005;31:26570. 13. Claffey E, Reader A, Nusstein J, et al. Anesthetic efcacy of articaine for inferior alveolar nerve blocks in patients with irreversible pulpitis. J Endod 2004;30: 56871. 14. Reitz J, Reader A, Nist R, et al. Anesthetic efcacy of the intraosseous injection of 0.9mL of 2% lidocaine (1:100,000 epinephrine) to augment an inferior alveolar nerve block. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;86: 51623. 15. Reader A, Nusstein, Hargreaves KM. Local anesthesia in endodontics. In: Cohen S, Hargreaves KM, eds. Pathways of the Pulp. 9th ed. St Louis, MO: Mosby Elsevier; 2006:691723. 16. Kanaa MD, Whitworth JM, Corbett IP, et al. Articaine buccal inltration enhances the effectiveness of lidocaine inferior alveolar nerve block. Int Endod J 2009;42: 23846. 17. Kanaa MD, Whitworth JM, Corbett IP, et al. Articaine and lidocaine mandibular buccal inltration anesthesia: a prospective randomized double-blind cross-over study. J Endod 2006;32:2968. 18. Foster W, Drum M, Reader A, et al. Anesthetic efcacy of buccal and lingual inltrations of lidocaine following an inferior alveolar nerve block in mandibular posterior teeth. Anesth Prog 2007;54:1639. 19. Jung IY, Kim JH, Kim ES, et al. An evaluation of buccal inltrations and inferior alveolar nerve blocks in pulpal anesthesia for mandibular rst molars. J Endod 2008; 34:113. 20. Meechan JG, Kanaa MD, Corbett IP, et al. Pulpal anaesthesia for mandibular permanent rst molar teeth: a double-blind randomized cross-over trial comparing buccal and buccal plus lingual inltration injections in volunteers. Int Endod J 2006;39: 7649. 21. Matthews R, Drum M, Reader A, et al. Articaine for supplemental buccal mandibular inltration anesthesia in patients with irreversible pulpitis when the inferior alveolar nerve block fails. J Endod 2009;35:3436. 22. Rosenberg PA, Amin KG, Zibari Y, et al. Comparison of 4% articaine with 1:100,000 epinephrine and 2% lidocaine with 1:100,000 epinephrine when used as a supplemental anesthetic. J Endod 2007;33:4035. 23. Modaresi J, Dianat O, Mozayeni MA. The efcacy comparison of ibuprofen, acetaminophen-codeine, and placebo premedication therapy on the depth of anesthesia during treatment of inamed teeth. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:399403. 24. Aggarwal V, Singla M, Kabi K. Comparative evaluation of effect of preoperative oral medication of ibuprofen and ketorolac on anesthetic efcacy of inferior alveolar nerve block with lidocaine in patients with irreversible pulpitis: a prospective, double-blind, randomized clinical trial. J Endod 2010;36:3758. 25. Oleson M, Drum M, Reader A, et al. Effect of preoperative ibuprofen on the success of the inferior alveolar nerve block in patients with irreversible pulpitis. J Endod 2010;36:37982. 26. Malamed SF. Handbook of Local Anesthesia. 5th ed. St Louis, MO: Mosby; 2004. 22754. 27. McCartney M, Reader A, Beck M. Injection pain of the inferior alveolar nerve block in patients with irreversible pulpitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;104:5715. 28. Parirokh M, Satvati A, Shari R, et al. Efcacy of combining a buccal inltration with an inferior alveolar nerve block for mandibular molars with irreversible pulpitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:46873. 29. Nakanishi T, Matsuo T, Ebisu S. Quantitative analysis of immunoglobulins and inammatory factors in human pulpal blood from exposed pulps. J Endod 1995; 21:1316. 30. Hargreaves KM, Karl Keiser, Byrne BE. Analgesics in endodontics. In: Cohen S, Hargreaves KM, eds. Pathways of the Pulp. 9th ed. St Louis, MO: Mosby; 2006:668. 31. Henry MA, Hargreaves KM. Peripheral mechanisms of odontogenic pain. Dent Clin North Am 2007;51:1944. 32. Wassef M, Pelage JP, Velzenberger E, et al. Anti-inammatory effect of ibuprofenloaded embolization beads in sheep uterus. J Biomed Mater Res B Appl Biomater 2008;86:6373. 33. Gould HJ 3rd, England JD, Soignier RD, et al. Ibuprofen blocks changes in Nav 1.7 and 1.8 sodium channels associated with complete Freunds adjuvant-induced inammation in rat. Pain 2004;5:27080. 34. Gibbs JL, Hargreaves KM. Mechanisms of odontogenic and non-odontogenic pain. In: Ingle JI, Backland LK, Baumgartner JG, eds. Endodontics. 6th ed. Hamilton, Ontario, Canada: BC Decker Inc; 2008:37691. 35. Seymour R, Ward P. Evaluation of different doses of soluble ibuprofen and ibuprofen tablets in postoperative dental pain. Br J Oral Maxillofac Surg 1996;34:1104. 36. Indomethacin web page. National Center for Biotechnology Information website. Available at: http://www.ncbi.nlm.nih.gov/bookshelf/br.fcgi?book=meds&log$= drug_bottom_one&part=a681027. Accessed February 15, 2010. 37. Shrestha M, Morgan D, Moreden J, et al. Randomized double-blind comparison of the analgesic efcacy of intramuscular ketorolac and oral indomethacin in the treatment of acute gouty arthritis. Ann Emerg Med 1995;26:6826. 38. Rogers MJ, Johnson BR, Remeikis NA, et al. Comparison of effect of intracanal use of ketorolac trimethamine and dexamethasone with oral ibuprofen on post treatment endodontic pain. J Endod 1999;25:3814. 39. Battrum D, Gutmann J. Efcacy of ketorolac in the management of pain associated with root canal treatment. J Can Dent Assoc 1996;62:3642. 40. Modaresi J, Dianat O, Soluti A. Effect of pulp inammation on nerve impulse quality with or without anesthesia. J Endod 2008;34:43841.

1454

Parirokh et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Diagnosis of Vertical Root Fractures in Endodontically Treated Teeth Based on Clinical and Radiographic Indices: A Systematic Review
Igor Tsesis, DMD,* Eyal Rosen, DMD,* Aviad Tamse, DMD,* Silvio Taschieri, MD, DDS, and Anda Kr, DMD*
Abstract
Introduction: The diagnosis of vertical root fracture (VRF) is at times complicated for lack of specic signs, symptoms, and/or radiographic features. The purpose of this study was to systematically search and evaluate the literature regarding the diagnostic accuracy of clinical signs and symptoms and radiographic indices for the diagnosis of VRF in endodontically treated teeth by means of a systematic review. Methods: An exhaustive literature search combined with strict inclusion and exclusion criteria was undertaken to identify clinical studies that assessed the diagnosis of VRF. Results: There is no substantial evidence regarding the accuracy of the clinical and radiographic indices for the diagnosis of VRF in endodontically treated teeth. Conclusions: Evidence-based data concerning the diagnostic accuracy and clinical effectiveness of clinical and radiographic dental evaluation for the diagnosis of VRF in endodontically treated teeth are lacking. The need for evidence-based research efforts to elucidate the currently unknown situation is of utmost signicance. (J Endod 2010;36:14551458)

Key Words
Diagnosis, systematic review, vertical root fracture

ertical root fracture (VRF) in endodontically treated teeth is one of the most frustrating complications of root canal therapy, which results in the tooth or root extraction (14). The VRF is a longitudinally oriented fracture of the root that originates from its apical end and propagates coronally (5) and is dened as one of the crack types (5). The incidence of vertically fractured teeth has not been described in the literature. However, the prevalence was reported to range from 11%20% of VRFs in extracted endodontically treated teeth (6, 7). VRF is usually diagnosed years after all endodontic and prosthetic procedures have been completed (8). The nal diagnosis of VRF is at times complicated for lack of specic signs, symptoms, and/or radiographic features and because several etiologic factors might be involved. Thus, the differential diagnosis from other pathologic entities might be difcult (3, 814). Evidence-based dentistry is an approach to oral healthcare that integrates the best available clinical evidence to support a practitioners clinical expertise for each patients treatment needs and preferences (1517). It is based on the process of systematically nding, apprising, and using research ndings as the basis for clinical decision-making. Systematic reviews constitute the basis for practicing evidence-based dentistry (15, 17, 18). The application of evidence-based dentistry in diagnosis should result in a reduction of errors in the clinical decision-making process (1518). Thus, evidence-based review of the available literature regarding the clinical and radiographic features of endodontically treated vertically root fractured teeth is of utmost importance (2, 4). The aim of this study was to systematically search and evaluate the literature regarding the diagnostic accuracy of clinical signs and symptoms and radiographic indices for the diagnosis of VRF in endodontically treated teeth by means of a systematic review.

Materials and Methods

From the *Department of Endodontology, Maurice and Gabriela Goldschleger School of Dental Medicine, Tel Aviv University, Tel Aviv, Israel; and IRCCS Istituto Ortopedico Galeazzi, Department of Odontology, University of Milan, Milan, Italy. Address requests for reprints to Dr Igor Tsesis, Department of Endodontology, Maurice and Gabriela Goldschleger School of Dental Medicine, Tel Aviv University, Tel Aviv, Israel. E-mail address: dr.tsesis@gmail.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.003

Criteria for Considering Studies for This Review This systematic review included clinical studies that assessed the diagnosis of VRF consistent with the following denition: complete or incomplete fracture initiated from the root at any level (5). Case reports, reviews, in vitro studies, and studies of teeth with vital pulps were not included in the review. The studies had to include patients with conrmed VRFs (target condition) in endodontically treated teeth. Studies with at least one of the following reference standards for VRF conrmation were included (reference standard dened as the best available method for establishing the presence or absence of the target condition (19)). The VRFs were conrmed during surgical ap procedure, the VRFs were conrmed after tooth extraction, or the VRFs were identied radiographically as clearly discernable separation of segments of fractured roots. The studies had to include detailed data regarding the clinical signs and symptoms and the radiographic ndings of the evaluated teeth. Table 1 summarizes the criteria for inclusion of studies in the systematic review. Search Methods for Identication of Studies The following electronic databases were searched. MEDLINE database was searched by using Evidents web-based search engine (http://medinformatics.uthscsa.

JOE Volume 36, Number 9, September 2010

Diagnosis of VRFs in Endodontically Treated Teeth

1455

Clinical Research
TABLE 1. Criteria for Inclusion of Studies in the Systematic Review
1. Clinical studies that assessed the diagnosis of VRF. 2. VRF denition corresponded with the description of complete or incomplete fracture initiated from the root at any level. 3. Only endodontically treated teeth were included. 4. VRF conrmation method was in accordance with the dened reference standards. 5. Detailed data were included regarding the clinical signs and symptoms and radiographic ndings of the evaluated teeth.
VRF, vertical root fractures.

edu/EviDents/), by using the key words vertical root fracture OR longitudinal root fracture, and applying Entrez PubMed limits to humans and English. MeSH received the following: (vertical[All Fields] AND (plant roots[MeSH Terms] OR (plant[All Fields] AND roots[All Fields]) OR plant roots[All Fields] OR root[All Fields]) AND (fractures, bone[MeSH Terms] OR (fractures[All Fields] AND bone[All Fields]) OR bone fractures[All Fields] OR fracture[All Fields])) OR (longitudinal[All Fields] AND (plant roots[MeSH Terms] OR (plant[All Fields] AND roots[All Fields]) OR plant roots[All Fields] OR root[All Fields]) AND (fractures, bone[MeSH Terms] OR (fractures[All Fields] AND bone[All Fields]) OR bone fractures[All Fields] OR fracture[All Fields])) NOT (animals[MeSH:noexp] NOT humans[MeSH Terms]) AND (humans[MeSH Terms] AND English[lang]). Scopus search (www. scopus.com) was searched by using the key words vertical root fracture OR longitudinal root fracture, by using these lters: articles, limited to dentistry, English language. Query received the following: TITLE-ABS-KEY(vertical root fracture OR longitudinal root fracture) AND (LIMIT-TO(DOCTYPE, ar)) AND (LIMIT-TO(SUBJAREA, DENT) OR LIMIT-TO(SUBJAREA, MULT)) AND (LIMIT-TO(LANGUAGE, English)). Embase database (http://www.embase.com) was searched by using the key words vertical root fracture OR longitudinal root fracture, with these limits: humans, English, article, Embase only. MeSH received the following: (vertical root fracture OR longitudinal root fracture) AND [article]/lim AND [humans]/lim AND [english]/ lim AND [embase]/lim. Related articles, literature reviews that appeared in the MEDLINE search engine, and textbook chapters were evaluated, and their reference lists were manually checked.

The following index tests, dened as tests performed to reduce the uncertainty about the presence of the target condition (19, 20), were to be recorded: tenderness to percussion and/or palpation, presence of a sinus tract, pain, swelling, presence of periradicular radiolucency, presence of an osseous defect/periodontal pocket, gender and age of the patient, tooth type, coronal restoration type, and presence of post. Methodological Quality Assessment and Data Synthesis and Analysis. On the basis of the study methodological parameters, an assessment of the methodological quality of the included studies was planned to be undertaken following the recommendations of the guest editorial on evidence-based dentistry published by the Journal of Endodontics in 2009 (15) and the recommendations of The Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy (20). The diagnostic accuracy was planned to be calculated by comparing the results of the index tests with the outcomes of the reference standards (16, 19, 21).

Results
The search in MEDLINE database by using Evidents web-based search engine covered all articles published in dental journals in English from 1971January 2010 and resulted in 197 articles, of which 19 were eligible for inclusion (3, 10, 14, 2237) on the basis of their titles and abstracts. The other 178 articles were rejected; 62 were found to be irrelevant to the topic of the present study, 14 were review articles, 52 were case reports, and 50 were ex vivo studies. The search with Scopus database resulted in 117 articles, of which 5 articles (1, 6, 8, 38, 39) were not identied in the MEDLINE database search and were found to be eligible for inclusion on the basis of their titles and abstracts. The search with Embase database yielded no additional articles. The manual search in the related articles and the reference lists of literature reviews and textbook chapters resulted in 3 additional articles that were not previously identied in the electronic search and that were eligible for inclusion on the basis of their titles and abstracts (4042).

Data Collection and Analysis Selection of Studies. The articles were initially evaluated for relevance, on the basis of their titles and abstracts, by 3 observers independently (I.T., A.T., E.R.). Possibly relevant studies were submitted to a full text evaluation. The full texts of the studies were obtained and reviewed for suitability. Cases of disagreement were discussed together until agreement was achieved. Eventually, the identied suitable articles were subjected to data extraction, assessment of the methodological quality, and data analysis. Data Extraction. Data were to be extracted by 3 observers independently. Cases of disagreement were to be subject to joint evaluation by the observers until an agreement was achieved. For each study the following methodological parameters were to be recorded: authors and date of publication; studys purpose; sample size; demographic details of the subjects including criteria for inclusion; study design; reference standard type; evaluation methods including the presence of multiple evaluators, calibration of evaluators, randomization, and evaluators blinding; homogeneity of the subjects; statistical analysis type and quality; and funding source.
1456
Tsesis et al.

Full Text Evaluation The combined search through the electronic databases and the manual search resulted in a total of 27 studies (1, 3, 6, 8, 10, 14, 2242), which were subjected to a suitability test in accordance with the criteria for considering studies for this review (Table 1). Six articles were found to be case reports, reviews, in vitro studies, or were found irrelevant to the topic of the present study. One article was not compatible with the VRF denition. In 5 articles non-endodontically treated teeth were included (and data could not be separated from endodontically treated teeth). In 7 articles the VRF was not conrmed according to the reference standards, and 14 articles provided no detailed data on clinical signs and symptoms and radiographic ndings of the evaluated teeth. Eventually, none of the articles submitted to full text evaluation fullled the inclusion criteria (Table 2). As a result, data extraction, methodological quality assessment, and data analysis could not be carried out. Thus, it was not possible to nd substantial evidence regarding the diagnostic accuracy of the index tests for the diagnosis of VRF in endodontically treated teeth.

Discussion
Numerous studies dealing with the diagnosis of VRF and the clinical and radiographic characteristics of the vertically root fractured teeth have been published (2, 3, 711, 14, 28, 29, 31, 34, 35, 39, 4244). However, their study design, VRF denition, VRF conrmation methods, and evaluated clinical and radiographic data have been extremely variable, which resulted in publication of

JOE Volume 36, Number 9, September 2010

Clinical Research
TABLE 2. Excluded Studies and the Reasons for Exclusion Study
Takeuchi et al, 2009 (33) Zadik et al, 2008 (37) Miyamoto et al, 2007 (30) Cohen et al, 2006 (25) Chambrone et al, 2006 (22) Tamse et al, 2006 (34) Cornelini R et al 2005 (26) Fox et al, 2004 (27) Llena-Puy et al, 2001 (28) Lustig et al, 2000 (29) Tamse et al, 1999 (3) Tamse et al, 1999 (14) Chan et al, 1999 (24) Youssefzadeh et al, 1999 (36) Chan et al, 1998 (23) Nicopoulou-Karayianni et al, 1997 (32) Testori et al, 1993 (35) Mors et al, 1990 (31) Meister et al, 1980 (10) Kawamura-Hagiya et al, 2008 (38) Fuss et al, 2001 (8) Fuss et al, 1999 (6) Walton et al, 1984 (1) Pitts et al, 1983 (39) Rud et al 1970, (40) Abou-Rass M et al, 1983 (41) Gher et al, 1987 (42)

Reason for exclusion*

3, 4 4 1 3 1 5 1 1 4 5 5 5 3, 4, 5 5 1 3, 4 4 4 5 5 5 5 5 1 2, 5 3, 5 5

*Based on incompatibleness to the inclusion criteria of the systematic review (Table 1): (1) case reports, reviews, in vitro studies, and studies irrelevant to the topic of the current study; (2) incompatible with the VRF denition; (3) non-endodontically treated teeth are included; (4) lack of VRF conrmation according to the reference standards; (5) no detailed data regarding the clinical signs and symptoms and radiographic ndings of the evaluated teeth.

inconsistent and confusing results (2, 4, 5, 9, 10, 25, 40). The aim of the present study was to systematically search and evaluate the available literature concerning the diagnostic accuracy of clinical and radiographic indices for the diagnosis of VRF in endodontically treated teeth by the means of a systematic review. Systematic reviews use a systematic approach and explicit methodology to review and synthesize research evidence, aimed to minimize bias, and explicitly address the issues of the completeness of the identied evidence, assess the quality of the included studies and the studies combinability (1518, 21, 45, 46). This systematic process requires a comprehensive literature search to identify as much of the relevant literature as possible (17, 18, 21, 47). It is recommended that a combined search of several electronic databases, together with other methods to retrieve studies (such as hand-searching of reference lists of relevant articles and book chapters), will be conducted (1618, 21, 45, 47). In the present study a combined comprehensive literature search of several electronic databases and hand-search of related articles, literature reviews, and textbook chapters was conducted. On the basis of the study titles and abstracts, 27 possibly relevant articles were identied. To overcome heterogeneity of information, strict inclusion and exclusion criteria were applied to the identied studies. Furthermore, it is crucial that the studies that are compatible with the inclusion criteria of the systematic review will be also subjected to methodological quality assessment to determine the strength of evidence (15). In the present study no article was in compliance with the criteria for entering the systematic review; thus methodological quality assessment was not carried out. The ability to assess the diagnostic accuracy and clinical usefulness of clinical and radiographic parameters for the diagnosis of a target

condition depends on the availability of a valid reference standard. The appropriate reference standard should be carefully chosen and has to be based on a technique different from that of the tests being evaluated (16). The right selection of a reference standard is of paramount signicance. In the present review, 3 types of reference standards were selected as acceptable for conrmation of the VRF (the target condition): conrmation during surgical ap procedure, conrmation after tooth extraction, and radiographic identication being a clearly discernable separation of segments of fractured root. In the current systematic review only endodontically treated teeth were included, because it is expected that in vital teeth the signs and symptoms would be different and reect pulp inammation without periradicular changes (5, 24, 48, 49). While considering the usefulness of a diagnostic evaluation, the relative value of possible health states resulting from the diagnosis and subsequent therapy should be taken into consideration (16). When VRF diagnosis is made, a quick decision to extract the tooth or root is necessary. The reason is that the inammation in the supporting tissues would otherwise lead to periodontal breakdown followed by the development of a deep osseous defect (1) and resorption of the bone facing the root fracture. Immediate treatment might prevent this unnecessary bone loss that might lead to complicated restoration of the area of extraction, if an implant should be considered the treatment of choice (2). A diagnostic process is based on the combination of the patients subjective complaints and objective clinical and radiographic evaluation (50). The clinician is supposed to correlate the subjective and objective ndings and formulate a diagnosis (50). In the case of VRF diagnosis, there is no known single pathognomonic sign, symptom, or radiographic feature to make the diagnosis easy and denitive (29). In 2006 in their article on demographic analysis of VRFs, Cohen et al (25) stated that one of the main reasons for the difcult diagnosis of VRF is the fact that the nal diagnosis is based on constellation of several signs and symptoms rather than a single pathognomonic one. Therefore, the present review mandated the inclusion of detailed data regarding both clinical and radiographic evaluations (index tests) of the diagnosed teeth. The most common signs and symptoms of VRF described in the literature are deep osseous defects especially on the buccal aspect of the susceptible teeth and roots, and highly located sinus tract (3, 10, 35). The deep osseous defect (deep probing), especially on the buccal aspect of the more susceptible teeth and roots (maxillary and mandibular premolars and mesial roots of the mandibular molars), were found in retrospective case series publications in high percentages and with statistical signicance: Meister et al (10), 93%; Tamse (13), 64%; Testori et al (35), 78%; and Tamse et al (3), 67%. The sinus tract in the gingiva, especially when it is coronally located closer to the gingival margin (as opposed to the sinus tract in failing endodontic cases), is also a typical sign for the diagnosis of VRF. This type of sinus tract was found by Tamse et al (3) in 35% and by Testori et al (35) in 42% of the VRF teeth. The most frequent radiographic features of VRF are the halo appearance, which is a combined periapical and perilateral radiolucency in one or both sides of the root, lateral periodontal radiolucency along the side of the root, or angular radiolucency from the crestal bone terminating along the root side (14, 32, 35). In mandibular molars, radiolucency in the furcation area can often be observed, coupled with the types of radiolucencies described above (34). In the present systematic review no articles passed the inclusion criteria. That means that unfortunately, evidence-based data concerning the diagnostic accuracy and clinical effectiveness of the daily used clinical and radiographic dental evaluation for the diagnosis of VRF in 1457

JOE Volume 36, Number 9, September 2010

Diagnosis of VRFs in Endodontically Treated Teeth

Clinical Research
endodontically treated teeth are lacking. Thus, the need for a cautious rigorous clinical approach for the diagnosis and follow-up of suspected VRFs and the need for evidence-based research efforts to elucidate the currently unknown are of utmost signicance.
21. Whiting P, Westwood M, Burke M, Sterne J, Glanville J. Systematic reviews of test accuracy should search a range of databases to identify primary studies. J Clin Epidemiol 2008;61:35764. 22. Chambrone LA, Chambrone L. Tooth loss in well-maintained patients with chronic periodontitis during long-term supportive therapy in Brazil. J Clin Periodontol 2006; 33:75964. 23. Chan CP, Jeng JH, Chang SH, Chen CC, Lin CJ, Lin CP. Cutaneous sinus tracts of dental origin: clinical review of 37 cases. J Formos Med Assoc 1998;97:6337. 24. Chan CP, Lin CP, Tseng SC, Jeng JH. Vertical root fracture in endodontically versus nonendodontically treated teeth: a survey of 315 cases in Chinese patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999;87:5047. 25. Cohen S, Berman LH, Blanco L, Bakland L, Kim JS. A demographic analysis of vertical root fractures. J Endod 2006;32:11603. 26. Cornelini R, Cangini F, Covani U, Wilson TG Jr. Immediate restoration of implants placed into fresh extraction sockets for single-tooth replacement: a prospective clinical study. Int J Periodontics Restorative Dent 2005;25:43947. 27. Fox K, Wood DJ, Youngson CC. A clinical report of 85 fractured metallic postretained crowns. Int Endod J 2004;37:56173. 28. Llena-Puy MC, Forner-Navarro L, Barbero-Navarro I. Vertical root fracture in endodontically treated teeth: a review of 25 cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;92:5535. 29. Lustig JP, Tamse A, Fuss Z. Pattern of bone resorption in vertically fractured, endodontically treated teeth. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;90:2247. 30. Miyamoto T, Morgano SM, Kumagai T, Jones JA, Nunn ME. Treatment history of teeth in relation to the longevity of the teeth and their restorations: outcomes of teeth treated and maintained for 15 years. J Prosthet Dent 2007;97:1506. 31. Mors AS. Vertical root fractures. Oral Surg Oral Med Oral Pathol 1990;69: 6315. 32. Nicopoulou-Karayianni K, Bragger U, Lang NP. Patterns of periodontal destruction associated with incomplete root fractures. Dentomaxillofac Radiol 1997;26:3216. 33. Takeuchi N, Yamamoto T, Tomofuji T, Murakami C. A retrospective study on the prognosis of teeth with root fracture in patients during the maintenance phase of periodontal therapy. Dent Traumatol 2009;25:3327. 34. Tamse A, Kaffe I, Lustig J, Ganor Y, Fuss Z. Radiographic features of vertically fractured endodontically treated mesial roots of mandibular molars. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;101:797802. 35. Testori T, Badino M, Castagnola M. Vertical root fractures in endodontically treated teeth: a clinical survey of 36 cases. J Endod 1993;19:8791. 36. Youssefzadeh S, Gahleitner A, Dorffner R, Bernhart T, Kainberger FM. Dental vertical root fractures: value of CT in detection. Radiology 1999;210:5459. 37. Zadik Y, Sandler V, Bechor R, Salehrabi R. Analysis of factors related to extraction of endodontically treated teeth. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:e315. 38. Kawamura-Hagiya Y, Yoshioka T, Suda H. Logistic regression equation to screen for vertical root fractures using periapical radiographs. Dentomaxillofac Radiol 2008; 37:2833. 39. Pitts DL, Natkin E. Diagnosis and treatment of vertical root fractures. J Endod 1983; 9:33846. 40. Rud J, Omnell KA. Root fractures due to corrosion: diagnostic aspects. Scand J Dent Res 1970;78:397403. 41. Abou-Rass M. Crack lines: the precursors of tooth fracturestheir diagnosis and treatment. Quintessence Int Dent Dig 1983;14:43747. 42. Gher ME Jr, Dunlap RM, Anderson MH, Kuhl LV. Clinical survey of fractured teeth. J Am Dent Assoc 1987;114:1747. 43. Dang DA, Walton RE. Vertical root fracture and root distortion: effect of spreader design. J Endod 1989;15:294301. 44. Kishen A. Mechanisms and risk factors for fracture predilection in endodontically treated teeth. Endodontic Topics 2006;13:5783. 45. Greenhalgh T, Peacock R. Effectiveness and efciency of search methods in systematic reviews of complex evidence: audit of primary sources. BMJ 2005;331:10645. 46. Suebnukarn S, Ngamboonsirisingh S, Rattanabanlang A. A systematic evaluation of the quality of meta-analyses in endodontics. J Endod 2010;36:6028. 47. Torabinejad M, Corr R, Handysides R, Shabahang S. Outcomes of nonsurgical retreatment and endodontic surgery: a systematic review. J Endod 2009;35:9307. 48. Cameron CE. Cracked-tooth syndrome. J Am Dent Assoc 1964;68:40511. 49. Cameron CE. The cracked tooth syndrome: additional ndings. J Am Dent Assoc 1976;93:9715. 50. Berman LH, Hartwell GR. Diagnosis. In: Cohen S, Hargreaves KM, eds. Pathways of the pulp. 9th ed. St Louis, MO: Mosby; 2006:239.

Conclusions
The evidence-based data regarding the diagnostic accuracy and usefulness of the commonly used clinical and radiographic evaluation methods for the diagnosis of VRF in endodontically treated teeth are lacking. That makes the determination of a fractured root often not distinctly objective and more of a prediction rather than a denitive diagnosis, as previously stated (25). In case a VRF is suspected, the clinician should be aware of the current conict between the signicant clinical importance of a quick and correct diagnosis and the lack of evidence-based data supporting the usefulness of common clinical and radiographic evaluation methods. Future evidence-based research regarding the diagnosis of VRF is needed and might possibly shed light on this clinical challenge.

References
1. Walton RE, Michelich RJ, Smith GN. The histopathogenesis of vertical root fractures. J Endod 1984;10:4856. 2. Tamse A. Vertical root fractures in endodontically treated teeth: diagnostic signs and clinical management. Endodontic Topics 2006;13:8494. 3. Tamse A, Fuss Z, Lustig J, Kaplavi J. An evaluation of endodontically treated vertically fractured teeth. J Endod 1999;25:5068. 4. Tamse A. Vertical root fractures of endodontically treated teeth. In: Ingle JI, Bakland LK, Baumgartner JC, eds. Ingles endodontics. 6th ed. Hamilton, Ontario, Canada: BC Decker Inc; 2008:67689. 5. Colleagues for excellence. Cracking the cracked tooth code: detection and treatment of various longitudinal tooth fractures. Chicago: American Association of Endodontics; 2008. 6. Fuss Z, Lustig J, Tamse A. Prevalence of vertical root fractures in extracted endodontically treated teeth. Int Endod J 1999;32:2836. 7. Coppens CRM, DeMoor RJG. Prevalence of vertical root fractures in extracted endodontically treated teeth. Int Endod J 2003;36:926. 8. Fuss Z, Lustig J, Katz A, Tamse A. An evaluation of endodontically treated vertical root fractured teeth: impact of operative procedures. J Endod 2001;27:468. 9. Cohen S, Blanco L, Berman L. Vertical root fractures: clinical and radiographic diagnosis. J Am Dent Assoc 2003;134:43441. 10. Meister F Jr, Lommel TJ, Gerstein H. Diagnosis and possible causes of vertical root fractures. Oral Surg Oral Med Oral Pathol 1980;49:24353. 11. Moule AJ, Kahler B. Diagnosis and management of teeth with vertical root fractures. Aust Dent J 1999;44:7587. 12. Sedgley CM, Messer HH. Are endodontically treated teeth more brittle? J Endod 1992;18:3325. 13. Tamse A. Iatrogenic vertical root fractures in endodontically treated teeth. Endod Dent Traumatol 1988;4:1906. 14. Tamse A, Fuss Z, Lustig J, Ganor Y, Kaffe I. Radiographic features of vertically fractured, endodontically treated maxillary premolars. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999;88:34852. 15. Gutmann JL. Evidence-based/guest editorial. J Endod 2009;35:1093. 16. Mileman PA, van den Hout WB. Evidence-based diagnosis and clinical decision making. Dentomaxillofac Radiol 2009;38:110. 17. Rosenberg W, Donald A. Evidence based medicine: an approach to clinical problem-solving. BMJ 1995;310:11226. 18. Sutherland SE, Matthews DC. Conducting systematic reviews and creating clinical practice guidelines in dentistry: lessons learned. J Am Dent Assoc 2004;135:74753. 19. Bossuyt PM, Leeang MM. Chapter 6: developing criteria for including studies. Cochrane handbook for systematic reviews of diagnostic test accuracy: the Cochrane Collaboration; 2008. 20. Reitsma JB, Rutjes AWS, Whiting P, Vlassov VV, Leeang MMG, Deeks JJ. Chapter 9: assessing methodological quality. In: Deeks JJ, Bossuyt PM, Gatsonis C, eds. Cochrane handbook for systematic reviews of diagnostic test accuracy: the Cochrane Collaboration; 2009.

1458

Tsesis et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Prospective, Randomized Single-blind Study of the Anesthetic Efcacy of 1.8 and 3.6 Milliliters of 2% Lidocaine with 1:50,000 Epinephrine for Inferior Alveolar Nerve Block
Maji Wali, BDS, MS, Melissa Drum, DDS, MS, Al Reader, DDS, MS, and John Nusstein, DDS, MS
Abstract
Introduction: The purpose of this prospective, randomized study was to compare the degree of pulpal anesthesia obtained in vital, asymptomatic teeth by using 1.8 mL and 3.6 mL of 2% lidocaine with 1:50,000 epinephrine compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine for inferior alveolar nerve (IAN) block. Methods: Thirty adult subjects randomly received IAN blocks of 1.8 mL and 3.6 mL of 2% lidocaine with 1:50,000 epinephrine and 1.8 mL of 2% lidocaine with 1:100,000 epinephrine at 3 separate appointments in a crossover design. An electric pulp tester was used to test for anesthesia in 3-minute cycles for 60 minutes of the rst molars, rst premolars, and lateral incisors. Anesthesia was considered successful when 2 consecutive 80 readings were obtained within 15 minutes, and the 80 reading was continuously sustained through the 60th minute. Results: By using 1.8 mL of 2% lidocaine with 1:50,000 epinephrine, successful pulpal anesthesia ranged from 33%50%, and when using 3.6 mL of 2% lidocaine with 1:50,000 epinephrine, success ranged from 40%60%. When using 1.8 mL of 2% lidocaine with 1:100,000 epinephrine, success ranged from 40%60%, with no signicant difference among the 3 anesthetic formulations. Conclusion: Increasing the epinephrine concentration to 1:50,000 epinephrine or increasing the volume to 3.6 mL of 2% lidocaine with 1:50,000 epinephrine did not result in more successful pulpal anesthesia when compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine by using the IAN block. (J Endod 2010;36:14591462)

he inferior alveolar nerve (IAN) block does not always result in successful pulpal anesthesia (16). Failure rates of 10%39% have been reported in experimental studies (1). Therefore, it would be advantageous to improve the success rate of the IAN block. The effect of epinephrine concentration on the success of an IAN block has been studied. Bou Dagher et al (7) evaluated 1.8 mL of 2% lidocaine with 3 concentrations of epinephrine, 1:50,000, 1:80,000, and 1:100,000, in IAN blocks. They found no significant differences between the 3 formulations for anesthetic success or failure by using a pulp tester to evaluate pulpal anesthesia. Likewise, Yared and Bou Dagher (8) evaluated 3.6 mL of 2% lidocaine with 3 concentrations of epinephrine, 1:50,000, 1:80,000, and 1:100,000, in IAN blocks by using the same methodology as Bou Dagher et al. They also found no signicant differences between the 3 formulations for anesthetic success or failure by using a pulp tester to evaluate pulpal anesthesia. However, Yared and Bou Dagher retrospectively compared the results of the 3.6 mL volume with the 1.8 mL volume by using the data from Bou Dagher et al and found statistically higher success rates with the 3.6 mL volume. The efcacy of 1:50,000 epinephrine in a 2% lidocaine formulation and an increased volume of a 2% lidocaine formulation with 1:50,000 epinephrine in providing pulpal anesthesia when administered in an IAN block needs further investigation to ensure its appropriate clinical use. The purpose of this prospective, randomized single-blind study was to compare the degree of pulpal anesthesia obtained in vital, asymptomatic teeth by using 1.8 mL and 3.6 mL of 2% lidocaine with 1:50,000 epinephrine compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine for the IAN block.

Materials and Methods

Thirty adult subjects participated in this study. The subjects were in good health and were not taking any medications that would alter their perception of pain. Exclusion criteria were as follows: younger than 18 years; older than 65 years; allergies to local anesthetics or sultes; pregnancy; history of signicant medical conditions; taking any medications that might affect anesthetic assessment (nonsteroidal anti-inammatory drugs, opioids, antidepressants, alcohol); active sites of pathosis in area of injection; and inability to give informed consent. The Ohio State University Human Subjects Review Committee approved both the protocol and informed consent document, and written informed consent was obtained from each subject. The subjects randomly received each of 3 solutions in an IAN block at 3 separate appointments, spaced at least 1 week apart, in a crossover design. The 3 solutions were 1.8 mL of 2% lidocaine with 1:100,000 epinephrine, 1.8 mL of 2% lidocaine with 1:50,000 epinephrine, and 3.6 mL of 2% lidocaine with 1:50,000 epinephrine. With the crossover design, there were 90 total injections administered, and each subject served as his or her own control. Fifteen sets of 3 IAN block injections were administered on the right side, and 15 sets of 3 IAN block injections were administered on the left side. The same side randomly chosen for the rst injection was used again for the second and third injections. The test teeth chosen for the experiment were the mandibular rst molar, rst premolar, and lateral incisor. The contralateral mandibular canine was used as the unanesthetized control to ensure that the electric pulp tester was operating properly and that the subject was responding appropriately during each

Key Words
Epinephrine concentrations, inferior alveolar nerve block, lidocaine, local anesthesia

From the Division of Endodontics, The Ohio State University, Columbus, Ohio. Maji Wali is currently in Baghdad, Iraq. Address requests for reprints to Dr Melissa Drum, Division of Endodontics, College of Dentistry, The Ohio State University, 305 West 12th Ave, Columbus, OH 43210. E-mail address: drum.13@osu.edu. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.012

JOE Volume 36, Number 9, September 2010

Anesthetic Efcacy of Lidocaine with Epinephrine for IAN Block

1459

Clinical Research
TABLE 1. Percentages and Numbers of Subjects who Experienced Anesthetic Success Anesthetic formulation 1.8 mL of 2% lidocaine with 1:50,000 epinephrine
Anesthetic success* First molar First premolar Lateral incisor 33% (10/30) 50% (15/30) 37% (11/30)

3.6 mL of 2% lidocaine with 1:50,000 epinephrine

40% (12/30) 60% (18/30) 47% (14/30)

1.8 mL of 2% lidocaine with 1:100,000 epinephrine

43% (13/30) 60% (18/30) 40% (12/30)

n = 30. *There were no signicant differences among the 3 anesthetic formulations.

experimental portion of the study. Clinical examinations indicated that all teeth were free of caries, large restorations, and periodontal disease; none had histories of trauma or sensitivity. Before the experiment, the 3 solutions were randomly assigned 4digit numbers from a random number table. Each subject was randomly assigned to the right or left side for the set of injections. The order of the anesthetic solutions was also randomly assigned to determine which solutions were to be administered at each appointment. Only the random numbers were recorded on the data collection and postinjection survey sheets to help blind the experiment. Five-milliliter Luer-Lok sterile syringes (Becton, Dickinson and Co, Rutherford, NJ) were loaded with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine (Xylocaine; Astra Pharmaceutical Products, Inc, Worchester, MA), 1.8 mL of 2% lidocaine with 1:50,000 epinephrine (Xylocaine), or 3.6 mL of 2% lidocaine with 1:50,000 epinephrine (Xylocaine) under sterile conditions. All anesthetic cartridges were checked to ensure that the anesthetic solution had not expired. An opaque tape was placed on each syringe, and the corresponding 4-digit code number was written on the tape. At the beginning of each appointment and before any injections were given, the experimental teeth and control contralateral canine were tested 3 times with the pulp tester (Kerr, Analytic Technology Corp, Redmond, WA) to record baseline vitality. After the tooth to be tested was isolated with cotton rolls and dried with gauze, toothpaste was applied to the probe tip, which was then placed midway between the gingival margin and the occlusal or incisal edge of the tooth. The current rate was set at 25 seconds to increase from no output (0) to the maximum output (80). The number associated with the initial

sensation was recorded. Trained research personnel performed all preinjection and postinjection tests. All subjects received a conventional IAN block as described by Jorgensen and Hayden (9). The needle used for all injections was a 27gauge 112 -inch needle (Becton, Dickinson and Company, Franklin Lakes, NJ). Once the target site was reached, each anesthetic formulation was deposited during a period of 2 minutes. All injections were given by the senior author (M.W.). At 1 minute after each block was given, the rst molar was pulp tested. At 2 minutes, the rst premolar and lateral incisor were tested. At 3 minutes, the contralateral canine was pulp tested, and the subject was asked if his/her lip was numb. This cycle of testing was repeated every 3 minutes for 60 minutes. At every third cycle the control tooth, the contralateral canine, was tested by a pulp tester without batteries to test the reliability of the subject. If profound lip numbness was not recorded within 20 minutes, the block was considered unsuccessful, and the subject was reappointed. All testing was stopped at 60 minutes after injection. No response from the subject at the maximum output (80 reading) of the pulp tester was used as the criterion for pulpal anesthesia. Anesthesia was considered successful when 2 consecutive 80 readings were obtained within 15 minutes, and the 80 reading was continuously sustained through the 60th minute. For most restorative procedures, we would want the patient numb within 15 minutes and to remain numb through the 60th minute. Data were analyzed statistically. All subjects used for data analysis had profound lip anesthesia. Comparisons of the 3 anesthetic formulations for anesthetic success, onset of lip and pulpal anesthesia, were

100

100

Percentage of 80 Readings

75

Percentage of 80 Readings

75

50

50

25

1.8 1 8 mL 2% Lidoc aine with 1:50,000 e pine phrine 3.6 mL 2% Lidoc aine with 1:50,000 e pine phrine 1.8 mL 2% lidoc aine with 1:100,000 e pine phrine
1 7 13 19 25 31 37 43 49 55

25

1.8 mL 2% Lidoc aine with 1:50,000 e pine phrine h i 3.6 mL 2% Lidoc aine with 1:50,000 e pine phrine 1.8 mL 2% lidoc aine with 1:100,000 e pine phrine

0 1 7 13 19 25 31 37 43 49 55

Time (M inute s)

Time (M inute s)

Figure 1. Incidence of rst molar anesthesia as determined by lack of response to electrical pulp testing at the maximum setting (percentage of 80/80s), at each postinjection time interval, for the 3 solutions.

Figure 2. Incidence of rst premolar anesthesia as determined by lack of response to electrical pulp testing at the maximum setting (percentage of 80/80s), at each postinjection time interval, for the 3 solutions.

1460

Wali et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
formulations. For pulpal anesthesia, there were no signicant differences among the 3 anesthetic formulations.

Discussion
We based our use of the electric pulp test reading of 80, signaling maximum output, as a criterion for pulpal anesthesia on the studies of Dreven et al (10) and Certosimo and Archer (11). These studies showed that no patient response to an 80 reading ensured pulpal anesthesia in vital asymptomatic teeth. In addition, Certosimo and Archer demonstrated that electric pulp test readings less than 80 resulted in pain during operative procedures in asymptomatic teeth. Therefore, using the electric pulp tester before beginning dental procedures on asymptomatic vital teeth will provide the clinician a reliable indicator of pulpal anesthesia. Because all subjects felt profound lip numbness but pulp testing revealed that subjects did not always have pulpal anesthesia (80 readings), asking the patient if the lip is numb only indicates soft tissue anesthesia but does not guarantee successful pulpal anesthesia. Regarding epinephrine concentrations, Bou Dagher et al (7) recorded success rates for 1.8 mL of 2% lidocaine with 1:50,000 epinephrine of 50% for the rst molar, 53% for the rst premolar, and 47% for the lateral incisor. In the current study, success rates were lower for the rst molar (33%) and lateral incisor (37%) but similar in the rst premolar (50%). Differences between studies might be due to population differences. When 1.8 mL of 2% lidocaine with 1:50,000 epinephrine was compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine, no differences were found (Table 1). Bou Dagher found similar results. Regarding the increased volume, Yared and Bou Dagher (8) recorded success rates for 3.6 mL of 2% lidocaine with 1:50,000 epinephrine of 77% for the rst molar, 80% for the rst premolar, and 70% for the lateral incisor. In the current study, success rates were lower for the rst molar (40%), lateral incisor (60%), and rst premolar (47%). When Yared and Bou Dagher retrospectively compared the results of the 3.6 mL volume with the 1.8 mL volume by using different subject populations, they found statistically higher success rates with the 3.6 mL volume. In the current study, we found no signicant difference between the 1.8 mL and 3.6 mL volumes for 2% lidocaine with 1:50,000 epinephrine. Therefore, we could not conrm the results of Yared and Bou Dagher. Generally, 1.8 mL of 2% lidocaine with 1:100,000 epinephrine had similar rates of anesthetic success as other studies of the IAN block (1). Nusstein et al (1) found success rates of 53% for the rst molar, 61% for the rst premolar, and 35% for the lateral incisor. These are similar to the success rates of 43% for the rst molar, 60% for the rst premolar, and 40% for the lateral incisor in the current study (Table 1). The recorded success rates of Yared and Bou Dagher (8) for 3.6 mL of 2% lidocaine with 1:100,000 epinephrine were 77% for the rst molar,

Figure 3. Incidence of lateral incisor anesthesia as determined by lack of response to electrical pulp testing at the maximum setting (percentage of 80/80s), at each postinjection time interval, for the 3 solutions.

analyzed nonparametrically by using exact McNemar tests that were adjusted by using the step-down Bonferroni method of Holm. With a nondirectional alpha risk of 0.05 and assuming a success rate of 40% and a change proportion of 80%, a sample size of 30 subjects was required to demonstrate a difference in anesthetic success of 30 percentage points with a power of 0.80. Comparisons were considered signicant at P <.05.

Results
Thirty adult subjects, 8 women and 22 men, aged 2244 years with an average age of 28 years, participated. One hundred percent of the subjects used for data analysis had profound lip anesthesia. Anesthetic success is presented in Table 1. The success rate for the 1.8 mL of 2% lidocaine with 1:50,000 epinephrine formulation was 33% for the rst molar, 50% for the rst premolar, and 37% for the lateral incisor. For the 3.6 mL of 2% lidocaine with 1:50,000 epinephrine formulation, success rates were 40% for the rst molar, 60% for the rst premolar, and 47% for the lateral incisor. The 1.8 mL of 2% lidocaine with 1:100,000 epinephrine success rates were 43% for the rst molar, 60% for the rst premolar, and 40% for the lateral incisor. There were no signicant differences among the 3 anesthetic formulations. Figs. 1 through 3 show the incidence of pulpal anesthesia (80 readings) for the 3 anesthetic formulations. Onset times of lip and pulpal anesthesia are shown in Table 2. For onset of lip anesthesia, there was a signicant difference when 1.8 mL of 2% lidocaine with 1:50,000 epinephrine was compared with the other 2

TABLE 2. Mean Onset Times (minutes) of Subjective Lip Anesthesia and Pulpal Anesthesia Anesthetic formulation 1.8 mL of 2% lidocaine with 1:50,000 epinephrine
Onset of lip numbness (min)* Onset of pulpal anesthesia (min) First molar First premolar Lateral incisor 5.9 0.5 15.9 3.4 15.5 2.7 14.0 2.3

3.6 mL of 2% lidocaine with 1:50,000 epinephrine

4.7 0.5 11.1 2.7 10.5 2.7 11.9 3.0

1.8 mL of 2% lidocaine with 1:100,000 epinephrine

4.4 0.4 13.3 2.7 12.5 2.5 11.6 2.8

*There was a signicant difference when 1.8 mL of 2% lidocaine with 1:50,000 epinephrine was compared with the other 2 formulations. Subjects were excluded from the data analysis if they did not achieve 2 consecutive 80 readings at any time during the 60 minutes. n = 24, rst molar; n = 26, rst premolar; n = 18, lateral incisor.

JOE Volume 36, Number 9, September 2010

Anesthetic Efcacy of Lidocaine with Epinephrine for IAN Block

1461

Clinical Research
80% for the rst premolar, and 67% for the lateral incisor. Nusstein et al recorded success rates for 3.6 mL of 2% lidocaine with 1:100,000 epinephrine of 44% for the rst molar, 67% for the rst premolar, and 32% for the lateral incisor. Therefore, the higher success rates of Yared and Bou Dagher with the increased volume of 2% lidocaine with 1:100,000 epinephrine were not conrmed by Nusstein et al. None of the 2% lidocaine formulations provided complete pulpal anesthesia for the mandibular teeth (Table 1; Figs. 13). This could present meaningful clinical problems because the teeth might not be numb for procedures requiring complete pulpal anesthesia. Practitioners should consider supplemental techniques such as intraosseous injections (12 16), periodontal ligament injections (17), or a buccal inltration of a cartridge of 4% articaine with 1:100,000 epinephrine for the rst molar (18) when a conventional IAN block fails to provide pulpal anesthesia for a particular asymptomatic tooth. Because we studied a young adult population, the results of this study might not apply to children or the elderly. Onset of pulpal anesthesia varied from 1415.9 minutes for the 1.8 mL of 2% lidocaine with 1:50,000 epinephrine formulation, from 10.511.9 minutes for the 3.6 mL of 2% lidocaine with 1:50,000 epinephrine formulation, and from 11.613.3 minutes for the 1.8 mL of 2% lidocaine with 1:100,000 epinephrine formulation, with no signicant differences between the 3 anesthetic formulations. Therefore, increasing the epinephrine concentration to 1:50,000 epinephrine or increasing the volume to 3.6 mL of 2% lidocaine with 1:50,000 epinephrine did not increase the onset time of pulpal anesthesia when compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine. Other studies that used 1.8 mL of 2% lidocaine with 1:100,000 epinephrine for IAN blocks have reported onset of pulpal anesthesia ranged from 8.413.2 minutes (19), 8.812.3 minutes (20), 10.8 17.2 minutes (21), and 8.213 minutes (22) for the rst molar, rst premolar, and lateral incisor. The values for onset are similar to the results of the current study. Pulp testing the tooth with an electric pulp tester or cold refrigerant in asymptomatic patients will give the clinician a reliable indicator of onset of pulpal anesthesia (10, 11, 23). Onset of lip anesthesia ranged from 4.45.9 minutes, with a significant difference between the 1.8 mL of 2% lidocaine with 1:50,000 epinephrine formulation when compared with the other 2 formulations (Table 2). However, the difference of a little more than 1 minute would have little clinical signicance. Other studies that used 2% lidocaine with 1:100,000 epinephrine for IAN blocks have reported onset of lip anesthesia of 5.0 minutes (21), 6.1 minutes (20), 8.8 minutes (19), and 4.7 minutes (22). The values are somewhat higher when compared with the results of the current study (4.4 minutes), and variations between studies are probably related to differences in populations. However, the time of onset of lip numbness might not indicate the onset of pulpal anesthesia (Table 1). We concluded that increasing the epinephrine concentration to 1:50,000 epinephrine in a 2% lidocaine solution or increasing the volume to 3.6 mL of 2% lidocaine with 1:50,000 epinephrine did not result in more successful pulpal anesthesia when compared with 1.8 mL of 2% lidocaine with 1:100,000 epinephrine by using the IAN block.

References
1. Nusstein J, Reader A, Beck M. Anesthetic efcacy of different volumes of lidocaine with epinephrine for inferior alveolar nerve blocks. Gen Dent 2002;50:3725. 2. Hannan L, Reader A, Nist R, Beck M, Meyers WJ. The use of ultrasound for guiding needle placement for inferior alveolar nerve blocks. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999;87:65865. 3. Goldberg S, Reader A, Drum M, Nusstein J, Beck M. Comparison of the anesthetic efcacy of the conventional inferior alveolar, Gow-Gates, and Vazirani-Akinosi techniques. J Endod 2008;34:130611. 4. Foster W, Drum M, Reader A, Beck M. Anesthetic efcacy of buccal and lingual inltrations of lidocaine following an inferior alveolar nerve block in mandibular posterior teeth. Anesth Prog 2007;54:1639. 5. Steinkruger G, Nusstein J, Reader A, Beck M, Weaver J. The signicance of needle bevel orientation in achieving a successful inferior alveolar nerve block. J Am Dent Assoc 2006;137:168591. 6. Mikesell P, Nusstein J, Reader A, Beck M, Weaver J. A comparison of articaine and lidocaine for inferior alveolar nerve blocks. J Endod 2005;31:26570. 7. Bou Dagher F, Yared GM, Machtou P. An evaluation of 2% lidocaine with different concentrations of epinephrine for inferior alveolar nerve block. J Endod 1997;23: 17880. 8. Yared GM, Bou Dagher F. Evaluation of lidocaine in human inferior alveolar nerve block. J Endod 1997;23:5758. 9. Jorgensen NB, Hayden J Jr. Premedication, local and general anesthesia in dentistry. 2nd ed. Philadelphia, PA: Lea & Febiger; 1967. 10. Dreven L, Reader A, Beck M, Meyers W, Weaver J. An evaluation of the electric pulp tester as a measure of analgesia in human vital teeth. J Endod 1987;13:2338. 11. Certosimo A, Archer R. A clinical evaluation of the electric pulp tester as an indicator of local anesthesia. Oper Dent 1996;21:2530. 12. Reisman D, Reader A, Nist R, Beck M, Weaver J. Anesthetic efcacy of the supplemental intraosseous injection of 3% mepivacaine in irreversible pulpitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;84:67682. 13. Nusstein J, Reader A, Nist R, Beck M, Meyers WJ. Anesthetic efcacy of the supplemental intraosseous injection of 2% lidocaine with 1:100,000 epinephrine in irreversible pulpitis. J Endod 1998;24:48791. 14. Parente SA, Anderson RW, Herman WW, Kimbrough WF, Weller RN. Anesthetic efcacy of the supplemental intraosseous injection for teeth with irreversible pulpitis. J Endod 1998;24:8268. 15. Nusstein J, Kennedy S, Reader A, Beck M, Weaver J. Anesthetic efcacy of the supplemental X-tip intraosseous injection in patients with irreversible pulpitis. J Endod 2003;29:7248. 16. Bigby J, Reader A, Nusstein J, Beck M, Weaver J. Articaine for supplemental intraosseous anesthesia in patients with irreversible pulpitis. J Endod 2006;32:10447. 17. Cohen HP, Cha BY, Spangberg LSW. Endodontic anesthesia in mandibular molars: a clinical study. J Endod 1993;19:3703. 18. Haase A, Reader A, Nusstein J, Beck M, Drum M. Comparing anesthetic efcacy of articaine versus lidocaine as a supplemental buccal inltration of the mandibular rst molar after an inferior alveolar nerve block. J Am Dent Assoc 2008;139: 122835. 19. Vreeland D, Reader A, Beck M, Meyers W, Weaver J. An evaluation of volumes and concentrations of lidocaine in human inferior alveolar nerve block. J Endod 1989; 15:612. 20. Hinkley SA, Reader A, Beck M, Meyers WJ. An evaluation of 4% prilocaine with 1:200,000 epinephrine and 2% mepivacaine with 1:20,000 levonordefrin compared with 2% lidocaine with 100,000 epinephrine for inferior alveolar nerve block. Anesth Prog 1991;38:849. 21. McLean C, Reader A, Beck M, Meyers WJ. An evaluation of 4% prilocaine and 3% mepivacaine compared with 2% lidocaine (1:100,000 epinephrine) for inferior alveolar nerve block. J Endod 1993;3:14650. 22. Chaney MA, Kerby R, Reader A, Beck FM, Meyers WJ, Weaver J. An evaluation of lidocaine hydrocarbonate compared with lidocaine hydrochloride for inferior alveolar nerve block. Anesth Prog 1991;38:2126. 23. Hsiao-Wu GW, Susarla SM, White RR. Use of the cold test as a measure of pulpal anesthesia during endodontic therapy: a randomized, blinded, placebo-controlled clinical trial. J Endod 2007;33:40610.

1462

Wali et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Photodynamic Therapy Associated with Conventional Endodontic Treatment in Patients with Antibiotic-resistant Microora: A Preliminary Report
Aguinaldo S. Garcez, PhD,* Silvia C. Nunez, PhD, Michael R. Hamblim, PhD,k Hideo Suzuki,* and Martha S. Ribeiro, PhD
Abstract
Introduction: This study reports the antimicrobial effect of photodynamic therapy (PDT) combined with endodontic treatment in patients with necrotic pulp infected with microora resistant to a previous antibiotic therapy. Methods: Thirty anterior teeth from 21 patients with periapical lesions that had been treated with conventional endodontic treatment and antibiotic therapy were selected. Microbiological samples were taken (1) after accessing the root canal, (2) after endodontic therapy, and (3) after PDT. Results: All the patients had at least 1 microorganism resistant to antibiotics. PDT used polyethylenimine chlorin(e6) as a photosensitizer and a diode laser as a light source (P = 40 mW, t = 4minutes, E = 9.6 J). Endodontic therapy alone produced a signicant reduction in numbers of microbial species but only 3 teeth were free of bacteria, whereas the combination of endodontic therapy with PDT eliminated all drug-resistant species and all teeth were bacteria-free. Conclusions: The use of PDT added to conventional endodontic treatment leads to a further major reduction of microbial load. PDT is an efcient treatment to kill multi-drug resistant microorganisms. (J Endod 2010;36:14631466)

n the case of endodontic treatment failure, retreatment, surgical treatment, or extraction usually is carried out with the use of antibiotics and antiseptics as adjunctive therapies, but the long-term use of these agents can be rendered ineffective by resistance developing in the target organism (1). Currently, there is an emergence of bacteria with multiple resistances, and there is a need for alternative antimicrobial approaches (26). The combination of conventional endodontic therapy and photodynamic therapy (PDT) has been shown as an effective approach in reducing bacterial load in in vitro and in vivo models (711). This study investigated the combination of PDT with endodontic treatment in patients with necrotic pulp harboring microora resistant to a previous antibiotic therapy.

Materials and Methods

Thirty teeth from 21 patients with periapical lesions who had been previously treated with endodontic treatment associated with antibiotic were selected. The patients were in good health and between the ages of 17 and 52 years. All the teeth presented sights and symptoms of periapical periodontitis and apical bone lesion detected by radiography, and some patients had pain by vertical percussion and/or local edema, all requiring root canal retreatment on teeth with closed apices. The same practitioner carried out this study in a private dental ofce in Sao Paulo, Brazil. The protocol was approved by the Institutional Review Board of the Sao Paulo University, and all procedures were conducted according to the principles of the Declaration of Helsinki.

Key Words
Antibiotic resistant bacteria, endodontic re-treatment, laser, photodynamic therapy

From the *Centro de Pesquisa e Pos-Graduacao Sao Leo poldo Mandic, Campinas, SP, Brazil; CETAO, Sao Paulo, SP, Brazil; Wellman Center for Photomedicine, Massachusetts General Hospital, Boston, Massachusetts; Department of Dermatology, Harvard Medical School, Boston, Massachusetts; k Harvard MIT Division of Health Science and Technology, Cambridge, Massachusetts; and Center of Lasers and Applications, IPEN-CNEN/SP, Sao Paulo, SP, Brazil. Address requests for reprints to Dr Aguinaldo Silva Garcez, Sao Leopoldo Mandic University, Campinas, SP, Brazil. E-mail address: garcez.segundo@terra.com.br. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.001

Endodontic Treatment Thirty root canals from anterior teeth were re-treated and received endodontic treatment followed by PDT. Microbiological samples were taken after accessing the root canal, after endodontic therapy, and after PDT. The rst microbiological sample conrmed that all the patients had at least 1 microorganism resistant to antibiotic medication. A periapical radiograph was taken for each case to determine the presence of apical lesion, the canal morphology, and its length. The access to the pulp chamber was gained after installation of a rubber dam, and then the surrounding area received prophylactic asepsis and was irrigated with 5 mL of chlorhexidine solution at 2% to ensure that the crown of the tooth had minimal microbial load (8). Once the canal was accessed, a Hedstrom le #15 (Maillefer Instruments SA, Ballaigues, Switzerland) was inserted inside the canal to remove the gutta-percha and root canal sealer obturation; then the root canal was irrigated with 1 mL of sterile saline solution. The canal was dried with 3 sterile paper points (Dentsply Latin America, Petropolis, Brazil) and left inside the root canal for 1 minute each. All 3 paper points were combined for microbiological analysis. This procedure was the rst microbiological sampling representing the initial contamination. The paper points were deposited in a fresh sterile bottle with sterile nutrient broth.

JOE Volume 36, Number 9, September 2010

Antimicrobial Effect of PDT Combined with Endodontic Treatment

1463

Clinical Research
The canals were prepared with manual instrumentation by K les (Maillefer Instruments SA) by using a standard crown-down technique working to 1 mm short of the working length (le #45 was the average apical preparation diameter). Ten milliliters of sodium hypochlorite at 2.5% and hydrogen peroxide at 3% was alternated between each instrumentation by using an endodontic needle (27-gauge). At the end of the procedure the root canals were irrigated with 5 mL of 17% ethylenediaminetetraacetic acid followed by irrigation with 5 mL of phosphatebuffered saline (PBS) solution to remove the smear layer (12). The canal was irrigated with 5 mL of sterile saline solution to remove the antimicrobial agent and dried with another 3 paper points by using the same methodology cited above (second microbiological sample).

Photosensitizer The photosensitizer used was a conjugate between polyethylenimine (PEI) and chlorin(e6), which has been previously described in detail (13). The photosensitizer was used in a PBS solution at 60 mmol/L. Light Source The illumination was performed with a disposable 200-mm diameter ber-coupled diode laser (MMOptics, Sao Paulo, Brazil). The laser delivered 660 nm light at a total power of 40 mW out of the ber. The ber was placed in the apical portion of the root canal at a point where resistance to the ber was just felt (usually 1 mm from the apex), and spiral movements, from apical to cervical, were manually performed to ensure even diffusion of the light inside the canal lumen (14, 15). After the endodontic procedure, the canal was irrigated with 0.5 mL of the photosensitizer and left inside the root canal for 2 minutes as an incubation time. The root canal was then irradiated for 240 seconds (total energy, 9.6 J), and the ber was changed between each patient. The root canal was again irrigated with 10 mL of sterile saline solution to remove the photosensitizer and dried as before (third microbiological sample). A calcium hydroxide paste (Ultradent Products, South Jordan, UT) was placed into the canals; cotton was placed in the pulp chamber, and the teeth were dressed with temporary restorative material (IRM; Dentsply Latin America). One week later, a second session of each therapy was performed without microbiological sampling. Thereafter, root canal was sealed by using conventional techniques with Sealer 26 (Dentsply Petropolis), and the tooth was restored with a composite resin Z250 (3M, Sumare, Brazil). This 1-week interappointment dressing approach was used by Garcez et al (8). Briey, the pH in the environment is increased; consequently, the live-time of reactive oxygen species increases, and the photodynamic effect is improved at the second session. Microbiological Analyses The samples were sent in a sterile bottle with fresh sterile nutrient broth (Viability Medium Goteborg Agar III) to a private microbiological facility for identication and to antibiogram analyses. The bacterial species were identied on the basis of Gram stain, aerotolerance, colony morphology, esculin hydrolysis, nitrate reduction, indole production, (alpha)-glycosidase and N-benzoyl-DL-arginine-2-naphthylamide (BANA) hydrolysis, oxidase and catalase activities. The antibiogram tested 17 different antibiotics by using the Kirby-Bauer method (16).

Figure 1. Means and standard deviations of multi-drug resistant bacteria inside root canal in each step of the treatment.

antibiotic therapy. The number of multi-drug resistant bacterial species did vary widely between individual teeth, with a mean value of 2.16 species per root canal sample (range, 41). This was probably due to differences in the geometry of the root canal systems and initial contamination. The mean values of the number of species for each step are given in Fig. 1. Among the initial samples, 33% were gram-negative, and 67% were gram-positive bacteria; moreover, 57% were facultative anaerobes, and 43% were obligate anaerobes. After the endodontic therapy the infectious burden was reduced to 0.8 species per root canal (range, 20). After PDT, microorganism growth was not detected on any of the samples from any of the root canals. Ten of the root canals treated had 100% bacterial elimination after endodontic treatment, whereas all 30 teeth showed total absence of microorganisms after the combination. The multi-drug resistant bacteria found in the initial samples were, in order of prevalence, Enterococcus sp, Prevotella sp, Actinomyces sp, Peptostreptococcus sp, Streptococcus sp, Fusobacterium sp, Porphyromonas sp, Enterobacter sp, and Propionibacterium sp. After the endodontic therapy, the species found were Enterococcus sp, Actinomyces sp, Peptostreptococcus sp, Fusobacterium sp, and Porphyromonas sp. All teeth were completely free of bacteria after the 2 combination therapies. Fig. 2 shows the number of bacterial species that grew in each sample from each stage of the therapy.

Results
The rst samples showed that all teeth harbored at least 1 resistant microorganism, indicating unsuccessful previous treatment and/or 1464
Figure 2. Bacteria species per patient in each step of the treatment.

Garcez et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
killed multi-drug resistant and native Escherichia coli strains equally and killed MRSA better than the sensitive strain. Our results showed that the combination of endodontic therapy and PDT killed all 9 multi-drug resistant bacterial species found in root canal infections. Therefore, PDT not only kills multi-drug resistant bacteria in vitro but is also effective in eliminating species resistant to diverse antibiotics in patients. The samples showed that the multi-drug resistant bacteria found consisted of facultative and obligate anaerobic species. However, it is well-known that aerobic microorganisms can deal better with reactive oxygen species, and the greater susceptibility of anaerobes to the reactive oxygen species produced during PDT could explain the 100% reduction of multi-drug resistant bacteria after the combination therapy. Furthermore, the majority of the species found were grampositive, and the literature has shown that PDT is more efcient in killing these microorganisms (6, 7, 13, 17). Nevertheless, the photosensitizer used in this study (PEI-ce6) has also a high efcacy in killing gramnegative species compared with alternative photosensitizers such as toluidine blue (31). In fact, despite several attempts to induce resistance, the use of PDT to kill bacteria has not resulted in the generation of any PDT resistance among treated species, suggesting that bacteria do not nd it easy to develop defenses against the reactive oxygen species generated during PDT (32). In addition, the literature has showed that it is safe to use PDT against microorganisms near normal cells, for example, cells from apical region. George and Kishen (33) showed that cytotoxicity was signicantly less in PDT compared with conventional antimicrobial irrigation. In an in vitro experiment, Enterococcus faecalis were killed at a faster rate than normal broblasts. PDT produced 97.7% bacterial killing and only 30% broblast dysfunction. Also Xu et al (34) suggested that there is a safe therapeutic window whereby PDT can inactivate endodontic pathogens without affecting host cell viability.

Figure 3. Number of species resistant to each type of antibiotic.

The antibiogram showed bacteria resistant to ampicillin, penicillin G, vancomycin cephalosporin, clindamycin, chloramphenicol, erythromycin, and tetracycline. Fig. 3 shows the number of species that were resistant to each antibiotic.

Discussion
Previous studies from our group (7, 8) and from other groups (9, 17, 18) showed that a combination of conventional endodontic therapy followed by antimicrobial PDT was effective in reducing bacterial load in ex vivo root canals (for planktonic and biolm endodontic microorganisms) and in patients. In both studies we used the same photosensitizer, a conjugate between PEI and chlorin(e6) (PEI-ce6) that has been designed to have a broad-spectrum antimicrobial effect under illumination (19). This study shows for the rst time, in vivo, the susceptibility of drug-resistant bacteria in root canal infections to PDT. The literature reports that endodontic therapy will have a 94% success rate when a negative microbiological culture is obtained from the root canal at the time of obturation. On the other hand, when obturation is performed and the cultures are positive, the success rate is reduced to 68%; in the case of a periapical lesion, the failure of healing is more likely when the canal is obturated in the presence of persistent infection (20, 21). Treatment procedures to eliminate the infection include root canal debridement and mechanical shaping or smoothing (22), irrigation with a disinfectant agent, the application of an interappointment dressing, and sealing of the root canal (23). In case of infection, the use of antibiotics and antiseptics is an alternative approach, but the long-term use of antimicrobial agents, however, can be rendered ineffective by resistance developing in the target organisms (2426). Our results conrm that the long-term use of antibiotics can lead to development of resistance in microorganisms. Endodontic treatment alone had an effect in reducing the number of multi-drug resistant species in root canals and produced a total bacterial elimination in 10 of 30 teeth, but the addition of PDT produced a reduction in bacterial burden leading to total elimination in all teeth. Previous studies compared photodynamic antimicrobial therapy of multi-drug resistant bacteria with wild-type strains. Maisch et al (27) found identical killing of methicillin-resistant Staphylococcus aureus (MRSA) and native strain. Wainwright et al (28) showed that PDT killed MRSA somewhat less efciently than the native strain; also Embleton et al (29) used a phage delivery system to carry out PDT with the photosensitizer Sn-ce6 and again found that MRSA was susceptible. Tang et al (30) showed that PDT with a polylysine-chlorin(e6)

Conclusions
Our results suggest that the use of PDT as an adjuvant to conventional endodontic treatment leads to a signicant further reduction of bacterial load and is effective against multi-drug resistant bacteria. PDT offers an efcient means of destroying multi-drug resistant bacteria remaining inside the root canal system after using conventional endodontic chemomechanical therapy.

Acknowledgments
The authors deny any conicts of interest.

References
1. Siqueira JF Jr. Endodontic infections: concepts, paradigms, and perspectives. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;94:28193. 2. Yoshikawa TT. Antimicrobial resistance and aging: beginning of the end of the antibiotic era? J Am Geriatr Soc 2002;50:s2269. 3. Hancock RE, Bell A. Antibiotic uptake into gram-negative bacteria. Eur J Clin Microbiol Infect Dis 1988;7:71320. 4. Livermore DM. Antibiotic resistance in staphylococci. Int J Antimicrob Agents 2001; 16:310. 5. Agarwal R, Athar M, Urban SA, Bickers DR, Mukhtar H. Involvement of singlet oxygen in chloroaluminum phthalocyanine tetrasulfonate-mediated photoenhancement of lipid peroxidation in rat epidermal microsomes. Cancer Lett 1991;56: 1259. 6. Phoenix DA, Sayed Z, Hussain S, Harris F, Wainwright M. The phototoxicity of phenothiazinium derivates against Escherichia coli and Staphylococcus aureus. FEMS Immun Med Microbiol 2003;39:1722. 7. Garcez AS, Ribeiro MS, Tegos GP, Nunez SC, Jorge AO, Hamblin MR. Antimicrobial photodynamic therapy combined with conventional endodontic treatment to eliminate root canal biolm infection. Lasers Surg Med 2007;39:5966.

JOE Volume 36, Number 9, September 2010

Antimicrobial Effect of PDT Combined with Endodontic Treatment

1465

Clinical Research
8. Garcez AS, Nunez SC, Hamblin MR, Ribeiro MS. Antimicrobial effects of photodynamic therapy on patients with necrotic pulps and periapical lesion. J Endod 2008;34:13842. 9. George S, Kishen A. Inuence of photosensitizer solvent on the mechanisms of photoactivated killing of Enterococcus faecalis. Photochem Photobiol 2008;84:73440. 10. Bergmans L, Moisiadis P, Huybrechts B, Van Meerbeek B, Quirynen M, Lambrechts P. Effect of photo-activated disinfection on endodontic pathogens ex vivo. Int Endod J 2008;41:22739. 11. Bonsor SJ, Nichol R, Reid TM, Pearson GJ. An alternative regimen for root canal disinfection. Br Dent J 2006;201:1015. 12. Haapasalo M, Orstavik D. In vitro infection and disinfection of dentinal tubules. J Dent Res 1987;66:13759. 13. Tegos GP, Anbe M, Yang C, et al. Protease-stable polycationic photosensitizer conjugates between polyethyleneimine and chlorin(e6) for broad-spectrum antimicrobial photoinactivation. Antimicrob Agents Chemother 2006;50:140210. 14. Gutknecht N, van Gogswaardt D, Conrads G, Apel C, Schubert C, Lampert F. Diode laser radiation and its bactericidal effect in root canal wall dentin. J Clin Laser Med Surg 2000;18:5760. 15. Silva Garcez A, Nunez SC, Lage-Marques JL, Jorge AO, Ribeiro MS. Efciency of NaOCl and laser-assisted photosensitization on the reduction of Enterococcus faecalis in vitro. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:e938. 16. Bauer AW, Kirby WM, Sherris JC, Turck M. Antibiotic susceptibility testing by a standardized single disk method. Am J Clin Pathol 1966;45:4936. 17. Fimple JL, Fontana CR, Foschi F, et al. Photodynamic treatment of endodontic polymicrobial infection in vitro. J Endod 2008;34:72834. 18. Soukos NS, Chen PS, Morris JT, et al. Photodynamic therapy for endodontic disinfection. J Endod 2006;32:97984. 19. Tegos GP, Anbe M, Yang C, et al. Protease-stable polycationic photosensitizer conjugates between polyethyleneimine and chlorin(e6) for broad-spectrum antimicrobial photoinactivation. Antimicrob Agents Chemother 2006;50:140210. 20. Sjogren U, Figdor D, Persson S, Sundqvist G. Inuence of infection at the time of root lling on the outcome of endodontic treatment of teeth with apical periodontitis. Int Endod J 1997;30:297306. Erratum in: Int Endod J 1998;31:148. 21. Nair PN, Sjogren U, Krey G, Kahnberg KE, Sundqvist G. Intraradicular bacteria and fungi in root-lled, asymptomatic human teeth with therapy-resistant periapical 22. 23. 24. lesions: a long-term light and electron microscopic follow-up study. J Endod 1990;16:5808. Bahcall JK, Barss JT. Understanding and evaluating the endodontic le. Gen Dent 2000;48:6902. Sedgley C. Root canal irrigation: a historical perspective. J Hist Dent 2004;52:615. Reynaud AF, Geijersstam AH, Ellington MJ, Warner M, Woodford N, Haapasalo M. Antimicrobial susceptibility and molecular analysis of Enterococcus faecalis originating from endodontic infections in Finland and Lithuania. Oral Microbiol Immunol 2006;21:1648. Pinheiro ET, Gomes BP, Drucker DB, Zaia AA, Ferraz CC, Souza-Filho FJ. Antimicrobial susceptibility of Enterococcus faecalis isolated from canals of root lled teeth with periapical lesions. Int Endod J 2004;37:75663. Figdor D. Microbial aetiology of endodontic treatment failure and pathogenic properties of selected species. Aust Endod J 2004;30:114. Maisch T, Bosl C, Szeimies RM, Lehn N, Abels C. Photodynamic effects of novel XF porphyrin derivatives on prokaryotic and eukaryotic cells. Antimicrob Agents Chemother 2005;49:154252. Wainwright M, Phoenix DA, Laycock SL, Wareing DR, Wright PA. Photobactericidal activity of phenothiazinium dyes against methicillin-resistant strains of Staphylococcus aureus. FEMS Microbiol Lett 1998;160:17781. Embleton ML, Nair SP, Heywood W, Menon DC, Cookson BD, Wilson M. Development of a novel targeting system for lethal photosensitization of antibiotic-resistant strains of Staphylococcus aureus. Antimicrob Agents Chemother 2005;49:36906. Tang HM, Hamblin MR, Yow CM. A comparative in vitro photoinactivation study of clinical isolates of multidrug-resistant pathogens. J Infect Chemother 2007;13:8791. Demidova TN, Hamblin MR. Effect of cell-photosensitizer binding and cell density on microbial photoinactivation. Antimicrob Agents Chemother 2005;49:232935. Lauro FM, Pretto P, Covolo L, Jori G, Bertoloni G. Photoinactivation of bacterial strains involved in periodontal diseases sensitized by porphycene-polylysine conjugates. Photochem Photobiol Sci 2002;1:46870. George S, Kishen A. Advanced noninvasive light-activated disinfection: assessment of cytotoxicity on broblast versus antimicrobial activity against Enterococcus faecalis. J Endod 2007;33:599602. Xu Y, Young MJ, Battaglino RA, et al. Endodontic antimicrobial photodynamic therapy: safety assessment in mammalian cell cultures. J Endod 2009;35:156772.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

1466

Garcez et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Antigenic Activity of Bacterial Endodontic Contents from Primary Root Canal Infection with Periapical Lesions against Macrophage in the Release of Interleukin-1b and Tumor Necrosis Factor a
Frederico C. Martinho, DDS, MSc,* Wanderson Miguel Maia Chiesa, DDS, MSc,* Fabio R.M. Leite, DDS, MSc, PhD, Joni A. Cirelli, DDS, MSc, PhD, and Brenda P.F.A. Gomes, DDS, MSc, PhD*
Abstract
Introduction: Periradicular tissue chronic stimulation by endotoxin may cause apical periodontitis. This study investigated the microbial prole and the levels of endotoxin found in primary root canal infection with apical periodontitis, determined their antigenicity against macrophages through the levels of interleukin (IL)-1b and tumor necrosis factor a (TNF-a), and evaluated their relationship with clinical and radiographic ndings. Methods: Samples were taken from 21 root canals with primary endodontic infection and apical periodontitis with paper points. PCR technique (16S rDNA) was used for the detection of the target bacteria. Limulus Amebocyte Lysate (LAL) was used to measure endotoxin. The amounts of IL-1/ TNF-alpha in macrophages supernatants were measured by enzyme-linked immunosorbent assay Duoset-kit (ELISA). Results: Prevotella nigrescens (13/21), Porphyromonas endodontalis (6/21), and Treponema socranskii (6/21) were the most frequently detected gram-negative bacterial species. The presence of the sinus tract (2/21) was related to the detection of Filifactor alocis (p < 0.05), whereas a tooth with a radiolucent area $2 mm was related to the detection of Treponema denticola. A correlation was found between the number of gram-negative bacteria and the levels of IL-1b/TNF-a (p < 0.05). Increased levels of endotoxin were followed by TNF-a release (p < 0.05). Higher levels of IL-1b (p < 0.05) and endotoxin contents were related to the larger size of the radiolucent area. Conclusion: The antigenicity of the endodontic contents is not only related to the amount of endotoxin found in the root canal but also to the number of different species of gram-negative bacteria involved in the infection. Moreover, a larger size ($2 mm) of the radiolucent area was related to IL-1b and endotoxin. (J Endod 2010;36:14671474)

Key Words
Antigenicity, bacteria, endodontic, endotoxin, macrophages

From the *Department of Restorative Dentistry, Endodontic Division, Piracicaba Dental School, State University of Campinas, Campinas, Brazil; Department of Semiology and Clinics, Periodontic Division, Dental School, Federal University of Pelotas, Pelotas, Brazil; and Department of Diagnosis and Oral Surgery, Periodontic Division, Araraquara Dental School, State University of Sao Paulo, Sao Paulo, Brazil. Supported by the Brazilian agencies FAPESP (07/58518-4, 08/58299-3, 08/ 57954-8, and 08/57551-0) and CNPq (3470820/2006-3, 471631/2008-6, and 302575/2009-0). Address requests for reprints to Dr Brenda P.F.A. Gomes, Piracicaba Dental School, State University of CampinasUNICAMP, Department of Restorative Dentistry, Endodontic Division, Av Limeira 901, Bairro Areiao, Piracicaba, Sao Paulo, Brazil CEP 13414-903. E-mail address: bpgomes@fop.unicamp. br. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.012

rimary endodontic infection is a polymicrobial infection caused predominantly by gram-negative anaerobic bacteria (1) that present lipopolysaccharide (LPS) on the outer layers of their cell walls. LPS is released during disintegration, multiplication, or bacterial death (2) and is capable of penetrating into the periradicular tissues (3), acting as endotoxin in the host organism (4) and leading to periradicular inammation and bone destruction (5). The lipid A is the bioactive component of LPS responsible for the majority of the immunoresponse (3). The accumulation of bacteria components in an infected area, particularly endotoxins (including lipoteichoic acid, peptidoglycan, lipopolysaccharide, and others), can stimulate the release of proinammatory cytokines by different cell lines through TLR2 and -4 activation (57). The inammatory tissue present in periradicular lesions is populated predominantly by macrophage (8, 9), which is the major source of interleukin-1b (IL-1b) (10), and almost the exclusive producer of tumor necrosis factor a (TNF-a) (11) in the presence of bacteria or LPS. Clinical investigations of primary endodontic infection have elucidated the strong correlation between oral bacteria LPS and the presence of apical periodontitis (6, 1215). Moreover, higher contents of endotoxins in root canals have been associated with spontaneous pain (6, 12, 16) and clinical signs/symptoms such as pain on palpation, tenderness to percussion, and exudation (6, 12, 14, 16). Previous in vitro investigations (57) have shown that oral bacterial LPS extracted from bacteria commonly found in root canal infection induces a potent inammatory response against different cell line cultures. IL-1b and TNF-a have been detected in periapical tissues (3, 9, 1719) and root canal exudates (2023) from primary root canal infection in the presence apical periodontitis. Higher contents of IL-1b have been detected in teeth with clinical signs/symptoms (3, 23) and larger size of radiolucent area corresponding to bone resorption (21, 22). However, studies correlating all these aspects have not yet been provided in the literature. Therefore, the aim of this clinical study was to investigate the microbial prole and the levels of endotoxin found in primary root canal infection with apical periodontitis and to determine their antigenicity against macrophages through the levels of IL-1b/TNF-a, evaluating their relationship with clinical and radiographic ndings.

JOE Volume 36, Number 9, September 2010

Antigenic Activity of Bacteria from Primary Root Canal Infection with Periapical Lesions

1467

Clinical Research
Material and Methods
Patient Selection Twenty-one patients who attended the Piracicaba Dental School, Piracicaba, SP, Brazil, in need of endodontic treatment were included in this research. The age of the patients ranged from 13 to 73 years. Samples were collected from 21 root canals with pulp necrosis and showing radiographic evidence of apical periodontitis. The selected teeth showed absence of periodontal pockets more than 4 mm. The following clinical/radiographic features were found: pain on palpation (9/21), tenderness to percussion (8/21), exudation (12/21), and radiolucent area $2 mm (11/21) and <2 mm (10/21). None of the patients reported spontaneous pain. A detailed dental history was obtained from each patient. Patients who had received antibiotic treatment during the last 3 months or who had any general disease were excluded. The Human Volunteers Research and Ethics Committee of the Piracicaba Dental School approved the protocol describing specimen collection for this investigation, and previously all patients signed an informed consent document. Sampling Procedures All materials used in this study were heat sterilized at 200 C for 4 hours to become apyrogenic. The method followed for the disinfection of the operative eld had been described previously (14, 15). Briey, the teeth were isolated with a rubber dam. The crown and the surrounding structures were disinfected with 30% H2O2 for 30 seconds followed by 2.5% NaOCl for an additional 30 seconds. Subsequently, 5% sodium thiosulphate was used to inactivate the irrigant. The sterility of the of the external surfaces of the crown was checked by taking a swab sample from the crown surface and streaking it on blood agar plates that were incubated aerobically and anaerobically. A two-stage access cavity preparation was made without the use of water spray but under manual irrigation with sterile/apyrogenic saline solution and by using sterile/apyrogenic high-speed diamond bur. The rst stage was performed to promote a major removal of contaminants. In the second stage, before entering the pulp chamber, the access cavity was disinfected following the protocol described earlier. The sterility of the internal surface of the access cavity was checked as previously described, and all procedures were performed aseptically. A new sterile and apyrogenic bur was used, accomplished by irrigation with sterile apyrogenic water to access the canal. The endotoxin sample was taken introducing sterile pyrogen-free paper points (size 15; DentsplyMaillefer, Ballaigues, Switzerland) into the full length of the canal (determined radiographically) and retained in position during 60 seconds. Immediately, the paper point was placed in a pyrogen-free glass and frozen at 80 C for future Limulus amebocyte lysate assay (LAL) and cell culture stimulation. The procedure was repeated with ve sterile paper points. The paper points were pooled in a sterile tube containing 1 mL of VMGA (Viability Medium Goteborg Agar) III transport medium and immediately processed for DNA extraction for the detection of target bacteria by molecular method (16S ribosomal DNA). Bacterial Detection (PCR 16S rDNA) Reference bacteria strains used in this study were purchased from the American Type Culture Collection (ATCC) and are listed as follows: Dialister pneumosintes (ATCC 33048), Prevotella intermedia (ATCC 25611), Prevotella nigrescens (ATCC 33099), Aggregatibacter actinomycetemcomitans (ATCC 43718), Porphyromonas gingivalis (ATCC 33277), Filifactor alocis (ATCC 35896), Tannerella forsythia
1468
Martinho et al.

(ATCC 43037), Prevotella tannerae (ATCC 51259), Treponema denticola (ATCC 35405), Porphyromonas endodontalis (ATCC 35406), Treponema socranskii (35536), and Parvimonas micra (ATCC 33270). Bacterial selection criteria were performed based on the most commonly found species in primarily root canal infection. DNA Extraction. Microbial DNA from endodontic samples as well as from ATCC bacteria were extracted and puried with the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers instructions. The DNA concentration (absorbance at 260 nm) was determined using a spectrophotometer (Nanodrop 2000; Thermo Scientic, Wilmington, DE). PCR Assay. The PCR reaction was performed in a thermocyler (MyCycler; Bio-Rad, Hercules, CA) thermocycler in a total volume of 25 mL containing 2.5 mL of 10X Taq buffer (1) (MBI Fermentas, Mundolsheim, France), 0.5 mL of dNTP mix (25 mmol/L of each deoxyribonucleoside triphosphatedATP, dCTP, dGTP, and dTTP) (MBI Fermentas, Hanover, MD), 1.25 mL of 25 mmol/L MgCl2, 0.25 mL of forward and reversal universal primers (0.2 mmol/L) (Invitrogen, Eugene, OR), 1.5 mL of sample DNA (1 mg/50 mL), 1.5 mL of Taq DNA polymerase (1 U) (MBI Fermentas), and 17.25 mL of nuclease-free water. The primer sequences and PCR cycling parameters were previously optimized (13-15) and listed in Table 1.

Determination of Endotoxin Concentration (Turbidimetric Test and LAL Assay) The turbidimetric test (BioWhitaker, Inc, Walkersville, MD) was used to measure endotoxin concentrations in the root canals using the LAL technique. First, as a parameter for the calculation of the amount of endotoxins in root canal samples, a standard curve was plotted using endotoxins supplied in the kit with a known concentration (100 EU/mL) and its dilutions with the following nal concentrations (ie, 0.01, 0.10, 1, and 10 EU/mL) following the manufacturers instructions. Test Procedure. All reactions were accomplished in duplicate to validate the test. A 96-well microplate (Corning Costar, Cambridge, MA) was used in a heating block at 37 C and maintained at this temperature throughout the assay. First, the endotoxin samplings were suspended in 1 mL of LAL water supplied on the kit and agitated in vortex for 60 seconds and serial diluted to the 101. Immediately, 100 mL of the blank followed the standard endotoxin solutions in concentrations (ie, 0.01, 0.10, 1, and 10 EU/mL), and 100 mL of the samples were added in duplicate in the 96-well microplate. The test procedure was performed following the manufacturers instructions. The absorbencies of endotoxin were measured individually using an enzyme-linked immunosorbent assay plate reader (Ultramark, BioRad Laboratories, Inc) at 340 nm. Calculation of Endotoxin Concentrations. Because the mean absorbance value of the standards was directly proportional to the concentration of endotoxins present, the endotoxin concentration was determined from the standard curve. Cell Culture and Cytokine Expression Macrophages (RAW 264.7) were cultured in 100-mm culture plates in Dulbeccos modied Eagle minimal essential medium supplemented (DMEM) with 100 IU/mL of penicillin, 100 mg/mL of streptomycin, and 10% heat-inactivated fetal bovine serum and maintained in a humidied atmosphere at 37 C and 5% CO2 until 90% conuence. Unless noted otherwise, all tissue culture reagents were obtained from Invitrogen (Carlsbad, CA). Macrophages were released from 100-mm plates with 0.25% trypsin and counted in a Newbauer chamber; a total of 104 macrophages were grown for 48 hours in each well of six-well
JOE Volume 36, Number 9, September 2010

Clinical Research
TABLE 1. PCR Primer Pairs and Cycling Parameters Used for Detection of Bacteria Species in Primary Root Canal Infection with Apical Periodontitis Target bacteria
Universal (16s rDNA)

Primer pairs (5- 3)

Forward: TCC TAC GGG AGG CAG CAG T Reverse: GGA CTA CCA GGG TAT CTA ATC CTG TT Forward: TTC TAA GCA TCG CAT GGT GC Reverse: GAT TTC GCT TCT CTT TGT TG Forward: TTT GTT GGG GAG TAA AGC GGG Reverse: TCA ACA TCT CTG TAT CCT GCG T Forward: ATG AAA CAA AGG TTT TCC GGT AAG Reverse: CCC ACG TCT CTG TGG GCT GCG A Forward: AAA CCC ATC TCT GAG TTC TTC TTC Reverse: ATG CCA ACT TGA CGT TAA AT Forward: AGG CAG CTT GCC ATA CTG CG Reverse: ACT GTT AGC AAC TAC CGA TGT Forward: CAG GTG GTT TAA CAA GTT AGT GG Reverse: CTA AGT TGT CCT TAG CTG TCT CG Forward: GCG TAT GTA ACC TGC CCG CA Reverse: TGC TTC AGT GTC AGT TAT ACC T Forward: CTT AGC TTG CTA AGT ATG CCG Reverse: CAG CTG ACT TAT ACT CCC G Forward: TAA TAC CGA ATG TGC TCA TTT ACA T Reverse: TCA AAG AAG CAT TCC CTC TTC TTC TTA Forward: GCT GCA GCT CAA CTG TAG TC Reverse: CCG CTT CAT GTC ACC ATG TC Forward: GAT CAC TGT ATA CGG AAG GTA GAC A Reverse: TAC ACT TAT TCC TCG GAC AG Forward: AGA GTT TGA TCC TGG CTC AG Reverse: ATA TCA TGC GAT TCT GTG GTC TC

Amplicon size
466 bp

Cycles
Initial denaturation at 95 C for 10 min and 40 cycles of 95 C for 10 s, 60 C for 10 s, and a nal extension step at 72 C for 25 s Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 55 C for 1 min, 72 C for 2 min, and a nal step 72 C for 2 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 58 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 58 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 94 C for 30 s and 36 cycles of: 95 C for 30 s, 55 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 60 C for 1 min, 72 C for 2 min, and a nal step 72 C for 2 min Initial denaturation at 95 C for 2 min and 26 cycles of 95 C for 30 s, 58 C for 1 min, 72 C for 1 min, and a nal step 72 C for 2 min Initial denaturation at 95 C for l min and 36 cycles of 95 C for 30 s, 60 C for 1 min, 72 C for 1 min, and a nal step 72 C for 2 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 55 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 60 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 58 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 56 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min Initial denaturation at 95 C for 2 min and 36 cycles of 94 C for 30 s, 60 C for 1 min, 72 C for 2 min, and a nal step 72 C for 10 min

Dialister pneumosintes

1105 bp 575 bp

Prevotella intermedia

Prevotella nigrescens

804 bp

Aggregatibacter actinomycetemcomitans Porphyromonas gingivalis

557 bp 404 bp

Filifactor alocis

594 bp

Tannerella forsythia

641 bp 550 bp

Prevotella tannerae

Treponema denticola

316 bp

Porphyromonas endodontalis Treponema socranskii

672 bp 288 bp

Parvimonas micra

207 bp

plates, deinduced by incubation for 8 hours in culture medium (DMEM) containing 0.3% fetal bovine serum, and stimulated with 60 mL of root canal contents during 24 hours in order to quantify the total amount of protein released in the culture media, IL-1b, and TNFa protein. The supernatants were collected and stored at 80 C until protein evaluation.

IL-1b and TNF-a Messenger RNA Expression The macrophage cell viability was tested in the present study by its capacity to express IL-1b and TNF-a messenger RNA after 24 hours of root canal contents stimulation. A total of 104 macrophages were grown for 48 hours in each well of six-well plates, deinduced by incubation for
JOE Volume 36, Number 9, September 2010

8 hours in culture medium (DMEM) containing 0.3% fetal bovine serum, and stimulated with 60 uL of primary infection contents for 24 hours for IL-1b and TNF-a messenger RNA expression. Total RNA was isolated from cells using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturers instructions. The quantity and purity of total RNA were determined on a Biomate 3 spectrophotometer (ThermoSpectronic, Rochester, NY). Complementary DNA was synthesized by reverse transcription of 500 ng of total RNA using 2.5 mmol/L Oligo (dT)12-18 primers and 1.25 U/mL Moloney murine leukemia virus reverse transcriptase in the presence of 3 mmol/L MgCl2, 2 mmol/L dNTPs, and 0.8 U/mL RNAse inhibitor according to the manufacturers protocol (Improm II; Promega, Madison, WI, USA). The PCR reaction was performed in a MyCycler (Bio-Rad) thermocycler using 2 mL of the 1469

Antigenic Activity of Bacteria from Primary Root Canal Infection with Periapical Lesions

Clinical Research
reverse transcription reaction product on a 20-uL total volume PCR reaction mix (GoTaq Flexi, Promega) in the presence of 100 pmol/mL of each genes primers (50 pmol/mL of sense and antisense primers) for IL-1b, TNF-a, and GAPDH genes yielding products of 494, 451, and 418 bp, respectively. The primer pair used for IL-1b (accession no.: NM031512) was sense 5-GACCTGTTCTTTGAGGCTGA-3, antisense 5-CGTTGCTTGTCTCTCCTTGT-3; TNF-alpha (accession no.: NM012675) sense 5-GGAGAACAGCAACTCCAGAA-3, and antisense 5-TCTTTGAGATCCATGCCATT-3 and for GAPDH (accession no.: BC083065) sense 5-CACCATGGAGAAGGCCGGGG-3, and antisense 5-GACGGACACATTGGGGTAG- 3. The optimized cycling conditions used for TNF-a and IL-1b were initial denaturation at 95 C for 2 minutes and 35 cycles of 95 C for 1 minute, 58 C for 1 minute, 72 C for 2 minutes, and a nal extension step at 72 C for 7 minutes in the presence of 1.5 mmol/L MgCl2. For GAPDH, conditions were as follows: initial denaturation at 95 C for 2 minutes and 25 cycles of 95 C for 1 minute, 52 C for 1 minute, 72 C for 1 minute, and a nal extension step at 72 C for 10 minutes in the presence of 1.5 mmol/L MgCl2. PCR products were resolved by electrophoresis on 1.5% (w/v) agarose gels containing ethidium bromide (0.5 mg/mL). The amplied DNA bands were analyzed densitometrically after digital imaging capture (Image Quant 100; GE Healthcare, Sunnyvale, CA) using ImageJ 1.32j software (National Institute of Health, http://rsb. info.nih.gov/ij/; Bethesda, MD). The density of the bands corresponding to TNF-a and IL-1b messenger RNA in each sample was normalized to the quantity of the housekeeping gene GAPDH and expressed as fold change over unstimulated control. linked immunosorbent assay reader at 450 nm and normalized with an abundance of standard solution. Each densitometric value was expressed as mean standard deviation and was obtained from three independent experiments.

Statistical Analysis The data collected for each case (clinical features and the bacteria isolated) were typed onto a spreadsheet and statistically analyzed using SPSS for Windows (SPSS, Inc., Chicago, IL). The Pearson chi-square test or the one-sided Fisher exact test, as appropriate, was chosen to test the null hypothesis that there was no relationship between bacteria species such as endodontic clinical signs/symptoms and the presence of a specic group of bacteria in the root canal samples. The Pearson coefcient was used to correlate the amount of LPS, IL-1b, and TNF-a levels with the size of the radiolucent area and the number of gram-negative bacteria present in root canals with apical periodontitis. The correlation between the presence of clinical/radiographic ndings with the median levels of LPS, IL-1b, and TNF-a was analyzed using the Student t test or the Mann-Whitney U test; p < 0.05 was considered statistically signicant.

Results
Bacterial Detection (16 rDNA) Bacterial DNA was detected in all root canal samples (21/21). The maximum of ve species was detected in the root canal samples. At least 1 gram-negative species was detected in 18 of 21 root canals (Table 2). Prevotella nigrescens (13/21), Porphyromonas endodontalis (6/ 21), and Treponema socranskii (6/21) were the three most frequently target gram-negative bacterial species detected. A combination of two or more gram-negative target species was detected in 8 of 21 root canals (Table 2). The Parvimonas micra positive samples (6/21) were associated in 100% with at least one gram-negative target bacterial species. Positive associations were found between P. endodontalis and Treponema denticola (p = 0.003, Odds Ratio = 2.000, Condence Bound = 0.899-4.452) and P. micra and Filifactor alocis (p = 0.008, OR = 1.667, CB = 0.815-3.409) in primary root canal infection. Teeth with sinus tract (2/21) were related to the presence of F. alocis (p = 0.040, OR = 18.000, CB = 0.585-553.586). A radiolucent area $2 mm was associated with the presence of T. denticola (p = 0.012, OR = 10.000, CB = 2.685-37.239). A correlation between the number of different gram-negative bacterial species and the levels of IL-1b (p < 0.05, Pearson r = 0.124) (Fig. 1A) and TNF-a (p < 0.05, Pearson r = 0.173) (Fig. 1B) was found. Determination of Endotoxin Concentration (Turbidimetric Test and LAL Assay) The LAL assay (turbidimetric test) indicated that endotoxin was present in 100% of the root canals samples (21/21). The median value of endotoxin contents was 7,490 pg/mL in root canals with periradicular lesions. A higher median value of endotoxin contents was detected in teeth with the presence of a radiolucent area $2 mm (9,190 pg/mL; range, 257-212,000 pg/mL) than in teeth with a radiolucent area <2 mm (3,480 pg/mL; range, 27-289,000 pg/mL). Teeth with exudation presented higher median levels of endotoxins (9,190 pg/mL; range, 355-289,000 pg/mL) than teeth with no exudation (2,620 pg/mL; range, 27-112,000 pg/mL). The median value of endotoxins in the presence of pain on palpation was 5,580 pg/mL (range, 27-269,000 pg/ mL), and in its absence it was 35.200 pg/mL (range, 59-289,000 pg/ mL). Moreover, in the presence of tenderness to percussion, the median value was 3,480 pg/mL (range, 27-289,000 pg/mL), and in the absence of tenderness to percussion, it was 9190 pg/mL (range, 59-232,000 pg/mL). Table 3 shows the median concentration of
JOE Volume 36, Number 9, September 2010

Measurements of Total Protein Levels Released to the Culture Media The total amount of protein released in the culture media following root canal contents stimulation was measured by the Coomassie (Bradford) Protein Assay kit (Rockford, IL). As a parameter for calculation of the amount of protein released to the culture media, a standard curve was plotted using bovine serum albumin standard supplied in the kit with a known concentration (2.0 mg/mL) with a series bovine serum albumin concentration (ie, 0, 25, 125, 250, 500, 750, 1,000, 1,500, and 2,000 mg/mL). The protein assay was performed following the manufacturers instructions. Calculation of Protein Concentration The protein standard and sample solutions were measured individually using an enzyme-linked immunosorbent assay plate reader (Ultramark) at 595 nm. Because this absorbance value was directly proportional to the concentration of protein, the protein concentration from the samples solutions was determined from the standard curve. Measurements of IL-1b and TNF-a Levels The amounts of IL-1b and TNF-a released to the culture media following root canal contents stimulation of macrophages were measured by enzyme-linked immunosorbent assay (Duoset kit; R&D, Minneapolis, MN). A medium of unstimulated macrophage culture was used as a negative control. Briey, standard, control, or sample solution was added to the enzyme-linked immunosorbent assay well plate, which had been precoated with specic monoclonal capture antibody. After shaking gently for 3 hours at room temperature, polyclonal antiTNF-a and IL-1b antibody, conjugated with horseradish peroxidase, was added to the solution, respectively, and incubated for 1 hour at room temperature. A substrate solution containing hydrogen peroxidase and chromogen was added and allowed to react for 20 minutes. The levels of cytokines were assessed by a microenzyme1470
Martinho et al.

Clinical Research
endotoxin according to the clinical ndings and the size of the radiolucent area. A correlation was found between endotoxin contents and the levels of TNF-a released in the culture media (p < 0.05, Pearson r = 0.740) (Fig. 1C).
S21
+ + +

S20

S19

Cell Culture and Cytokine Expression IL-1b and TNF-a Messenger RNA Expression. The macrophage cell viability after 24 hours of root canal contents stimulation was conrmed in the present study tested by its capacity to express IL-1b and TNF-a messenger RNA. Measurements of IL-1b and TNF-a Levels. IL-1b and TNFa were detected in all culture media after stimulation with root canal contents. The median level of IL-1b (24.835 pg/mL) was present in almost 90-fold higher than TNF-a (0.2830 pg/mL). A higher median level of IL-1b was detected when one of the following clinical symptoms/radiographic ndings was present: pain on palpation (25.528 pg/mL), tenderness to percussion (25.528 pg/mL), or size of the radiolucent area $2 mm (25.291 pg/mL) was present (Table 3). A higher level of TNF-a was found in teeth with exudation (0.2870 pg/mL) than in its absence (0.2340 pg/mL). A correlation between the levels of IL-1b released on the culture media and the size of radiolucent area was found in this study (p < 0.05, Pearson r = 0.028) (Fig. 1D). The median concentration of endotoxin, IL-1b, and TNF-a according to clinical ndings and the size of radiolucent area is shown in Table 3.

S18

S17

S16

S15

S14

+ + + + + + + +

S13

S12

+ + + +

S11

TABLE 2. PCR Detection of Bacteria Species in 21 Root Canals with Primary Root Canal Infection and Apical Periodontitis

+ +

Discussion
Data obtained in the present study revealed that a wide variety of gram-negative bacterial species do play a role in primary root canal infection with apical periodontitis, detecting at least 1 of 11 gramnegative target bacterial species in 18 of 21 root canals investigated, with a predominance of P. nigrescens, P. endodontalis, and T. socranskii. The high frequency of P. nigrescens in primary endodontic infection with apical periodontitis seems to be related to its LPS potential in causing bone resorption (24). The almost exclusive presence of P. endodontalis in endodontic infections suggests a specic association with bone resorption and activation of macrophages cells (5, 25). Moreover, P. endodontalis LPS has been detected in a very high percentage in severe endodontic infection (26). The combination of two or more gram-negative bacterial species found in 8 of 21 root canals investigated indicates that different bacterial LPS with different toxicity structure (lipid A) (27) can be involved in the root canal infection, enhancing or even inhibiting each others antigenicity activity over periradicular tissues. For instance, P. endodontalis seems to enhance the Fusobacterium nucleatum LPS toxicity (5), whereas Porphyromonas (27) and Bacteroides fragilis (28) are limited in the presence of Escherichia coli LPS. The association of Parvimonas micra (gram-positive bacteria) with at least 1 gram-negative target bacteria (eg, F. alocis) found in the present study turns endodontic contents even more complex and immunogenic to the immune system. Peptidoglycan (PGN), which is present in a signicant amount in the gram-positive bacterial cells, plays a synergistic effect on LPS antigenic activity when they activate TLR-2 and -4, respectively (29). In this case, more macrophages differentiate into osteoclast-like cells through the Receptor Activator for Nuclear Factor k B Ligand (RANKL):Osteoprotegerin (OPG) ratio increase (29). Yoshioka et al (30) reported that LPS can bind to P. micra cells conferring to this gram-positive bacteria the capacity to induce a strong TNFa response in macrophage-like cells. Such critical nding stresses the importance of considering gram-positive non-LPS components 1471
+ + + + + +

S7

S8

S9

S10

S6

S5

+ + + + + + + +

+ + + + + + + + + +
S, samples; +, positive detection.

S4

S3

S2

+ + +

S1

ATCC

JOE Volume 36, Number 9, September 2010

Universal (16 rDNA) Dialister pneumosintes Prevotella intermedia Prevotella nigrescens Aggregatibacter actinomycetecomitans Porphyromonas gingivalis Filifactor alocis Tannerella forsythia Prevotella tannerae Treponema denticola Porphyromonas endodontalis Treponema socranskii Parvimonas micra

Target bacteria

Antigenic Activity of Bacteria from Primary Root Canal Infection with Periapical Lesions

Clinical Research

Figure 1. The correlation between the number of gram-negative species, endotoxins, and the size of the radiolucent area (mm) and the levels of cytokines. (A) The correlation between the number of gram-negative bacteria (n) and the levels of IL-1b detected in the macrophage supernatant (pg/mL). (B) The correlation between the number of gram-negative bacteria (n) and the levels of TNF-a detected in the macrophage supernatant (pg/mL). (C) The correlation between the endotoxin concentration (pg/mL) and the levels of TNF-a detected in the macrophage supernatant (pg/mL). (D) The correlation between the size of the radiographic area (mm) and the levels of IL-1b in the macrophage supernatant (pg/mL).

when interpreting ndings on the expression of proinammatory cytokines. Previous in vitro investigations have attempted to extract LPS in order to determine the antigenic activity of the endodontic contents (57). However, the clinical signicance of these investigations is unclear considering the complexity of the antigens involved in endodontic infection described earlier. Moreover, in vitro models, using bacterial growth culture media, fail in reproducing the infection environment, particularly regarding the hemin concentration in the infection site (31). Hemin concentration (in the hemoglobin form) varies considerably depending on the inammatory response and blood vessels integrity, and it might modulate the lipid A structure in the LPS molecule, which is responsible for the majority of IL-1 induction (27). To address these points, the present study stimulated macrophage cells with material individually isolated from 21 infected root canals presenting primary endodontic infection and apical periodontitis, concomitantly investigating the microbiota and the endotoxin content of each infected root canal. The analysis revealed that the increase in the number of gram-negative bacteria was signicantly followed by an increase in the IL-1b and TNF-a levels. These ndings suggest that the presence of clinical signs/symptoms such as bone destruction involved in apical periodontitis evoked by the immune system in response to LPS is not only associated with the amount of endotoxin elucidated by previous clinical investigations (12, 1415) but also with the presence of a different number and heterogeneity of gramnegative bacteria, which acting synergistically can lead to a stronger immune response in periapical tissues. 1472

P. nigrescens contains a very potent LPS molecule for prostaglandin E2 stimulation (24) in inamed pulp tissue (32) and in acutely inamed periapical tissue (17). Additionally, IL-1b and TNF-a released from macrophages supernatants treated with infectious material derived from teeth with exudation are potent stimuli for prostaglandin E2 release (33). Our results agree with Ataoglu et al (22) who reported no association between teeth with exudates and higher levels of IL-1b and are inconsistent with Kuo et al (23). Furthermore, higher signicantly levels of endotoxin were found in root canal exudation in this current study corroborating with previous clinical investigations (6, 12, 16). Even though higher levels of endotoxins were expected particularly in pain on palpation and tenderness to percussion, it was not found in the present study. These particular data might be explained by the characteristics of the samples investigated in which a greater number of cases in the absence of these clinical features were found, contributing to such ndings. Teeth with a larger size of periradicular lesions ($2 mm) were related to the detection of T. denticola in the root canal positively associated with P. endodontalis exhibiting a potent biological activity in cell culture (5) and with chronic bone resorption (29). Higher levels of endotoxin were found in those teeth, which is in agreement with Schein and Schilder (16) who reported that the endotoxin contents of teeth with radiolucent area were ve times greater than in teeth without such area. The larger size of periradicular lesions ($2 mm) was also correlated with higher levels of IL-1b in accordance to Tani-Ishii et al (34).

Martinho et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
This nding might be related to the PGN present in gram-positive and -negative bacteria, which inhibits the differentiation of monocytes/ macrophages into mature cells. Besides this inhibition, PGN promotes cytokine production by undifferentiated precursors (29), consequently increasing localized osteolysis particularly at sites with larger number of macrophages (35). More studies that analyze the mechanisms involved in the production of cytokines and bone resorption are important for the development of new therapies. The interference in toll-like receptors and interleukin-1 receptor downstream signaling pathways should be considered to control the development and progression of endodontic lesions. In the future, the use of the RNA interference can reduce the differentiation of macrophages into osteoclasts; the induction of cytosolic phospholipase-A (cPLA) and beta-defensin2; and, consequently, the release of TNF-a, IL-1b, and IL-6 (3638). Another strategy is the superexpression of specic regulator proteins such as SOCS3, interferon-gamma, and interleukin-10 through local delivery of DNA vectors (39). Overall, the present study suggested that the antigenicity of the endodontic contents against macrophages (IL-a and IL-1b release) is not only associated with the amount of endotoxin but also with the number of gram-negative bacteria involved in the infection. Further investigations should be performed in order to assess the effect of antimicrobial agents over different lipid A structures isolated from species commonly found in endodontic infections and their mechanism of action in different cell types.
Size of radiolucent area <2 mm n = 10
15.007* 0.2870 24.835 0.2870 24.835 0.2870 25.291 0.2340 25.291* 0.2575

$2 mm n = 11 Absent n=9 Present n = 12 Absent n = 13

TABLE 3. The Median Concentration of Endotoxin, IL-1, and TNF-a in pg/mL According to the Clinical Findings and Size of the Radiolucent Area

Exudation

Tenderness to percussion

9,190

9,190

2,620

9,190

3,480

Acknowledgments
25.528 0.2340

Present n=8

3,480

The authors thank Ana Regina de Oliveira Polay, Fernanda Barrichello Tosello, Thais Mageste Duque, Geovania Caldas Almeida, and Juliana Melo da Silva for technical support. We are also thankful to Cambrex for the Kinetic-QCL equipment.

References
1. Fabricius L, Dahlen G, Holm SE, et al. Inuence of combinations of oral bacteria on periapical tissues of monkeys. Scand J Dent Res 1982;90:2006. 2. Nair PNR. Pathogenesis of apical periodontitis and the causes of endodontic failures. Crit Rev Oral Biol Med 2004;15:34881. 3. Lim GC, Torabinejad M, Kettering J, et al. Interleukin 1 b in symptomatic and asymptomatic human perirradicular lesions. J Endod 1994;20:2257. 4. Rietschel ET, Brade H. Bacterial endotoxins. Sci Am 1992;267:5461. 5. Hong CY, Lin SK, Kok SH, et al. The role of lipopolysaccharide in infectious bone resorption of periapical lesion. J Oral Pathol Med 2004;33:1629. 6. Horiba N, Maekawa Y, Abe Y, et al. Cytotoxicity against various cell lines of lipopolysaccharides puried from bacteroides, fusobacterium, and veillonella isolated from infected root canals. J Endod 1989;15:5304. 7. Hosoya S, Matsushima K. Stimulation of interleukin-1 beta production of human dental pulp cells by Porphyromonas endodontalis lipopolysaccharide. J Endod 1997;23:3942. 8. Matsuo T, Ebisu S, Shimabukuro Y, et al. Quantitative analysis of immunocompetent cells in human periapical lesions: correlations with clinical ndings of the involved teeth. J Endod 1992;18:497500. 9. Artese L, Piattelli A, Quaranta M, et al. Immunoreactivity for interleukin 1-beta and tumor necrosis factor-alpha and ultrastructural features of monocytes/macrophages in periapical granulomas. J Endod 1991;17:4837. 10. Stashenko P, Dewhirst FE, Rooney ML, et al. Interleukin-1 beta is a potent inhibitor of bone formation in vitro. J Bone Miner Res 1987;2:55965. 11. Beutler B, Cerami A. Cachectin and tumour necrosis factor as two sides of the same biological coin. Nature 1986;320:5848. 12. Jacinto RC, Gomes BPFA, Shah HN, et al. Quantication of endotoxins in necrotic root canals from symptomatic and asymptomatic teeth. J Med Microbiol 2005;54:77783. 13. Vianna ME, Hortz HP, Conrads G, et al. Effect of root canal procedures on endotoxins and endodontic pathogens. Oral Microbiol Immunol 2007;22:18. 14. Martinho FC, Gomes BP. Quantication of endotoxins and cultivable bacteria in root canal infection before and after chemomechanical preparation with 2.5% sodium hypochlorite. J Endod 2008;34:26872.

Absent n = 12

Pain on palpation

35,200 5,580 7,490

Total Amount

Present n=9

n = 21

24.835 0.2830

25.528 0.2605
*Statistically signicant difference (p < 0.05).

Endotoxin (pg/ mL) IL-1 (pg/mL) TNF-a (pg/mL)

24.835 0.3150

JOE Volume 36, Number 9, September 2010

Antigenic Activity of Bacteria from Primary Root Canal Infection with Periapical Lesions

1473

Clinical Research
15. Gomes BP, Martinho FC, Vianna ME. Comparison of 2.5% sodium hypochlorite and 2% chlorhexidine gel on oral bacterial lipopolysaccharide reduction from primarily infected root canals. J Endod 2009;35:13503. 16. Schein B, Schilder H. Endotoxin content in endodontically involved teeth. J Endod 1975;1:1921. 17. McNicholas S, Torabinejad M, Blankenship J, et al. The concentration of prostaglandin E2 in human periradicular lesions. J Endod 1991;17:97100. 18. Shimauchi H, Takayama S, Miki Y, et al. The change of periapical exudate prostaglandin E2 levels during root canal treatment. J Endod 1997;23:7558. 19. Barkhordar RA, Hussain MZ, Hayashi C. Detection of interleukin-1 beta in human periapical lesions. Oral Surg Oral Med Oral Pathol 1992;73:3346. 20. Safavi KE, Rossomando EF. Tumor necrosis factor identied in periapical tissue exudates of teeth with apical periodontitis. J Endod 1991;17:124. 21. Matsuo T, Ebisu S, Nakanishi T, et al. Interleukin-1 alpha and interleukin-1 beta periapical exudates of infected root canals: correlations with the clinical ndings of the involved teeth. J Endod 1994;20:4325. 22. Ataoglu T, Ungor M, Serpek B, et al. Interleukin-1beta and tumour necrosis factoralpha levels in periapical exudates. Int Endod J 2002;35:1815. 23. Kuo ML, Lamster IB, Hasselgren G. Host mediators in endodontic exudates. I. Indicators of inammation and humoral immunity. J Endod 1998;24:598603. 24. Chung YH, Chang EJ, Kim SJ, et al. Lipopolysaccharide from Prevotella nigrescens stimulates osteoclastogenesis in cocultures of bone marrow mononuclear cells and primary osteoblasts. J Periodontal Res 2006;41:28896. 25. van Winkelhoff AJ, van Steenbergen TJ, de Graaff J. Porphyromonas (Bacteroides) endodontalis: its role in endodontal infections. J Endod 1992;18:4314. 26. Hanazawa S, Sagiya T, Kitami H, et al. Monoclonal antibody against Porphyromonas (Bacteroides) endodontalis lipopolysaccharide and application of the antibody for direct identication of the species. J Clin Microbiol 1991;29:25503. 27. Dixon DR, Darveau RP. Lipopolysaccharide heterogeneity: innate host responses to bacterial modication of lipid a structure. J Dent Res 2005;84:58495. 28. Magnuson DK, Weintraub A, Pohlman TH, et al. Human endothelial cell adhesiveness for neutrophils, induced by Escherichia coli lipopolysaccharide in vitro, is inhibited by Bacteroides fragilis lipopolysaccharide. J Immunol 1989;143:302530. 29. Jiang J, Zuo J, Hurst IR, et al. The synergistic effect of peptidoglycan and lipopolysaccaride on osteoclast formation. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:73843. 30. Yoshioka M, Grenier D, Mayrand D. Binding of Actinobacillus actinomycetemcomitans lipopolysaccharides to Peptostreptococcus micros stimulates tumor necrosis factor alpha production by macrophage-like cells. Oral Microbiol Immunol 2005;20:11821. 31. Al-Qutub MN, Braham PH, Karimi-Naser LM, et al. Hemin-dependent modulation of the lipid A structure of Porphyromonas gingivalis lypopolysacaccharide. Infect Immun 2006;74:447485. 32. Cohen JS, Reader A, Fertel R, et al. A radioimmunoassay determination of the concentrations of prostaglandins E2 and F2 alpha in painful and asymptomatic human dental pulps. J Endod 1985;11:3305. 33. Richards D, Rutherford RB. The effects of interleukin 1 on collagenolytic activity and prostaglandin-E secretion by human periodontal-ligament and gingival broblast. Arch Oral Biol 1988;33:23743. 34. Tani-Ishii N, Wang CY, Stashenko P. Immunolocalization of bone-resorptive cytokines in rat pulp and periapical lesions following surgical pulp exposure. Oral Microbiol Immunol 1995;10:2139. 35. Gowen M, Wood DD, Ihrie EJ, et al. An interleukin 1 like factor stimulates bone resorption in vitro. Nature 1983;306:37880. 36. Fahid FS, Jiang J, Zhu Q, et al. Application of small interfering RNA for inhibition of lipopolysaccharide-induced osteoclast formation and cytokine stimulation. J Endod 2008;34:5639. 37. Fukushima A, Kajiya H, Izumi T, et al. Pro-inammatory cytokines induce suppressor of cytokine signaling-3 in human periodontal ligament cells. J Endod 2010;36:10048. 38. Kim YS, Min KS, Lee HD, et al. Effect of cytosolic phospholipase A(2) on proinammatory cytokine-induced bone resorptive genes including receptor activator of nuclear factor kappa B ligand in human dental pulp cells. J Endod 2010;36:63641. 39. Kim YS, Min KS, Lee SI, et al. Effect of proinammatory cytokines on the expression and regulation of human beta-defensin 2 in human dental pulp cells. J Endod 2010; 36:649.

1474

Martinho et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Molecular Fingerprinting Reveals the Presence of Unique Communities Associated with Paired Samples of Root Canals and Acute Apical Abscesses
Francisco Montagner, DDS, MSc, PhD,* Brenda P.F.A. Gomes, DDS, MSc, PhD,* and Purnima S. Kumar, DDS, PhD
Abstract
Introduction: Acute primary endodontic infections are polymicrobial infections that affect both the root canal (RC) system and apical tissues. It is known that these communities cannot be detected by conventional culturing methods. The aim of this study was to examine the prole of microbial communities in necrotic RCs and acute apical abscesses (AAAs) using an open-ended molecular approach to compare the diversity and composition of the microbiota of these two communities. Methods: Paired samples of RC and PA exudates were collected from 20 subjects and analyzed by terminal restriction fragment length polymorphism (tRFLP). The number of peaks, the peak areas, and the community diversity were compared between RCs and PAs. The similarity of the microbial prole of each pair of RCs and PAs was assessed by computing the number of shared peaks and the Bray-Curtis Similarity Index. Results: A total of 103 and 75.5 unique fragments (tRFs) were detected in RC and PA samples, respectively. RCs and PAs were not different in the number of species or in the community diversity; however, very few species were shared between RC and PA samples. No single t-RF fragment was detected in all samples, and the majority was detected in only one sample. Low diversity of species was observed in the RCs of smokers. Subjects with previous pain showed fewer species and greater community diversity. Conclusion: The microbial proles of the RC and PA communities are distinct and diverged between all subjects, suggesting that acute endodontic infections are microbiologically heterogeneous. (J Endod 2010;36:14751479)

Key Words
Acute endodontic infections, acute apical abscess, community, terminal restriction fragment length polymorphism

cute apical abscesses (AAAs) are one of the most frequently treated conditions in endodontic emergency procedures (1). Clinically, they are characterized by spontaneous pain, tenderness to percussion, and pain on palpation (2). The presence of soft-tissue swelling indicates diffusion through bone and may result in lifethreatening conditions if immediate treatment is not provided (3). Evidence indicates that the acute primary endodontic diseases are polymicrobial infections, predominantly involving strictly anaerobic, gram-negative cocci, and gramnegative rods, along with facultative streptococci (4, 5). At least 40% of the species remain as-yet-cultivated (6). Several methods have been used to characterize microbial communities associated with endodontic infections. In the recent years, the advent of detection methods based on DNA ngerprints within the 16S ribosomal RNA (rRNA) gene has revealed an unexpected diversity of microbial species (7). Data collected from cultivationbased and culture-independent molecular studies have shown that 486 unique bacteria, one archaeal, and 9 fungal taxa are associated with pulpal and periradicular infections (8). The diversity of acute endodontic infections has been assessed by direct DNA-DNA hybridization (9) and polymerase chain reaction (PCR)-based techniques (eg, reversecapture checkerboard hybridization [10], denaturing gradient gel electrophoresis [11], terminal restriction fragment length polymorphism [12], and 16S cloning and sequencing [1315]). Terminal restriction fragment length polymorphism (t-RFLP) is an open-ended molecular approach that provides information on the structure and prole of microbial communities (16). This technique is based on PCR amplication of a unique region of the genome with uorescent-labeled broad-range primers followed by a restriction enzyme digestion that generates different fragment lengths, each fragment representing a unique species (17). It has been used as a rapid, high-throughput tool to analyze complex microbial communities with several as-yet-uncultivable members because changes in microbial community structure can be monitored by the presence, absence, and quantity of specic fragments or peaks (16, 18, 19). Even though many studies have investigated the microbial community in root canal infection (4) and in acute periapical abscess (5), none of them has studied their similarity in the same patient. Therefore, the aim of this study was to examine the prole of microbial communities in necrotic root canals (RCs) and acute apical abscesses (AAAs) from the same patient using t-RFLP of the 16S rRNA gene to compare the diversity and composition of the microbiota of these two communities.

From the *Endodontic Division, Piracicaba Dental School, State University of Campinas, Piracicaba, Sao Paulo, Brazil; and Division of Periodontology, College of Dentistry, The Ohio State University, Columbus, OH, USA Supported by a seed grant from the College of Dentistry, The Ohio State University to Dr Purnima Kumar and the Brazilian Agencies FAPESP (2007/58518-4, 2008/ 56425-1, 2008/06162-4; 2008/57551-0) and CNPq (305437/2006-2, 470820/2006-3, 471631/2008-6; 302575/2009-0). Address requests for reprints to Dr Brenda P.F.A. Gomes, Piracicaba Dental School, University of Campinas, Endodontic Division, Av Limeira 901, Bairro Areiao, Piracicaba, Sao Paulo, Brasil CEP 13414-903. E-mail address: bpgomes@fop.unicamp.br. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.004

JOE Volume 36, Number 9, September 2010

Unique Communities Associated with Paired Samples of Root Canals and Acute Apical Abscesses

1475

Clinical Research
Materials and Methods
Patient Selection and Clinical Features The present study was approved by the Ethics Committee in Research of Piracicaba Dental School (State University of Campinas, Pi racicaba, Sao Paulo, Brazil), and informed consent was obtained from all subjects. Twenty patients presenting to the emergency clinic with pulp necrosis, determined by the vitality tests; acute apical abscesses (2); and soft-tissue swelling participated in the study. Subjects with systemic disease and those who had used antibiotics in the last 3 months were excluded from the study. The impossibility of tooth isolation, the presence of dental caries or coronal destruction that allowed the communication between the root canal and the oral cavity, previous endodontic manipulation, poor access to the apical region for sampling (as determined by previous radiographic examination), or the presence of marginal periodontitis were also exclusion factors. Demographic and clinical information was collected. Sample Collection From RCs All clinical procedures were performed under local anesthesia with 2% lidocaine and epinephrine (1:100,000) as described previously by Gomes et al (20). Briey, all dental caries and restorations were removed without pulp exposure. The tooth was individually isolated from the oral cavity with a previously disinfected rubber dam. Disinfection of the rubber dam and tooth was performed using 30% hydrogen peroxide and then 2.5% sodium hypochlorite. The solution was inactivated with 5% sodium thiosulfate in order to avoid interference with the sampling procedure. The pulp chambers were exposed using sterile diamond burs under manual irrigation with sterile saline. Root canals were classied as wet or dry based on the presence of exudate. Samples were collected by inserting sterile paper points to the radiographic working length. All samples were immediately transferred to 1 mL of Viability Medium Goteborg Agar III transport medium. In multirooted teeth, the largest canal or the one with periapical radiolucency was chosen. Sample Collection From AAAs AAA samples were collected following sampling from RCs. After disinfection with 2% chlorhexidine gel, samples were taken intraorally through intact mucosa by aspiration with a sterile and disposable 5-mL plastic syringe with a 21-G needle before surgical drainage as described by Rocas et al (5) and Riggio et al (15). Exudates were classied as bloody or purulent. One hundred microliters of the exudate was immediately placed into a test tube containing 900 mL of Viability Medium Goteborg Agar III. All samples were frozen at 20 C until analysis. DNA Isolation Three hundred fty microliters of the transport medium was used to isolate DNA with a Qiagen DNA MiniAmp Kit (Qiagen, Valencia, CA) using the tissue protocol according to the manufacturers instruction. The DNA was eluted in 100 mL of TrisHCl + EDTA. t-RFLP Analysis t-RFPL analysis was performed as described previously by Fullmer et al (19). Bacterial 16S rRNA genes were amplied by polymerase chain reaction (PCR) with uorescent-labeled broad-range bacterial primers A18 (5-TTTGATCCTGGCTCAG-HEX-3) and 317 (5-FAM-AAGGAGGTGATCCAGGC-3) (Applied Biosystems, Foster City, CA). PCR was performed by adding 4 mL of community DNA to a reaction mixture (50-mL nal volume), containing 20 nmol of each primer, 40 nmol of deoxynucleotide triphosphates, and 1U of Taq polymerase. The cycling conditions used were 22 cycles of denaturation at 94 C for 1
1476
Montagner et al.

Figure 1. The number of species in 20 pairs of RC and AAA samples. Pairs of samples are connected by matching lines.

minute, annealing at 42 C for 2 minutes, and elongation at 72 C for 3 minutes followed by a nal 10-minute elongation at 72 C. The amplicons were puried with the use of a Qiaquick Kit (Qiagen). A 10-mL aliquot of puried PCR product was digested with 20 U of either HhaI or MspI enzymes in a total volume of 20 mL at 37 C for 3 hours. Ten microliters of the restriction digestion product was puried by AMPure beads (Agencourt Bioscience Corporation, Beverly, MA) according to the manufacturers protocol. Five microliters of the puried product was denatured with 10 mL of deionized formamide and mixed with 0.2 mL of GeneScan 1200 LIZ size standard (Applied Biosystems, Foster City, CA). Fragment lengths were determined on an Applied Biosystems 3730 DNA Analyzer in GeneScan mode. The number of peaks and the height and area of each peak were determined with GeneMapper 4.0 Software (Applied Biosystems, Foster City, CA). All reactions were replicated.

Figure 2. The diversity index in RC and AAA samples. Pairs of samples are connected by matching lines.

JOE Volume 36, Number 9, September 2010

Clinical Research

Figure 3. The distribution of number of species and diversity index in RC and AAA samples, considering smoking status, previous pain, and RC or AAA exudates during sampling. A and B represent RC samples, whereas C and D are AAA samples (*p < 0.05).

Data Analysis Fragments with a peak height less than 50 uorescence units were excluded from analysis. Peak areas were standardized by converting the raw values to a proportion of the total area, as previously described (21). Peaks representing less than 1% of the total area were assigned a value of zero, and the percentages of the remaining peaks were recalculated. Peak areas and the number of peaks generated by each enzyme were averaged for the two uorophores. The mean number of peaks and the peak areas were used to compute the Shannon-Wiener Diversity Index. The similarity of the microbial prole of each pair of RCs and AAAs was assessed by computing the number of shared peaks and the Bray-Curtis Similarity Index. EstimateS 8.2.0 (University of Connecticut, Storrs, CT) was used to perform the analysis. The Bray-Curtis Similarity Index is a quantitative comparative measure that uses the number of shared species and the levels of each species in both communities. This index is weighted for levels of rare and abundant species, which a simple percentage similarity will not achieve. Statistical analysis was performed with JMP (SAS Institute Inc., Cary, NC). The Wilcoxon matched pair signed rank test was used to compare total peaks and community diversity between pairs of RC and PA samples. The Wilcoxon matched pair signed rank test was used to compare shared species and the Bray-Curtis index to compare samples from RCs and PAs.

Results
The mean age of the subjects was 30.5 years (range, 19-54 years). Seventy-ve percent of the subjects were female, and 25% were current smokers. All subjects reported a history of spontaneous pain; however, no other systemic signs (eg, pyrexia, malaise, and chills) were

observed. Nine of 20 subjects reported the presence of pain in the involved tooth followed by spontaneous remission of the symptoms before the dental visit and were categorized as previous pain. RCs of 9 incisors, 2 canines, 4 premolars, and 5 molars were sampled. Teeth were caries free (3/20), had caries (7/20), or had restorations (10/ 20). Gingivitis was associated with 9 of 20 teeth. Seventy percent of the teeth had mobility because of the presence of AAA. Radiographic examination showed widening of the apical periodontal ligament space (6/20) or apical radiolucency (14/20). During sampling, 13 teeth presented wet root canals, and 7 were dry; 11 of these showed purulent exudate, and two teeth had bloody exudate. Nine teeth had bloody exudate in the AAA. The number of unique terminal-restriction fragments (t-RFs), representing the number of individual species, differed between the two uorescent labels. One hundred eleven FAM-labeled fragments and 95 HEX-labeled fragments were generated in RC samples, whereas 84 and 67 fragments were generated in AAA samples. These differences were not statistically different (Wilcoxon matched pair signed rank test, p > 0.05). No unique t-RF was ubiquitously found in all samples. One HEX-labeled fragment of 274 bp was detected in 17 of 20 RC samples and 18 of 20 AAA samples. The majority of the t-RFs was present in only one sample (RC = 49.7%, AAA = 34.4%). The number of unique species in each pair of RCs and AAAs is shown in Figure 1. Paired samples are connected by lines. The mean number of t-RFs was 13.55 (6.49 standard deviation) in RC and 12.68 (7.79 standard deviation) in AAA samples. There was no statistically signicant difference between the samples (Wilcoxon matched pair signed rank test, p = 0.7017). The diversity of the bacterial communities in RCs and AAAs is shown in Figure 2. Paired samples are connected by lines. The mean 1477

JOE Volume 36, Number 9, September 2010

Unique Communities Associated with Paired Samples of Root Canals and Acute Apical Abscesses

Clinical Research
ties. It has been used to examine the microbial diversity in root canals with symptomatic/asymptomatic primary infections (6, 12, 22). This technique is based on the size and uorescence of the terminal fragment that is generated after digestion with a restriction enzyme (16). Each unique species is represented by a different fragment length. The same species might have different fragment proles depending on the primer set and also on the restriction enzyme that digested the PCR product (23). Schutte et al (18) described that both in silico or in vitro analysis can predict the fragment length associated with a specic microbial species. These data were not presented in the study because individual species proles have not been determined yet for the A18 and 317 amplicons when digested by MspI or HhaI. Furthermore, the aim of this study was to depict the prole of microbial communities in acute endodontic infections in paired samples of RCs and AAAs from the same patient, presenting their diversity, without attempting to the description of the individual species composing the consortia. RCs harbored an average of 103 t-RFs. This is in agreement with previous reports using t-RFLP analysis to examine root canals of symptomatic teeth (6, 12). However, the number of t-RFs from AAAs in the present study was signicantly greater than that previously reported using 16S rRNA sequencing (15, 24). In the present study, several tRFs were detected once or twice in each sample, indicating that these species are of low abundance. Therefore, it is possible that these species would have been below the detection threshold of 16S cloning and sequencing. The distribution of species appears to be unique for RC and AAA samples in each subject because no single t-RF fragment was detected in all samples, and the majority was detected in only one sample. This is consistent with previous studies on acute endodontic infections that revealed different bacterial proles among samples (11-14), with the great majority of the taxa detected in only one sample. The data suggested that different microbial communities may form the etiologic basis of a single disease (12, 25). Further studies must be undertaken to determine the levels of individual species in RC and periradicular infections and their relative importance in the balance of each community. It has been shown that different species may have similar fragment lengths, which could underestimate the number of closely related species (18). The use of two uorescent primers and more than one restriction enzyme can reduce these limitations (25). In the present study, the PCR product was digested with HhaI and MspI enzymes, and two uorophores were used for each enzyme (19). The data were averaged over the two uorophores in an effort to limit the underestimation of community diversity. More fragments were generated by digestion with MspI when compared with HhaI. Because averaging data from HhaI and MspI resulted in an underestimation of the number of species and the community diversity, only results from MspI digestion were used in statistical analysis. The mean number of species and the diversity index were not different between the RC and AAA samples; however, very few species were shared between RC and AAA samples. This suggests that these two communities are compositionally different. One reason might be the limitations in the current sampling methods, especially for RC. The sampling procedures for RCs are restricted to bacteria in the planktonic state suspended in the main RC or attached to the dentin walls. Alves et al (26) suggested a protocol in which a DNA was extracted from root powder samples. However, this method can only be applied to extracted teeth. It is known that low-abundant members inside dentin tubules or RC irregularities may contribute to the community balance (27). However, they might be found at higher counts at the apical tissues, favoring their detection by molecular methods. Selective

Figure 4. The mean number of shared species and the Bray-Curtis Community Similarity Index in paired samples (*p < 0.05).

diversity index was 2.647 (0.506) in the RC sample and 2.650 (0.675) in the AAA sample. There was no statistically signicant difference between the two environments (Wilcoxon matched pair signed rank test, p = 0.9841). Figure 3 shows the distribution of number of species and the Shannon-Wiener Diversity Index in RCs and AAAs by clinical presentation. RC samples are shown in Figure 3A and B, whereas acute apical abscess samples are shown in Figure 3C and D. A greater number of species was observed in root canal samples from patients who reported previous pain (p = 0.0156). However, these subjects showed low species diversity (p = 0.0190). The distribution of species shared by each pair of RCs and AAAs and the Bray-Curtis Community Similarity Index between each pair of RC and AAA communities is shown in Figure 4. Current smokers showed more similar community proles in RC and AAA samples than never smokers (p = 0.048).

Discussion
Although it is known that RCs and AAAs are polymicrobial infections with several uncultivated organisms, there is little evidence on the microbial prole of RCs and AAA infections in paired samples. Using an open-ended molecular approach allowed comparisons to be made of these largely uncultivated microbiomes. t-RFLP is an open-ended molecular technique that enumerates cultivable, difcult-to-grow and as-yet-uncultivated bacterial taxa, allowing comparisons to be made of complex polymicrobial communi1478

Montagner et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
pressures related to nutrition may also inuence the microbial composition of both sites (28). Furthermore, although bacteria colonizing root canals are protected, bacteria invading the periradicular tissues are under direct inuence of host defense mechanisms (6, 8, 10). Our results indicate that the presence of different microbial communities in paired RC and AAA infections that may participate in the same disease. Previous studies observed different microbial communities in samples from different geographic locations (5) and in samples from resected root ends and apical chronic periodontitis (29). It has also been shown that there is a variation in the species distribution in the coronal and apical thirds of the root canal in the same patient (26). However, none of them have investigated the effect of local factors over microbial species distribution in endodontics. Smoking is considered an important risk factor for endodontic disease. Krall et al (30) showed in a longitudinal analysis that smokers were 1.7 times more likely to have RC treatment than nonsmokers. In the present study, low species diversity was observed in RC of smokers. The oral environment and supragingival dental plaque are the main sources of bacteria in root canal infections (31). It has been shown that the microbiological composition of the subgingival microora in smokers is different from that of nonsmokers (32), with higher levels of Parvimonas micra, Fusobacterium nucleatum, Campylobacter rectus, Porphyromonas gingivalis, Actinobacillus actinomycetemcomitans and Bacteroides forsythus (32, 33). It is possible that smoking exerts a similar effect by enriching pathogens in the RC environment. Smoking can decrease important leukocyte functions such as chemotaxis and phagocytosis (34) that could favor pulp and periradicular tissue colonization. Further large-scale studies examining the microbial ora in RCs of smokers and nonsmokers are warranted. Subjects with previous pain showed fewer species and greater community diversity. Because the Shannon-Wiener Diversity Index is a measure of both number of species as well as the abundance of each species, the data suggest that subjects with previous pain had greater numbers of certain species. This is in agreement with previous report by Gomes et al (35), indicating that asymptomatic RC infections in patients with previous pain harbored a more complex microbiota, with facultative and strict anaerobes. In summary, the microbial proles of the RC and AAA communities are distinct, except in smokers. Both communities exhibit the same number of species and similar diversity; however, subjects with previous pain showed greater diversity in the RC system but not the AAA. The microbial proles of the two communities diverged between all subjects, suggesting that acute endodontic infections are microbiologically heterogeneous.
7. Siqueira JF Jr, Rocas IN. Exploiting molecular methods to explore endodontic infections: part 1current molecular technologies for microbiological diagnosis. J Endod 2005;31:41123. 8. Siqueira JF Jr, Rocas IN. Diversity of endodontic microbiota revisited. J Dent Res 2009;88:96981. 9. Siqueira JF Jr, Rocas IN, Souto R, et al. Microbiological evaluation of acute periradicular abscesses by DNA-DNA hybridization. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;92:4517. 10. Siqueira JF Jr, Rocas IN. The microbiota of acute apical abscesses. J Dent Res 2009; 88:615. 11. Machado de Oliveira JC, Siqueira JF Jr, Rocas IN, et al. Bacterial community proles of endodontic abscesses from Brazil and USA subjects as compared by denaturing gradient gel electrophoresis analysis. Oral Microbiol Immunol 2007;22:148. 12. Saito D, Marsh TL, de Souza Cannavan F, et al. Assessment of intraradicular bacterial composition by terminal restriction fragment length polymorphism analysis. Oral Microbiol Immunol 2009;24:36976. 13. Munson MA, Pitt-Ford T, Chong B, et al. Molecular and cultural analysis of the microora associated with endodontic infections. J Dent Res 2002;81:7616. 14. Jacinto RC, Gomes BP, Desai M, et al. Bacterial examintaion of endodontic infections by clonal analysis in concert with denaturing high-performance liquid chromatography. Oral Microbiol Immunol 2007;22:40310. 15. Riggio MP, Aga H, Murray CA, et al. Identication of bacteria associated with spreading odontogenic infections by 16S rRNA gene sequencing. Oral Sur Oral Med Oral Pathol Oral Radiol Endod 2007;103:6107. 16. Marsh TL. Terminal restriction fragment length polymorphism (t-RFLP): an emerging method for characterizing diversity among homologous populations of amplication products. Curr Opin Microbiol 1999;2:3237. 17. Osborn AM, Moore ERB, Timmis KN. An evaluation of terminal-restriction fragment length polymorphism (t-RFLP) analysis for the study of microbial community structure and dynamics. Environ Microbiol 2000;2:3950. 18. Schutte UME, Abdo Z, Bent SJ, et al. Advances in the use of terminal restriction frag ment length polymorphism (t-RFLP) analysis of 16S rRNA genes to characterize microbial communities. Appl Microbiol Biotechnol 2008;80:36580. 19. Fullmer SC, Preshaw PM, Heasman PA, et al. Smoking cessation alters subgingival microbial recolonization. J Dent Res 2009;88:5248. 20. Gomes BP, Pinheiro ET, Gade-Neto CR, et al. Microbiological examination of infected dental root canals. Oral Microbiol Immunol 2004;19:716. 21. Rees GN, Baldwin SK, Watson GO, et al. Ordination and signicance testing of microbial community composition derived from terminal restriction fragment length polymorphisms: application of multivariate statistics. Antonie Van Leeuwenhoek 2004; 86:33947. 22. Hommez GMG, Verhelst R, Vaneechoutte M, et al. Terminal restriction fragment length polymorphism analysis of the microora in necrotic teeth of patients irradiated in the head and neck region. J Endod 2008;34:104851. 23. Avaniss-Aghajani E, Jones K, Holtzman A, et al. Molecular technique for rapid identication of mycobacteria. J Clin Microbiol 1996;34:98102. 24. Dymock D, Weightman AJ, Scully C, et al. Molecular analysis of microora associated with dentoalveolar abcesses. J Clin Microbiol 1996;34:53742. 25. Siqueira JF Jr, Rocas IN, Debelian GJ, et al. Proling root canal bacterial communities associated with chronic apical periodontitis from Brazilian and Norwegian subjects. J Endod 2008;34:145761. 26. Alves FR, Siqueira JF Jr, Carmo FL, et al. Bacterial community proling of cryogenically ground samples from the apical and coronal root segments of teeth with apical periodontitis. J Endod 2009;35:48692. 27. Siqueira JF Jr, Rocas IN. Community as the unit of pathogenicity: an emerging concept as to the microbial pathogenesis of apical periodontitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;107:8708. 28. Grenier D, Mayrand D. Nutritional relationships between oral bacteria. Infect Immun 1986;53:61620. 29. Subramanian K, Mickel AK. Molecular analysis of persistent periradicular lesions and root ends reveals a diverse microbial prole. J Endod 2009;35:9507. 30. Krall EA, Abreu Sosa C, Garcia C, et al. Cigarette smoking increases the risk of root canal treatment. J Dent Res 2006;85:3137. 31. Okayama H, Nagata E, Ito HO, et al. Experimental abscess formation caused by human dental plaque. Microbiol Immunol 2005;49:399405. 32. Van Winkelhoff AJ, Bosch-Tijhof CJ, Winkel EG, et al. Smoking affects the subgingival microora in periodontitis. J Periodontol 2001;72:66671. 33. Zambon JJ, Grossi SG, Machtei EE, et al. Cigarette smoking increases the risk for subgingival infection with periodontal pathogens. J Periodontol 1996;67:10504. 34. Sasagawa S, Suzuki K, Sakatani T, et al. Effect of nicotine on the functions of human polymorphonuclear leukocytes in vitro. J Leukoc Biol 1985;37:493502. 35. Gomes BP, Lilley JD, Drucker DB. Associations of endodontic symptoms and signs with particular combinations of specic bacteria. Int Endod J 1996;29:6975.

References
1. Boykin MJ, Gilbert GH, Tilashalski KR, et al. Incidence of endodontic treatment: a 48-month prospective study. J Endod 2003;29:8069. 2. Walton RE, Torabinejad M. Principles and Practice of Endodontics. 3rd ed. Philadelphia: Saunders; 2002. 3. Flynn TR, Shanti RM, Levi MH, et al. Severe odontogenic infections, part 1: prospective report. J Oral Maxillofac Surg 2006;64:1093103. 4. De Sousa EL, Ferraz CC, Gomes BP, et al. Bacteriological study of root canals associated with periapical abscesses. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:3329. 5. Rocas IN, Baumgartner JC, Xia T, et al. Prevalence of selected bacterial named species and uncultivated phylotypes in endodontic abscesses from two geographic locations. J Endod 2006;32:11358. 6. Sakamoto M, Rocas IN, Siqueira JF Jr, et al. Molecular analysis of bacteria in asymptomatic and symptomatic endodontic infections. Oral Microbiol Immunol 2006;21: 11222.

JOE Volume 36, Number 9, September 2010

Unique Communities Associated with Paired Samples of Root Canals and Acute Apical Abscesses

1479

Clinical Research

A Cone-Beam Computed Tomography Study of Maxillary First Permanent Molar Root and Canal Morphology in a Chinese Population
Qing-hua Zheng, DDS,* Yao Wang, DDS,* Xue-dong Zhou, DDS, PhD, Qian Wang, DDS,* Guang-ning Zheng, DDS, and Ding-ming Huang, DDS, PhD
Abstract
Aim: This study evaluated root and canal morphology of permanent maxillary rst molars in a Chinese population using cone-beam computed tomography scanning. Methodology: The sample included 775 cone-beam computed tomography images of maxillary rst molars; 627 of the subjects had unilateral qualifying molars and 74 had bilateral qualifying molars. The following observations were made: (1) frequency of root and canal numbers, (2) frequency of additional canals in the mesiobuccal root by sex, age, and tooth position, and (3) unilateral and bilateral occurrence of additional canals in the mesiobuccal root. Results: Fused roots were present in 2.71% of unilateral qualifying molars. Multiple canals were present in the following frequencies: two canals in 0.31%, three canals in 47.21%, four canals in 50.40%, ve canals in 1.75%, and six canals in 0.31% of teeth. Additional canals were detected in 52.24% of mesiobuccal roots, 1.12% of distobuccal roots, and 1.76% of palatal roots. Patients aged 20 to 30 years showed a higher prevalence of additional mesiobuccal root canals. This prevalence did not differ with sex and tooth position. Most (71.11%) of the additional mesiobuccal root canals in subjects with bilateral qualifying molars were symmetric. Conclusion: Conebeam computed tomography scanning is an effective method for studying external and internal dental morphology. The root and canal congurations of maxillary rst molars in this Chinese population were consistent with previously reported data. More attention should be given to the detection of additional canals in patients between 20 and 30 years of age. These data may facilitate successful endodontic treatment. (J Endod 2010;36:14801484)

Key Words
Canal morphology, cone-beam computed tomography, maxillary rst molar, root morphology

uccessful endodontic treatment depends on the adequate cleaning, shaping, and lling of the root canal system. A thorough knowledge of root canal morphology is essential to achieve this goal (1). The inability to detect, debride, and obturate all extant canals is a major cause of endodontic failure (2). Maxillary rst molars have the most complex root and canal morphology of the maxillary dentition (3), and many studies have attempted to assess their anatomic characteristics. It is now generally accepted that the most common form of maxillary rst molar has three roots and four canals (4). Most reported incidences of a second canal (MB2) in the mesiobuccal root (MBR) exceed 50% (57). These teeth also exhibit some anatomic variation. Sert and Bayirli (8) assessed 200 extracted maxillary rst molars by clearing and reported that 9.5% of the distobuccal roots (DBRs) had more than two canals. Kottoor et al (9) reported one case in which the maxillary rst molar exhibited three roots and seven canals (three in the MBR, two in the DBR, and two in the palatal root [PR]). The internal anatomy of dental roots and canals has been assessed by clearing (6,10,11), in vitro endodontic access with radiography and instruments (12), radiographic examination (4), sectioning (13), in vitro macroscopic examination (14), and in vivo root canal therapy (RCT) with magnication (15). Cone-beam computed tomography (CBCT) scanning was introduced in the eld of endodontics in 1990 (16). This noninvasive three-dimensional (3D) imaging technique has many endodontic applications, including morphologic analyses. It uses a cone-shaped beam of radiation to acquire data in a single 360 rotation, which reveals the internal architecture of an object. Compared with conventional CT imaging, CBCT allows increased accuracy, higher resolution, and scan time and dose reductions (17). Tachibana and Matsumoto (16) reviewed the many potential endodontic applications of CBCT, including the analysis of canal morphology. Baratto et al (18) used three methods (ex vivo, clinical, and CBCT) to assess the internal morphology of maxillary rst molars and concluded that CBCT was effective for the initial identication of such morphology. Trevor et al (19) found CBCT to be a reliable method for the detection of the MB2 canal when compared with the gold standard of physical sectioning of the specimen. CBCT can also provide personal data such as sex, age, and tooth position, which may have important implications in the preoperative evaluation of canal morphology for nonsurgical RCT (8, 20). The certain identication of tooth position can also prevent the recruitment of inappropriate tooth samples in studies using extracted teeth, thus ensuring the accuracy of the research. The aim of this study was to evaluate root and canal morphology of permanent maxillary rst molars in a Chinese population using CBCT. The prevalence of additional

From the *State Key Laboratory of Oral Diseases, Sichuan University, Chengdu, China; and Departments of Conservative Dentistry and Endodontics and Radiology, West China College of Stomatology, Sichuan University, Chengdu, China. Address requests for reprints to Dr. Ding-ming Huang, Department of Conservative Dentistry, 14# 3rd Section, Renmin South Road, Chengdu, Sichuan Province 610041, China. E-mail address: dingminghuang@163.com; and Dr. Guang-ning Zheng, Department of Radiology, 14# 3rd Section, Renmin South Road, Chengdu, Sichuan Province 610041, China. E-mail address: gnzheng@163.com 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.018

1480

Zheng et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
TABLE 1. Frequency Distribution of Root Morphology in Maxillary First Molars Root morphology
All roots separate MBR fused with DBR MBR fused with PR DBR fused with PR All roots fused

Maxillary rst molar (n = 627) (%)

610 (97.29) 10 (1.59) 1 (0.16) 3 (0.48) 3 (0.48)

canals in the MBR was also evaluated by sex, age, tooth position, and unilateral or bilateral occurrence.

Materials and Methods

We screened CBCT images from 2,575 subjects that had been obtained in the medical imaging center of West China Hospital of Stomatology, Chengdu, between June 2009 and March 2010. The 3D Accuitomo CBCT machine (MCT-1[EX-2F], J. Morita Manufacturing Corp, Kyoto, Japan) used for tooth identication provided a grayscale image of 14 bits and had a voxel size of 0.125 mm. All images used a 1-mm slice thickness. Subjects with fully erupted permanent maxillary rst molars were selected. Qualied maxillary molars had fully formed apices and lacked root canal llings, posts, and crown restorations. CBCT images of 775 maxillary molars from 701 patients of Chinese descent were identied in the database; 74 of these patients had bilateral qualifying molars. Bilateral molar data were only used to analyze the distribution of unilateral and bilateral occurrence of additional canals in the MBR. Axial, coronal, and sagittal two-dimensional section images were displayed on a monitor and inspected by two endodontists using One Data Viewer software (J. Morita Manufacturing Corp). An intraexaminer calibration based on the anatomic diagnosis of CBCT images had been previously performed to assess data reliability. The two endodontists, who were assisted by a radiologist with endodontic experience, examined 20 previously selected CBCT images of morphologically diverse maxillary rst molars (three, four, and ve root canals). They evaluated the images twice, with a 1-week interval between the assessments. The reliability data were analyzed with a kappa test. After intra-examiner calibration, the two endodontists separately evaluated the images in the study sample. Disagreement in the interpretation of images was rst addressed in a discussion between the two endodontists. If consensus could not be reached, a radiologist with endodontic experience helped to make the decision. The frequency of root and canal numbers; the incidence of additional canals in the MBR; and correlations with sex, age, tooth position; and bilateral and unilateral appearance were determined and assessed by the chisquare test.

shown in Table 1. Most (97.29%) molars had three separate roots, whereas all roots were fused in 0.48% of teeth. Table 2 shows the frequency of the root canal numbers in the 627 unilateral qualifying molars. The most frequent pattern in this sample was four canals (50.40%) followed by three canals (47.21%). Teeth with two, ve, and six canals comprised 2.37% of the sample. The frequencies of root canal and apical foramen numbers in different roots are shown in Table 3. Most teeth had one foramen per root (MBR = 87.98%, DBR = 99.04%, and PR = 98.88%). One tooth presented with only two canals in the DBR and PR. Another tooth exhibited six canals, with two in each root (MBR, DBR, and PR). The teeth with ve canals had two canals in the MBR and an additional canal in either the DBR or PR (Figs. 1 and 2). Because the canals in teeth with all roots fused cannot be classied as mesial, buccal, and palatal, the following analysis of canal congurations in different roots (MBR, DBR, and PR) excludes these three teeth (n = 624). Single canals were the most frequent conguration in DBRs and PRs. Additional canals were detected in 52.24% of MBRs, 1.12% of DBRs, and 1.76% of PRs. Table 4 shows the distribution of additional canals in the MBR of maxillary rst molars by sex, age, and tooth position. The incidence of additional canals in the MBR was 54.27% for men and 50.00% for women. Multiple MBR canals were found in 54.13% of left maxillary rst molars and 49.8% of right molars. No statistical differences were observed in these data. The MBR of teeth in patients between 20 and 30 years of age showed a signicantly higher prevalence of additional canals than those of the other age groups (p < 0.05). The lowest frequency (40%) of additional canals was found in the group of patients who were more than 60 years of age. There were no statistical differences among groups other than the 20- to 30-year group. The distribution of unilateral and bilateral occurrence of additional MBR canals among subjects with bilateral permanent maxillary rst molars is shown in Table 5. Of 775 subjects examined, 74 subjects showed bilateral permanent maxillary rst molars (Fig. 2). The bilateral occurrence among patients with additional root canals in the MBR was 71.11%, which was much higher than the unilateral occurrence (28.89%). The frequency of bilateral distribution did not differ with sex and tooth position (Table 6).

Discussion
Prior knowledge of root and canal anatomy facilitates the accurate detection of all root canals in a tooth during endodontic treatment. Many studies of the root and canal morphology of the maxillary rst molar have been conducted because this tooth presents complex morphology that often increases the difculty of treatment (5, 6). However, most of these studies have either completely destroyed the tooth during examination or have gained only two-dimensional anatomic information. Moreover, few studies have incorporated information such as ethnic background, age, and sex of the study population, which may have important clinical implications for treatment. The recent use of CBCT has made it possible to conduct a nondestructive 3D global analysis of the external and internal morphology of the root and canal system. The 3D Accuitomo CBCT machine used in this study offers high resolution and is well suited for endodontic applications (16).

Results
Kappa test values for intraexaminer variability were greater than 0.76. The 627 patients with unilateral qualifying maxillary rst molars were between 10 and 86 years of age, with a mean age of 30.2 years. The frequency distribution of maxillary rst molar root morphology is
TABLE 2. Frequency of Root Canal Numbers in 627 Maxillary First Molars

No. of canals
No. of teeth and frequency (%) 1 0 2 2 (0.31) 3 296 (47.21) 4 315 (50.40) 5 11 (1.75) 6 2 (0.31)

JOE Volume 36, Number 9, September 2010

Maxillary First Permanent Molar Root and Canal Morphology in a Chinese Population

1481

Clinical Research
TABLE 3. Frequency of Root Canal and Apical Foramen Numbers in Different Roots No. of canals and frequency (%) 1
MBR (n = 624) DBR (n = 624) PR (n = 624) 297 (47.59) 617 (98.88) 613 (98.24)

No. of foramina and frequency (%) 3

3 (0.48) 0 0

2
324 (51.92) 7 (1.12) 11 (1.76)

1
556 (89.10) 622 (99.68) 617 (98.88)

2
68 (10.90) 2 (0.32) 7 (1.12)

3
0 0 0

This study found that 2.71% of maxillary rst molars among patients of Chinese descent had fused roots. The MBR had a higher frequency of fusion than other roots. This result was consistent with ndings in Ugandan (21) and, to a lesser extent, Irish (22) populations. However, previous studies of Burmese (23) and Thai (5) populations found three separate roots in all maxillary rst molars. These differences highlight the inuence of ethnic background on maxillary molar root morphology. Frequencies of additional canals in the MBR have been reported by many studies (5, 8, 24). The high frequency (52.40%) of additional MBR canals in this study is largely consistent with ndings from clearing studies of Asian populations from Thailand (65.0%) and Japan (58.0%) (5, 24). Our result was lower than that of a Turkish

population (93.5%) (8). Our results contrasted with those of a recent study of a Chinese population from the Guanzhong area, which used a new modied root canal staining technique (25) and found that 33.3% of the MBRs had additional canals. This variation may be caused by the differences in sample sizes, methods, and/or regional populational diversity. The present study found that 79.20% of the additional canals in MBR exhibited a single foramen. This result agreed with that of Vertucci (26) who reported that 82% of the additional MBR canals fused into a single apical foramen. Variations in additional canals in the DBR and PR have been less frequently observed. Our study found additional canals in 1.12% of DBRs and 1.17% of PRs. These results were similar to these (2.50%

Figure 1. A case of unilateral permanent maxillary rst molar with a second canal in mesiobuccal root (MB2) as viewed in the coronal, sagittal, and axial direction by using One Data Viewer software. (A) Coronal view with the mesiobuccal root exhibiting 2 canals, (B) sagittal view, (C) axial view (the arrow indicates the maxillary rst molar with MB2), and (D) scanning details.

1482

Zheng et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Figure 2. Cases of maxillary rst molar with root and canal variation in axial section; the black arrows indicate the examined tooth. (A) Two canals and fused roots, (B) ve canals, (C) six canals, and (D) bilateral occurrence of MB2.

TABLE 4. Numbers and Frequency of Additional Root Canals in the MBR of the Maxillary First Molars by Sex, Age, and Tooth Position Sex Male n = 328
No. of teeth Frequency (%)
*p < 0.05.

Tooth position Female n = 296

148 50.00

Age (y) 10$20 n = 225

113 50.22

Left n = 351
190 4.13

Right n = 273
136 49.82

20$30 n = 123
84 68.29*

30$40 n = 84
43 51.19

40$50 n = 102
43 42.16

50$60 n = 50
22 44.00

>60 n = 40
16 40.00

178 54.27

TABLE 5. Distribution of Unilateral and Bilateral Occurrence Among 45 Patients With Additional Root Canals in the MBR Number of patients
MBR with more than one canal (unilateral, left) (unilateral, right) (bilateral) Total 5 8 32 45

Frequency (%)
11.11 17.78 71.11 100

and 1.20%) reported by al Shlabi et al (22) in a study that used the clearing method. Sert and Bayirli (8) conducted a clearing study of human permanent teeth from Turkish patients and concluded that sex was an important factor to consider in the preoperative evaluation of canal morphology for nonsurgical RCT. In contrast, the incidence of additional MBR canals in our sample did not differ by the sex of the patient. This is in concordance with the results of a previous study reported by Neaverth et al (27). We also found that the number of MBR canals was not affected by tooth position, which is in agreement with the latter study.

JOE Volume 36, Number 9, September 2010

Maxillary First Permanent Molar Root and Canal Morphology in a Chinese Population

1483

Clinical Research
TABLE 6. Distribution of Unilateral and Bilateral Occurrence of Additional Root Canals in the MBR Among 74 Patients With Bilateral Permanent Maxillary First Molars MBR with additional canal Unilateral Left n
No. of patients Male Female Total No. of teeth 1 4 5 5

MBR with one canal Bilateral

Right %
3.22 9.30 6.76 3.38

Total % n
17 20 37 74

n
2 6 8 8

%
54.84 46.51 50.00 50.00

n
11 13 24 61

%
35.48 30.23 32.43 41.22

n
31 43 74 148

6.45 13.95 10.81 5.41

Hess (28) has reported that a broad canal may diverge into two canals as a patient grows older because of the apposition of dentin between the canal walls at the narrowest points. After 40 years of age, the canal morphology appeared to again become less complicated because of calcication of the canal ramications. Our study found signicantly more additional MBR canals among patients between 20 and 30 years of age (68.29%, p < .05). This result was similar to that of Neaverth et al (27) who found a higher frequency of two MBR canals in the maxillary rst molars of patients between 20 and 40 years of age. In our study, less than half of the MBR was found to have additional canals in patients older than 40 years. Additional MBR canals were symmetrically distributed in 71.11% of our Chinese population. If a patient has a known MB2 in one maxillary rst molar, the dentist should be aware of the likelihood that an MB2 is present in the opposite molar. We found no differences in bilateral occurrence by sex or tooth position. The incidence of additional canals in subjects with bilateral maxillary rst molars (58.78%) was higher than that of subjects with unilateral maxillary rst molars (52.40%). This may partly be caused by the sample size and the high incidence of bilateral distribution of the additional canals, which augments the result.

Conclusions
Within the limitations of this study, it can be concluded that more than half of maxillary rst molars have four canals and three roots in the Chinese population. Most of the additional canals were located in the MBR and had one foramen. The incidence of additional MBR canals did not differ with sex or tooth position and was usually bilateral. More additional canals were detected by CBCT in patients between 20 and 30 years of age. These values may help dentists to locate additional canals in the maxillary rst molar and thereby achieve better outcomes for the endodontic treatment of these teeth.

References
1. Baisden MK, Kulild JC, Weller RN. Root canal conguration of the mandibular rst premolar. J Endod 1992;18:5058. 2. Vertucci FJ. Root canal morphology and its relationship to endodontic procedure. Endodontic Topics 2005;10:329. 3. Vertucci FJ, Haddix JE, Britto LR. Tooth morphology and access cavity preparation. In: Cohen S, Hargreaves KM, eds. Pathways of the Pulp. 9th ed. St. Louis: Mosby Elsevier; 2006:203. 4. Walton R, Torabinejad M. Principles and practice of endodontics. 2nd ed. Philadelphia: WB Saunders Co; 1996. 5. Alavi AM, Opasanon A, Ng YL, et al. Root and canal morphology of Thai maxillary molars. Int Endod J 2002;35:47885.

6. Pineda F, Kutler Y. Mesiodistal and buccolingual roentgenographic investigation of 7275 root canals. Oral Surg Oral Med Oral Pathol 1972;33:10110. 7. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral Med Oral Pathol 1984;58:58999. 8. Sert S, Bayirli GS. Evaluation of the root canal congurations of the mandibular and maxillary permanent teeth by gender in the Turkish population. J Endod 2004;30: 3918. 9. Kottoor J, Velmurugan N, Sudha R, et al. Maxillary rst molar with seven root canals diagnosed with cone-beam computed tomography scanning: a case report. J Endod 2010;36:91521. 10. Gulabivala K, Aung TH, Alavi A, et al. Root canal morphology of Burmese mandibular molars. Int Endod J 2001;33:35970. 11. Omer OE, Al Shalabi RM, Jennings M, et al. A comparison between clearing and radiographic techniques in the study of the root-canal anatomy of maxillary rst and second molars. Int Endod J 2004;37:2916. 12. Kulild JC, Peters DD. Incidence and conguration of canal systems in themesiobuccal root of maxillary rst and second molars. J Endod 1990;16:3117. 13. Seidberg BH, Altman M, Guttuso J, et al. Frequency of two mesiobuccal root canals in maxillary permanent rst molars. J Am Dent Assoc 1973;87:8526. 14. Pecora JD, Woelfel JB. Sousa Neto MD. Morphologic study of themaxillarymolars. 1.External anatomy. Braz Dent J 1991;2:4550. 15. Buhrley LJ, Barrows MJ, BeGole EA, et al. Effect of magnication on locating the MB2 canal in maxillary molars. J Endod 2002;28:3247. 16. Tachibana H, Matsumoto K. Applicability of x-ray computerized tomography in endodontics. Endod Dent Traumatol 1990;6:1620. 17. Scarfe WC. Imaging of maxillofacial trauma: evolutions and emerging. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;100:S7596. 18. Baratto Filho F, Zaitter S, Haragushiku GA, et al. Analysis of the internal anatomy of maxillary rst molars by using different methods. J Endod 2009;35:33742. 19. Trevor CB, Nathan G, Charles CL, et al. Efcacy of cone-beam computed tomography as a modality to accurately identify the presence of second mesiobuccal canal in maxillary rst and second molars: a pilot study. J Endod 2010;36:86770. 20. Fogel HM, Peikoff MD, Christie WH. Canal conguration in the mesiobuccal root of the maxillary rst molar: a clinical study. J Endod 1994;20:1357. 21. Rwenyonyi CM, Kutesa AM, Muwazi LM, et al. Root and canal morphology of maxillary rst and second permanent molars teeth in a Ugandan population. Int Endod J 2007;40:67983. 22. al Shalabi RM, Omer OE, Glennon J, et al. Root canal anatomy of maxillary rst and second permanent molars. Int Endod J 2000;33:40514. 23. Ng YL, Aung TH, Alavi A, et al. Root and canal morphology of Burmese maxillary molars. Int Endod J 2001;34:62030. 24. Degerness RA, Bowles WR. Dimension, anatomy and morphology of the mesiobuccal root canal system in maxillary molars. J Endod 2010;36:9859. 25. Weng XL, Yu SB, Zhao SL, et al. Root canal morphology of permanent maxillary teeth in the Han nationality in Chinese Guanzhong area: a new modied root canal staining technique. J Endod 2009;35:6516. 26. Vertucci FJ. Root canal morphology of the human permanent teeth. Oral Surg Oral Med Oral Pathol 1984;58:58999. 27. Neaverth EJ, Kolter LM, Kaltenbach RF. Clinical investigation (in vivo) of endodontically treated maxillary rst molars. J Endod 1987;13:50612. 28. Hess W. The Anatomy of the Root Canals of the Teeth of Permanent Dentition. London: John Bale, Sons & Danielsson Ltd; 1925.

1484

Zheng et al.

JOE Volume 36, Number 9, September 2010

Clinical Research

Prevalence and Activity of Epstein-Barr Virus and Human Cytomegalovirus in Symptomatic and Asymptomatic Apical Periodontitis Lesions
Katinka Hernadi, MD,* Anita Szalmas, MSc,* Richard Mogyorosi, MD, Levente Czompa, MD, Gyorgy Veress, PhD,* Eszter Csoma, PhD,* Ildiko Marton, MD, PhD, DSci, and Jozsef Konya, MD, PhD*
Abstract
Introduction: Apical periodontitis is a polymicrobial inammation with a dominant ora of opportunistic Gram-negative bacteria; however, a pathogenic role of human herpesviruses such as Epstein-Barr virus (EBV) and human cytomegalovirus (HCMV) has been implicated recently. The aims of this study were to determine the prevalence, activity, and disease association of EBV and HCMV in apical periodontitis in an Eastern Hungarian population. Methods: Forty samples with apical periodontitis (17 symptomatic and 23 asymptomatic) and 40 healthy pulp controls were collected. EBV and HCMV prevalences were measured by polymerase chain reaction (PCR) detection of the viral DNA and viral activity was tested by reverse-transcription PCR amplication of viral messenger RNA. Results: EBV DNA and EBNA-2 messenger RNA were found in apical periodontitis lesions at signicantly (p < 0.0001) higher frequencies (72.5% and 50%, respectively) than in controls (both 2.5%). The occurrence of HCMV infection was rare in both apical lesions (10%) and controls (0%). The presence of EBV DNA in apical lesions was associated signicantly with large ($5 mm) lesion size (p = 0.02) but not with symtoms (p = 0.30). Symptomatic manifestation was signicantly associated with the co-occurrence (odds ratio [OR], 8.80; 95% condence interval [CI], 1.69-45.76) but not the sole occurrences of EBNA-2 messenger RNA (OR, 2.29; 95% CI, 0.48-11.06) and large lesion size (OR, 4.02; 95% CI, 0.81-19.89). Conclusion: EBV infection is a frequent event in apical periodontitis, whereas the involvement of HCMV still remains to be elucidated. This study showed that symptomatic manifestation was likely to occur if a large-sized apical periodontitis lesion is aggravated with active EBV infection. (J Endod 2010;36:14851489)

erpesviruses are supposed to be involved in several oral diseases in immunocompetent hosts, including mucosal inammations, irreversible pulpitis, and apical and marginal periodontitis (13). Apical periodontitis is essentially a polymicrobial inammation caused by opportunistic endodontic bacteria, but other microbes, such as Epstein-Barr virus (EBV) and human cytomegalovirus (HCMV), can play a pathogenic role (1, 47). Like other herpesviruses, these microbes are capable of persisting in a lifelong latency after primary infection: EBV initially infects oronasopharyngeal epithelial cells in a lytic form, and then infects B lymphocytes establishing type I latency, which does not inuence host cell functions (810). HCMV establishes latent infection in CD34+ bone marrow myeloid progenitor cells, dendritic cell (DC) precursors, and monocytes. For both EBV and HCMV, typical sites of reactivation are the salivary glands and the mucosal surfaces (8, 11). During periapical inammation, there is an inux of leukocytes, of which the mononuclear cells may carry latent herpesviruses. The local environment may allow herpesviral activation from the inactive latent phase. In the EBV-infected B lymphocytes, either a phenotype change called type III latency or lytic reactivation can occur. Several studies tested the activity of EBV infection by detecting EBNA-2 messenger RNA, which is expressed only in type III latency. In this latency stage, tumor necrosis factor a (TNF-a) is produced by infected B cells, which is a major mediator of inammation and tissue destruction. In these studies, EBNA-2 expression was indeed associated with symptomatic periods of apical periodontitis (1, 4, 5, 12). Regarding HCMV, the differentiation of infected monocytes into tissue macrophages or the maturation of infected DCs is followed by the reactivation of latent infection (11). The released infectious virions are capable of infecting further macrophages, T lymphocytes, endothelial cells, and connective tissue cells (13, 14). The aims of this study were to determine the prevalence and activity of EBV and HCMV by the detection of viral DNA and messenger RNA in apical periodontitis samples and to analyze the association with clinical symptoms.

Material and Methods

Forty apical periodontitis samples and 40 impacted third molars used as healthy pulp controls were collected from patients seeking dental care at the Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, University of Debrecen. The Local Ethical Committee approved the study (approval number: 2885-2008). Patients were anonymously coded for identication, and they lled out informed consents. Patients with apical periodontitis also lled out questionnaires, which contained questions about general diseases, medications, history of the involved tooth, and related symptoms.

Key Words
Apical periodontitis, Epstein-Barr virus type III viral latency, herpesviruses, human cytomegalovirus, polymerase chain reaction

From the *Department of Medical Microbiology and Faculty of Dentistry, Medical and Health Science Center, University of Debrecen, Debrecen, Hungary. Address requests for reprints to Dr J Konya, PO Box 17 Debrecen 4012 Hungary. E-mail address: konya@med.unideb.hu 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.008

JOE Volume 36, Number 9, September 2010

EBV and HCMV in Apical Periodontitis Lesions

1485

Clinical Research
The inclusion criteria for patients were as follows: individuals in good health (American Society of Anesthesia I or II) with no severe systemic disease requiring surgical apicoectomy because of the failure of conventional root canal therapy. Patients with either poor general status or systemic diseases or periodontally involved teeth (probing depth >4 mm) were excluded. Teeth with apical periodontitis were classied into symptomatic and asymptomatic groups. Symptomatic lesions were characterized by acute pain, discomfort on biting, sensitivity by percussion, or palpation at the apical region of mucosa. The asymptomatic lesions did not have any clinical symptoms with the exception of periapical radiolucent area on radiographs. Based on the radiographic size of the lesion, samples were divided into two subgroups: more or equal to 5 mm ($5 mm) and less than 5 mm (<5 mm) diameter of lesions. Before administrating local anesthetics for apicoectomy, patients rinsed their mouth with 0.2% chlorhexidine mouthwash for 1 minute, and the teeth, gingiva, and mucosa of the sample area were washed with 0.2% chlorhexidine. Using a sterile no. 15 blade, gingival incisions were extended one or two teeth mesially or distally from the studied tooth followed by a vertical releasing incision. A fullthickness mucoperiosteal ap was then reected, and the periapical lesion was exposed with a sterile round barr using sterile saline as coolant. A sterile curette was used to obtain the periapical sample, which was put into sterile Eppendorf tube with RNAlater RNA stabilization reagent (Applied Biosystems, Foster City, CA) and then was immediately frozen to 70 C. All control teeth were matured impacted third molars, with no signs of caries, cracking, or pulpal inammation. Before administrating local anesthetics, patients rinsed their mouth with 0.2% chlorhexidine mouthwash for 1 minute and the teeth, gingiva, and mucosa of the sample area were washed with 0.2% chlorhexidine. Using a sterile no. 15 blade, an alveolar reach incision and gingival incisions were extended one or two teeth mesially from the involved tooth followed by a vertical releasing incision. A full-thickness mucoperiosteal ap was then reected, and the impacted third molar was exposed with a sterile round barr using sterile saline as coolant. Each removed tooth was put in a Falcon tube with sterile saline and was immediately frozen to 70 C. To avoid sample to sample contamination in the laboratory, each extracted tooth were packed separately into sterile plastic bag, and it was fractured inside the bag and thereafter pulp tissues were removed. Homogenized tissue samples were divided into two portions: one for RNA and the other for DNA isolation. DNA was isolated by High Pure Viral Nucleic Acid Kit (Roche, Switzerland) according to the manufacturers instructions. The total RNA was extracted with TRI Reagent (Sigma, St Louis, MO) according to the manufacturers protocol. The extracted RNA was turned to complementary DNA (cDNA) by using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems) with random hexamers. The effectiveness of DNA and RNA isolation were controlled with polymerase chain reaction (PCR) and reversetranscription PCR detection of human b-globin and h36B4 housekeeping genes, respectively. Five microliters of DNA or cDNA was amplied in 20 mL nal volume using 1 U Red Taq Polymerase (Sigma, St Louis, MO) according to the manufacturers instructions. Nested PCR amplication was performed with a 1-mL aliquot of the primary reaction. The primers and PCR conditions are shown in Table 1 (1517). The signicance of differences were analyzed using Yates corrected chi-square statistics. If an expected cell value was less than ve in a contingency table, the Fisher exact test was used. For multivariate analysis, logistic regression statistics were used.

Results
A total of 40 apical periodontitis samples were collected from 36 patients (age, 18-80 years; mean age, 49 years) and 40 healthy pulp

TABLE 1. Primer Sequences and PCR Conditions PCR conditions Target

EBV-DNA: BamH1-W P-f P-r N-f N-r EBV-mRNA: EBNA-2 P-f P-r N-f N-r HCMV-DNA: pp65 P-f P-r N-f N-r HCMV-mRNA: pp65 P-f P-r N-f N-r

Primer sequence

Size of amplicon (reference)

Temperature prole

cycles

5 GAGACCGAAGTGAAGGCCCT 3 5 GGTGCCTTCTTAGGAGCTGT 3 5 GCCAGAGGTAAGTGGACTTTAAT 3 5 GAGGGGACCCTGAGACGGGT 3 5 TCCACCACACCCAGGCAC 3 5 TGGAGAGGTCAGGTTACTTAC 3 5 ACACACACCCACCCGTCTCA 3 5 TTGCTGGACGAGGACCCTT 3 5 TCACCTGCATCTTGGTTGCG 3 5 TGCCGCTCAAGATGCTGAAC 3 5 GGAAACACGAACGCTGACGT 3 5 TGCCGCTCAAGATGCTGAAC 3 5 ACGCGCTGCCGCTCAAGAT 3 5 TGTAGTAGACGTCGGGCTCTTT 3 5 CGCTGCCGCTCAAGATGCTGAA 3 5 ATCTGCTTGCCCGACGCGTGAA 3

171 bp (15) 97 bp (15)

94 C 30 s 94 C 30 s 94 C 30 s 94 C 30 s 94 C 30 s 94 C 30 s 94 C 30 s 94 C 30 s

58 C 30 s 60 C 30 s 54 C 30 s 63 C 30 s 64 C 30 s 65 C 30 s 60 C 30 s 70 C 30 s

72 C 90 s 72 C 90 s 72 C 90 s 72 C 90 s 72 C 90 s 72 C 90 s 72 C 90 s 72 C 90 s

30 35

205 bp (16) 120 bp

30 35

309 bp (17) 220 bp (17)

30 35

195 bp (16) 124 bp

30 35

Where reference is not shown, the primers were designed by the authors. P-f, primary-forward primer; P-r, primary-reverse primer; N-f, nested-forward primer; N-r, nested reverse primer; mRNA, messenger RNA.

1486

Hernadi et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
TABLE 2. Prevalence (DNA) and Activity (mRNA) of EBV and HCMV in Symptomatic, Asymptomatic, Larger ($5 mm)-, and Smaller (<5 mm)-Sized Apical Periodontitis Lesions EBV n (%) DNA
Apical periodontitis (n = 40) Symptomatic (n = 17) Asymptomatic (n = 23) Lesion size $5 mm (n = 21) Lesion size <5 mm (n = 19) Healthy controls (n = 40) 29* (72.5) 14* (82) 15* (65) 19* (91) 10* (53) 1 (2.5)

HCMV n (%) mRNA

20* (50) 12* (71) 8 (35) 16* (76) 4 (21) 1 (2.5)

EBV+ HCMV n (%) DNA

3 (7.5) 1 (6) 2 (9) 2 (10) 1 (5) 0 (0)

DNA
4 (10) 1 (6) 3 (13) 3 (14) 1 (5) 0 (0)

mRNA
0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Using healthy controls as the reference group, statistical tests revealed *p < 0.0001, p < 0.001 and p < 0.05. The nonlabeled frequencies are not signicantly different from those of the healthy controls

control samples from 25 patients (age, 17-29 years; mean age, 23 years). EBV DNA was detected in apical periodontitis lesions at a signicantly higher frequency than in healthy pulp controls (72.5% vs 2.5%, p < 0.0001). The local activation of EBV infection as detected by EBNA-2 messenger RNA expression was signicantly more frequent in apical periodontitis lesions (50% vs 2.5%, p < 0.0001). Samples with apical periodontitis were classied into subgroups according to symptoms and the size of radiographic bone destruction, respectively (Table 2.). Both markers of EBV infection had signicantly higher frequencies in each subgroup compared with healthy controls. Table 2 also shows that EBV DNA and EBNA-2 messenger RNA were detected more frequently in large lesions compared with small lesions and in symtomatic lesions compared with asymptomatic ones. Both EBV DNA (p = 0.02) and EBNA-2 messenger RNA (p = 0.002) were signicantly associated with a large lesion size.

Lesion size itself can also be an important clinical determinant of symptomatic manifestation. Therefore, we assessed the disease association of EBV infectious markers together with that of lesion size in logistic regression statistics allowing both univariate and multivariate analysis (Table 3). Based on the crude odds ratios and condence intervals, large lesion size and EBNA-2-messenger RNA but not EBVDNA appeared to be signicant determinants for symptomatic apical periodontitis in the univariate analysis. However, the adjusted odds ratios and condence intervals of the multivariate analysis indicated that the strength of association has been reduced to a tendency level (p = 0.08) for lesion size, and the association between EBNA messenger RNA and symptomatic manifestation has been eliminated (Table 3). Because EBNA messenger RNA and lesion size mutually weakened each others effect, we hypothesized that the disease association could be determined rather by complementary than separated effects of the investigated parameters. Therefore, the involved parameters were

TABLE 3. Determinants of Symptomatic Periapical Periodontitis Independent variable

Lesion size <5 mm $5 mm EBV DNA No Yes EBNA mRNA No Yes Variables Combined size mRNA 15 9 16 3 (20) 3 (33) 11 (69) 1.0 (ref) 1.99 (0.31-13.06) 8.80 (1.69-45.76) p = 0.027 NA NA NA

Lesions n
19 21

Symptomatic n (%)
4 (21) 13 (62)

Odds ratio (95% condence interval)

1.0 (ref) 6.09 (1.49-24,99) p = 0.01 1.0 (ref) 2.49 (0.55-11.30) p = 0.24 1.0 (ref) 4.50 (1.17-17.37) p = 0.03

Adjusted odds ratio (95% condence interval)

1.0 (ref) 4.02 (0.81-19.89)* p = 0.08 1.0 (ref) 1.11 (0.19-6.18) p = 0.90 1.0 (ref) 2.29 (0.48-11.06) p = 0.30

11 29

3 (27) 14 (48)

20 20

5 (28) 12 (60)

<5 mm No <5 mm Yes $5 mm No $5 mm Yes

NA, not applicable. Adjusted to )EBNA messenger RNA and lesion size. 95% condence interval and the corresponding p value of the calculated OR unless p for trend calculated.

JOE Volume 36, Number 9, September 2010

EBV and HCMV in Apical Periodontitis Lesions

1487

Clinical Research
combined, and the odds ratios for symptomatic manifestation were calculated separately in subgroups with none, any, and both of the hypothesized disease determinants (Table 3). Indeed, the cooccurrence of large lesion size and EBNA-2 messenger RNA was strongly associated (odds ratio, 8.80) with symptomatic manifestation. Neither the age nor the sex of the patients, tooth localization (left-right or upperlower), or type of teeth (incissors, canines, or premolars) inuenced the lesion size, the symptoms, or the incidence of EBV DNA and EBNA-2 messenger RNA, respectively (data not shown). Only four apical periodontitis lesions (10%) carried HCMV DNA, and, although HCMV was not found in healthy pulp controls, the difference was not signicant in either the whole group of lesions (p = 0.12) or the subgroups. We could not detect HCMV messenger RNA in pathological or healthy samples (Table 2). Based on the results by Sunde et al (13), we expected a low viral load in the samples analyzed in this study. Therefore, we applied nested PCR to increase the specicity and sensitivity of the virus detection, and the cycle number in the primary PCR was limited to 30 to reduce the risk of in-procedure cross-contamination (20). With regard to the underlying mechanisms, EBV-infected B lymphocytes in type III latency stage are known to produce TNF-a, transforming growth factor b, and interleukin-10 (21). TNF-a has the ability to increase bone resorption and can induce powerful hyperalgesia (22, 23). Ttransforming growth factor b is able to impair antiviral host defenses through the repression of lymphocyte proliferation, cytotoxic T-cell functions, toll-like receptor signaling, and so on. Interleukin-10 is able to inhibit macrophage activation and the antigen-presenting functions (24). A further aim of this study was to provide data obtained on healthy tissue samples. We detected only a negligible occurrence of EBV DNA and EBNA-2 messenger RNA in healthy pulp tissues. Based on these results and the similar results of a recent study (16), the most probable source of EBV infection in apical periodontitis is the immigrating Blymphocyte population. The literature on the prevalence of HCMV shows two peaks of distribution. Several studies found the frequencies of HCMV messenger RNA to be between 40% and 100% in apical periodontitis lesions (1, 4, 5, 12, 17). Our data are more consistent with the ndings of other studies, which found HCMV occurrence between 0% and 15.9% in apical periodontitis (13, 16, 18). Because these investigations uniformly tested the pp65 matrix protein coding sequences on HCMV DNA or messenge RNA, the observed biological diversity is probably because of the recent alteration of HCMV epidemiology at certain geographic regions. Periapical areups may develop as a result of complex immune responses against the herpesviral-bacterial coinfection (5, 18). The cumulative effects of herpesviruses, endopathogenic bacteria, and proinammatory immune mechanisms may be manifested in increased resorption of the periapical alveolar bone and in clinical symptoms, such as acute pain, discomfort on biting, and sensitivity to palpation at the apical region of mucosa. In conclusion, EBV infection is a frequent event in apical periodontitis, whereas the involvement of HCMV still remains to be elucidated. If EBV infection is present, the infected cells have a remarkable chance to be at type III latency stage, which, in turn, leads to the production of inammatory cytokines. This study showed that symptomatic manifestation was likely to occur if EBV infection at type III latency aggravates a large-sized apical periodontitis lesion.

Discussion
In apical periodontitis, the active participation of EBV was shown previously by EBNA-2 messenger RNA expression and immunohistochemical detection of EBV latent membrane proteins (6, 18). Both markers indicate that a remarkable proportion of inammated periapical lesions harbors EBV infected lymphocytes in type III latency stage. In this study, not only the activity of EBV infection was analyzed on the basis of EBNA-2 messenger RNA detection but also the occurrence of EBV DNA was measured to detect the prevalence of the virus regradless of the infectious stage. Our EBNA-2 messenger RNA results revealed viral activity at a similar level to other studies (1, 4, 5, 12). This study showed that approximately two thirds of the EBV DNApositive periapical lesions had EBNA-2 messenger RNA expression, the remaining one third carried EBV in infectious stages other than type III latency. Most studies on EBV infection in apical periodontitis included the symptomatic manifestation as the only clinical parameter in the analysis, and they also revealed the association between symptoms and EBNA-2 messenger RNA expression (1, 4, 5, 12). The studies by Sabieti et al (1, 4) analyzed the lesion size, and the pairwise analysis revealed that both symptomatic manifestation and large lesion size were associated with EBNA-2 messenger RNA expression. In this study, we pointed out that symptomatic manifestation and a large lesion size tend to coexist, and, therefore, their association with EBV infection was evaluated in a multivariate statistical analysis. According to the results of the multivariate analysis, we hypothesized a complementary effect of a large lesion size and EBNA-2 expression upon symptomatic manifestation. Subgrouping of the lesions according to this hypothesis allowed us to prove that large-sized apical periodontitis lesions aggravated with EBV infection at type III latency stage will most likely be symptomatic. On the other hand, the sole occurrence of either determinant has little if any effect on symptomatic stage. Another study showed that the lytic stage as detected by the expression of a viral tegument protein also occured frequently in apical periodontitis and tended to correlate with lesion size but not with symptomatic manifestation (16). From methodologic points, we minimized the contamination of samples with high EBV containing compartments such as saliva or periodontal pocket ora. The surgical technique itself is performed in a way to avoid saliva contamination from the oral cavity just to prevent postoperative wound infection. Furthermore, saliva is deprived of live replicating cells capable of hosting EBV infection in type III latency. Periodontally involved teeth were excluded from the study to avoid a major source of external EBV contamination (19). Because the same gingival incision technique was applied during operation, contamination by gingival sulcus should have affected equally the diseased and the contol group of samples. 1488
Hernadi et al.

References
1. Sabeti M, Valles Y, Nowzari H, et al. Cytomegalovirus and Epstein-Barr virus DNA transcription in endodontic symptomatic lesions. Oral Microbiol Immunol 2003; 18:1048. 2. Saygun I, Yapar M, Ozdemir A, et al. Human cytomegalovirus and Epstein-Barr virus type 1 in periodontal abscesses. Oral Microbiol Immunol 2004;19:837. 3. Slots J, Contreras A. Herpesviruses: a unifying causative factor in periodontitis? Oral Microbiol Immunol 2000;15:27780. 4. Sabeti M, Slots J. Herpesviral-bacterial coinfection in periapical pathosis. J Endod 2004;30:6972. 5. Sabeti M, Simon JH, Slots J. Cytomegalovirus and Epstein-Barr virus are associated with symptomatic periapical pathosis. Oral Microbiol Immunol 2003;18:3278. 6. Slots J, Sabeti M, Simon JH. Herpesviruses in periapical pathosis: an etiopathogenic relationship? Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:32731. 7. Sabeti M, Simon JH, Nowzari H, et al. Cytomegalovirus and Epstein-Barr virus active infection in periapical lesions of teeth with intact crowns. J Endod 2003;29:3213. 8. Slots J, Saygun I, Sabeti M, et al. Epstein-Barr virus in oral diseases. J Periodontal Res 2006;41:23544.

JOE Volume 36, Number 9, September 2010

Clinical Research
9. Kenney S, Theodore E. Woodward Award: development of novel, EBV-targeted therapies for EBV-positive tumors. Trans Am Clin Climatol Assoc 2006;117:5573. discussion 734. 10. Bollard CM, Cooper LJ, Heslop HE. Immunotherapy targeting EBV-expressing lymphoproliferative diseases. Best Pract Res Clin Haematol 2008;21:40520. 11. Sinclair J, Sissons P. Latency and reactivation of human cytomegalovirus. J Gen Virol 2006;87:176379. 12. Slots J, Nowzari H, Sabeti M. Cytomegalovirus infection in symptomatic periapical pathosis. Int Endod J 2004;37:51924. 13. Sunde PT, Olsen I, Enersen M, et al. Human cytomegalovirus and Epstein-Barr virus in apical and marginal periodontitis: a role in pathology? J Med Virol 2008;80: 100711. 14. Chen V, Chen Y, Li H, et al. Herpesviruses in abscesses and cellulitis of endodontic origin. J Endod 2009;35:1828. 15. Kis A, Feher E, Gall T, et al. Epstein-Barr virus prevalence in oral squamous cell cancer and in potentially malignant oral disorders in an eastern Hungarian population. Eur J Oral Sci 2009;117:53640. 16. Li H, Chen V, Chen Y, et al. Herpesviruses in endodontic pathoses: association of EpsteinBarr virus with irreversible pulpitis and apical periodontitis. J Endod 2009;35:239. 17. Yazdi KA, Sabeti M, Jabalameli F, et al. Relationship between human cytomegalovirus transcription and symptomatic apical periodontitis in Iran. Oral Microbiol Immunol 2008;23:5104. 18. Saboia-Dantas CJ, Coutrin de Toledo LF, Sampaio-Filho HR, et al. Herpesviruses in asymptomatic apical periodontitis lesions: an immunohistochemical approach. Oral Microbiol Immunol 2007;22:3205. 19. Sahin S, Saygun I, Kubar A, et al. Periodontitis lesions are the main source of salivary cytomegalovirus. Oral Microbiol Immunol 2009;24:3402. 20. Olmos A, Esteban O, Bertolini E, et al. Nested rt-PCR in a single closed tube. Methods Mol Biol 2003;226:15160. 21. Rochford R, Cannon MJ, Sabbe RE, et al. Common and idiosyncratic patterns of cytokine gene expression by Epstein-Barr virus transformed human B cell lines. Viral Immunol 1997;10:18395. 22. Rittner HL, Machelska H, Stein C. Leukocytes in the regulation of pain and analgesia. J Leukoc Biol 2005;78:121522. 23. Kawashima N, Stashenko P. Expression of bone-resorptive and regulatory cytokines in murine periapical inammation. Arch Oral Biol 1999;44:5566. 24. Hahn CL, Liewehr FR. Innate immune responses of the dental pulp to caries. J Endod 2007;33:64351.

JOE Volume 36, Number 9, September 2010

EBV and HCMV in Apical Periodontitis Lesions

1489

Clinical Research

Long-term Survival of Indirect Pulp Treatment Performed in Primary and Permanent Teeth with Clinically Diagnosed Deep Carious Lesions
Rene Gruythuysen, DDS, PhD, Guus van Strijp, DDS, PhD, and Min-Kai Wu, MSD, PhD
Abstract
Introduction: This retrospective study examined clinically and radiographically the 3-year survival of teeth treated with indirect pulp treatment (IPT) performed between 2000 and 2004. Methods: Sixty-six uncooperative children (4-18 years old) with at least one tooth with clinically diagnosed deep caries were included. Radiographically, the lesion depth was greater than two thirds of the dentin thickness. Incomplete excavation was performed leaving infected carious dentin at the center of the cavity. After placement of a layer of resin-modied glass ionomer as liner, the teeth were restored. A 3-year survival analysis (Kaplan-Meier) was performed. Failure was dened as the presence of either a clinical symptom (pain, swelling, or stula) or radiologic abnormality at recall. In total, 86 of 125 (69%) treated primary molars and 34 of 45 (76%) treated permanent teeth were available for both clinical and radiographic evaluation. Results: The survival rate was 96% for primary molars (mean survival time, 146 weeks) and 93% for permanent teeth (mean survival time, 178 weeks). Conclusion: This study shows that IPT performed in primary and permanent teeth of young patients may result in a high 3-year survival rate. (J Endod 2010;36:14901493)

Key Words
Caries detector, caries management, deep caries, indirect pulp capping, indirect pulp therapy, indirect pulp treatment, infected dentin, resin-modied glass ionomer, ultraconservative caries treatment

From the Department of Cariology Endodontology Pedodontology Oral Microbiology, Academic Centre for Dentistry Amsterdam, The Netherlands. Address requests for reprints to Dr Rene Gruythuysen, Department of Cariology Endodontology Pedodontology Oral Microbiology, Academic Centre for Dentistry Amsterdam (ACTA). Gustav Mahlerlaan 3004, 1081 LA Amsterdam, The Netherlands. E-mail address: R.Gruythuysen@acta.nl. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.006

he discussion of how much carious dentin must be removed in order to arrest the carious process exists already more than 150 years (1). With reference to the outcome of the histologic studies of Reeves and Stanley (2), infected dentine should be completely removed in order to arrest the carious process. However, it has been reported that carious progression was arrested for at least 10 years when bonded, and sealed composite restorations were placed directly over frank cavitated lesions extending into the dentin (3). Therefore, complete dentine carious removal may not be a prerequisite to arrest carious progression (4, 5). A randomized clinical trial with the Hall technique, in which there is no carious excavation, showed that caries is a disease that will arrest in the right circumstances (6). Taking into account the benecial outcome in ve clinical studies (7), indirect pulp treatment (IPT) is recommended as an appropriate procedure for treating primary teeth with deep carious lesions and asymptomatic pulp inammation provided that the restoration seals the cavity properly. In summary, the traditional concept of complete carious removal has been challenged. With IPT, carious dentin near the pulp is preserved to avoid pulp exposure and is covered with a biocompatible material (8). Pulpal inammation is inevitable once the dentin is affected. The remaining dentin thickness after carious excavation is a key determinant regarding the state of the pulp (9, 10). There is poor correlation, however, between the histologic ndings and the clinical diagnosis of pulpal injury. Subjacent to deep carious lesions, the pulp presents chronic inammatory exudates, including lymphocytes, macrophages, and plasma cells (11), indicating that pulpitis has been developed even in absence of unprovoked pain. On the other hand, for many years, the importance of inammation in maintaining pulpal health has been underestimated. Inammation was considered an undesirable side effect, frequently leading to pulp necrosis. In view of recent results, the inammatory process should be re-examined to understand its potentially benecial effect on pulp regeneration (12). Despite extensive pulpal inammation because of deep caries, a conservative approach can still generate a favorable prognosis for pulpal repair. When pulp exposure has occurred during complete excavation in primary or young permanent teeth, (partial) pulpotomy is a treatment option for teeth diagnosed with reversible pulpitis (7, 13). In case of symptoms referring to irreversible pulpitis, such as unprovoked pain, vital pulp techniques are associated with poor clinical outcomes. In those cases, vital pulpectomy is usually needed to save the tooth. Because of the difculties in cleaning and lling morphologically complex root canal systems, the placement of a root canal lling does not always prevent coronal bacterial contamination (14, 15). As a result, in some cases, periapical lesions may emerge. In previous outcome studies and systematic reviews, the average success rate for root canal therapy performed in teeth with vital pulps is rather high (16 18). In Toronto studies, a success rate of 93% was recorded (17). Interestingly, many outcome studies reported both periapical index scores 1 and 2 as healed or successful (17, 19) despite score 2 representing mild periapical inammation (20). When score 2 would not be considered successful, the success rate for vital teeth would drop to 70% (21). In animal studies (2224), posttreatment apical periodontitis was present in a high percentage of teeth up to 12 months after performing vital pulpectomy. Recent studies showed IPT-treated teeth remaining symptomless and free of radiologic abnormalities for years (5, 7). The pulp maintains its healing potential and

1490

Gruythuysen et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
defence capacity against advancing carious lesions (25) and acts as the best barrier against bacterial invasion (26). In addition, performance of IPT is simpler, more patient friendly, and cheaper than a root canal treatment. This retrospective study evaluated treatment success of deep carious lesions in primary and permanent teeth. The purpose was to investigate the 3-year survival rate of functional (clinical and radiological successful) primary and permanent teeth treated with IPT. rpm) with uoride toothpaste. The Caries Detector (Kuraray Co, Tokyo, Japan) applied in accordance with the manufacturers instructions was used only to check the proper excavation of the dentin-enamel junction. Infected dentin, showing intact parts and dark red staining after application of the Caries Detector (Fig. 1B), was retained and covered by a layer of resin-modied glass ionomer (RMGI) liner (Vitrebond; 3M ESPE, St. Paul, MN). All permanent teeth were restored with adhesive lling materials (Z100 or Ketac-Molar, 3M ESPE). Nine primary teeth were restored with a preformed metal crown (Stainless Steel Crowns, 3M ESPE) cemented with glass ionomer (Ketac-Cem, 3M ESPE). All other primary teeth were restored with adhesive lling materials (Dyract, Dentsply Caulk, York, PA, and Ketac-Molar, 3M ESPE). Survival was dened as teeth without clinical or radiologic signs or symptoms. The recall interval was 3 to 6 months depending on the level of dental hygiene. At recalls, the clinical performance of the teeth was examined by percussion and palpation along with the presence of signs of inammation (pain, abscess, sinus tract, and abnormal mobility). Treatment was considered radiologically successful when (1) the contours width and structure of the periodontal margin were normal and (2) no signs of pathological tooth resorption were present. Pulp sensitivity tests could not be performed in many of the very young children, and test outcomes were not reliable in the anxious (older) children. Radiographic treatment outcomes and chart notes for each tooth were independently reviewed by two calibrated investigators. When opinions differed, discussion was followed by consensus agreement. In case of doubt, the treatment was regarded as a failure. The Cohen kappa coefcient was used to assess the inter- and intraobserver agreement of the radiographic assessments. Kaplan-Meier survival analyses were performed on the censored data of both primary and permanent teeth. Statistical analyses were performed using SPSS for Windows, version 12.0.1 (SPSS Inc, Chicago, IL).

Materials and Methods

Study records were selected from a patient pool visiting the faculty clinic of the Department of Cariology Endodontology Pedodontology of Academic Centre for Dentistry Amsterdam. Sixty-six children (40 boys and 26 girls, 4-18 years old) with at least one tooth with clinically diagnosed deep caries were referred to one operator between 2000 and 2004. IPT was performed in teeth that met the following criteria: 1. Restorable permanent and posterior primary teeth without a history of spontaneous, persistent pain or sensitivity to palpation/percussion and/or presence of other clinical signs of inammation (abscess, sinus tract, and abnormal mobility). 2. Radiographically, the absence of signs of periapical or furcation pathology and/or of pathological resorption. However, in 30% of the children, no appropriate preoperative radiograph could be made because of limited cooperation. 3. Teeth at risk of pulp exposure when complete removal of carious dentine would be performed. In total, 125 primary molars and 45 permanent teeth (9 front teeth, 8 premolars, and 28 molars) were treated with IPT. In primary molars and permanent teeth, 79.1% and 41.2%, respectively, of the restorations comprised more than one surface. Patients cooperation did not inuence the indication for IPT. All patients with deep carious lesions were treated consecutively. Pulpotomy was only performed when a vital pulp was exposed during the excavation procedure (estimated 5% of all deep carious lesions). Subsequently, the tooth was excluded from the study. In one primary tooth, pulp exposure led to the diagnosis pulp necrosis, also resulting in exclusion from the study. Referring to the age of the children, about 20% to 30% of the permanent teeth treated with IPT were immature at the start of treatment. The criteria for radiographic evaluation were (1) the presence of an appropriate posttreatment radiograph and (2) the lesion depth was more than two thirds of the dentin thickness assessed on a preoperative radiograph or the restoration depth was greater than two thirds of the dentin thickness assessed on postoperative radiograph. Eighty-six (68.8%) of 125 primary molars and 34 (75.6%) of 45 permanent teeth met these criteria and were available for both clinical and radiologic evaluation. In primary as well as permanent teeth, IPT was performed similarly. After administering local anesthesia and placing a rubber dam, the dentin-enamel junction was completely excavated. The biomass near the pulp was removed with caution. In one in vivo study, specic tactile information was given about the carious removal procedure; excavation in deep cavities (with excavator) of primary molars was stopped when the remaining dentin showed increased resistance to manual instrumentation, coming out in scales or chips (27). The treatment protocol in the present study included also the use of a prophy brush to achieve this level of excavation. In order to remove the biomass (Fig. 1A) from the infected dentin, a spoon excavator (#153/154; Ash Lustra, Dentsply, Addlestone, UK) was used and/or a prophy brush (Screw type black brushes, Crescent, Dentsply Rinn, Elgin, IL; 2000

Results
The assessment of the radiographs revealed an inter- and intraexaminer agreement of 0.72 and 0.80 (Cohen kappa), respectively. The 3year survival analyses of IPT treatment, assessed radiographically, showed a survival rate of 96% for primary molars (2 failures; mean [standard deviation] survival time, 145.6 weeks [2.4 weeks]) and 93% for permanent teeth (Fig. 2A-D) (1 failure; mean [standard deviation] survival time, 178.1 weeks [8.5 weeks]). Failures in primary teeth were caused by intraradicular bone resorption. The permanent tooth failed because of apical bone resorption. During the evaluation period, no clinical symptoms were observed in primary molars as well as permanent teeth referring to inammation of the pulp. The approximal enamel wall of one primary and one permanent molar restored with a class I RMGI restoration fractured after 16 months and 12 months, respectively. Repair of the restorations did not inuence the survival of the teeth.

Discussion
This study contributes to the debate among cariologists and endodontologists whether IPT is a valuable approach to pulp preservation in teeth with deep carious lesions (28). The conservative child-friendly IPT approach limits the discomfort for uncooperative children by reduced treatment time. In the present study, deep carious lesions were dened as lesions comprising more than two thirds of the dentin thickness. Pulp exposure could be expected if direct complete excavation were pursued. Either pre- or postoperative radiographs were used to assure the lesion depth. 1491

JOE Volume 36, Number 9, September 2010

Indirect Pulp Treatment in Teeth with Clinically Diagnosed Deep Carious Lesions

Clinical Research

Figure 1. (A) A second mandibular primary molar after excavation of the dentin-enamel junction (DEJ). The biomass is still present in the center of the cavity. (B) The same tooth after removing the biomass with a prophy brush and uoride toothpaste only. The Caries Detector was used for inspection of the DEJ. After drying the cavity, the RMGI liner was applied, and the cavity was restored with a compomer.

Because no pulpal excavation was performed, lesion depth could also be assessed by postoperative radiographs. As reported by Leksell et al (29) and Kassa et al (30), up to 50% of deep carious lesions may actually have severe pulp involvement. Therefore, it seems reasonable to assume that many teeth in the present study suffered severe pulp involvement. Ninety-three percent of IPT-treated permanent teeth and 96% of primary molars remained nevertheless symptomless and free of intraradicular/periapical radiolucency for up to 3 years. Future clinical investigations, using sophisticated diagnostic tools (19), may provide more detailed information about the outcomes of vital pulp techniques. Pulpal necrosis can develop without clinical symptoms (11, 31). A low incidence of pulpal necrosis after IPT was reported (5) for 1 (3%) of 32 IPT-treated teeth. In this study, using radiographic criteria, more than 90% of the treated teeth remained free of apical radiolucency. As reported in the present study and in other investigations (5, 7), clinical outcomes achieved by IPT, as treatment for asymptomatic pulpal inammation, were not inferior to those of pulpectomy treatment (15, 21, 19). Recently, it has been questioned (19) whether the tradi-

tional strategy for treating deep caries (complete excavation resulting in pulp exposure and vital pulpectomy) still provides the best solution. Furthermore, in some cases with deep caries, without any pretreatment symptoms, spontaneous or persistent pain can develop after complete excavation. A recent cohort study supports this view, reporting a greater incidence of adverse events in deep cavities and pulpally exposed teeth than in teeth with moderately deep or shallow cavities (odds ratio = 7.8) (9). Therefore, performing IPT may avoid or delay root canal infection and thus root canal treatments. An adequate sealing of the remaining infected dentine is of the utmost importance in the investigated treatment strategy (IPT) for deep carious lesions. It has been shown that carious dentin beneath a restoration contains a decreasing number of viable bacteria over time and dries out (5, 32), which is a parameter for lesion arrest (1, 5). In addition, a shift toward a less cariogenic microora was observed (33). In another study (34), a more microorganisms were detected in teeth submitted to partial carious removal compared with the complete carious removal group. However, after sealing the cavity, the level of bacterial colonization was similar in the two groups (35).

Figure 2. (A) The rst mandibular permanent tooth before IPT in a 6.5-year-old girl. (B) The same (immature) tooth 1 week after IPT. (C) Almost 2.5 years after IPT. (D) The rst mandibular permanent tooth completely matured 5 years after IPT (2 years after nishing the clinical evaluation).

1492

Gruythuysen et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
Underneath the restoration, a number of microorganisms may survive but not in sufcient quantity to advance the disease or they are no longer carious active. Sealing of carious dentin arrested the carious process in deep carious lesions, promoted deposition of tertiary dentin, and induced mineral gain in the radiolucent zone (36). Studies on stepwise excavation advocate removal of carious tissue after the reaction of the pulpodentin complex after the rst excavation step (36, 37). This study supports the conclusion of Maltz et al (5) that the indication for reopening the cavity to remove remaining carious dentin lacks biological support. Calcium hydroxide is traditionally the material of choice in deep carious treatment because of its alkaline biocompatible properties and the induction of pulpodentin remineralization. However, in a study investigating the pulpodentin complex response to a RMGI liner applied in deep cavities, this material was found to be biocompatible comparable to calcium hydroxide (38). This was conrmed in a 4-year follow-up study (39). Teeth were treated with an IPT technique combined with either a calcium hydroxide or RMGI liner. No signicant difference in success rate (89% and 93%, respectively) was found. Additionally and in contrast to calcium hydroxide, RMGI prevents microleakage (40). Since starting to perform IPT 10 years ago, we apply this approach in every patient (anxious or not) who presents with one or more deep cavities in vital primary or permanent teeth that meet the criteria described in this study. Our 10-year experience with incomplete carious excavation conrms the results of the present study.
11. Bjrndal L, Mjor IA. Pulp-dentin biology in restorative dentistry. Part 4: dental cariescharacteristics of lesions and pulpal reactions. Quintessence Int 2001; 32:71736. 12. Goldberg M, Farges JC, Lacerda-Pinheiro S, et al. Inammatory and immunological aspects of dental pulp repair. Pharmacol Res 2008;58:13747. 13. Barrieshi-Nusair KM, Qudeimat MA. A prospective clinical study of mineral trioxide aggregate for partial pulpotomy in cariously exposed permanent teeth. J Endod 2006;32:7315. 14. Moller AJ. Microbiological examination of root canals and periapical tissues of human teeth. Methodological studies. Odontol Tidskr 1966;74(suppl):1380. 15. Molander A, Reit C, Dahlen G, et al. Microbiological status of root-lled teeth with apical periodontitis. Int Endod J 1998;31:17. 16. Sjogren U, Hagglund B, Sundqvist G, et al. Factors affecting the long-term results of endodontic treatment. J Endod 1990;16:498504. 17. Marquis VL, Dao T, Farzaneh M, et al. Treatment outcome in endodontics: the Toronto Study. Phase III: initial treatment. J Endod 2006;32:299306. 18. Ng YL, Mann V, Rahbaran S, et al. Outcome of primary root canal treatment: systematic review of the literaturepart 2. Inuence of clinical factors. Int Endod J 2008; 41:631. 19. Wu MK, Shemesh H, Wesselink PR. Limitations of previously published systematic reviews evaluating the outcome of endodontic treatment. Int Endod J 2009;42: 65666. 20. Orstavik D, Kerekes K, Eriksen HM. Clinical performance of three endodontic sealers. Endod Dent Traumatol 1987;3:17886. 21. rstavik D, Qvist V, Stoltze K. A multivariate analysis of the outcome of endodontic treatment. Eur J Oral Sci 2004;112:22430. 22. Holland GR. Periapical innervation of the ferret canine one year after pulpectomy. J Dent Res 1992;71:4704. 23. Katebzadeh N, Hupp J, Trope M. Histological periapical repair after obturation of infected root canals in dogs. J Endod 1999;25:3648. 24. Garcia de Paula-Silva FW, Hassan B, Bezerra da Silva LA, et al. Outcome of root canal treatment in dogs determined by periapical radiography and cone-beam computed tomography scans. J Endod 2009;35:7236. 25. Sims S, Duruturk L. A ow cytometric analysis of the biodefensive response of xek deciduous tooth pulp to carious stimuli during physiological root resorption. Arch Oral Biol 2005;50:4618. 26. Ricucci D, Langeland K. Apical limit of root canal instrumentation and obturation, part 2. A histological study. Int Endod J 1998;31:394409. 27. Massara ML, Alves JB, Brandao PR. Atraumatic restorative treatment: clinical, ultrastructural and chemical analysis. Caries Res 2002;36:4306. 28. Seale NS, Glickman GN. Contemporary perspectives on vital pulp therapy: views from the endodontists and pediatric dentists. J Endod 2008;34(7 suppl):S5761. 29. Leksell E, Ridell K, Cvek M, et al. Pulp exposure after stepwise versus direct complete excavation of deep carious lesions in young posterior permanent teeth. Endod Dent Traumatol 1996;12:1926. 30. Kassa D, Day P, High A, et al. Histological comparison of pulpal inammation in primary teeth with occlusal or proximal caries. Int J Paediatr Dent 2009;19:2633. 31. Bergenholtz G, Nyman S. Endodontic complications following periodontal and prosthetic treatment of patients with advanced periodontal disease. J Periodontol 1984; 55:638. 32. Bjrndal L, Larsen T. Changes in the cultivable ora in deep carious lesions following a stepwise excavation procedure. Caries Res 2000;34:5028. 33. Paddick JS, Brailsford SR, Kidd EA, et al. Phenotypic and genotypic selection of microbiota surviving under dental restorations. Appl Environ Microbiol 2005;71: 246772. 34. Lula EC, Monteiro-Neto V, Alves CM, et al. Microbiological analysis after complete or partial removal of carious dentin in primary teeth: arandomized clinical trial. Caries Res 2009;43:3548. 35. Alves LS, Fontanella V, Damo AC, et al. Qualitative and quantitative radiographic assessment of sealed carious dentin: a 10-year prospective study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:13541. 36. Magnusson BO, Sundell SO. Stepwise excavation of deep carious lesions in primary molars. J Int Assoc Dent Child 1977;8:3640. 37. Bjrndal L, Larsen T, Thylstrup A. A clinical and microbiological study of deep carious lesions during stepwise excavation using long treatment intervals. Caries Res 1997;31:4117. 38. Costa CA, Giro EM, do Nascimento AB, et al. Short-term evaluation of the pulpodentin complex response to a resin-modied glass-ionomer cement and a bonding agent applied in deep cavities. Dent Mater 2003;19:73946. 39. Marchi JJ, de Araujo FB, Froner AM, et al. Indirect pulp capping in the primary dentition: a 4 year follow-up study. J Clin Pediatr Dent 2006;31:6871. 40. Pereira CL, Cenci MS, Demarco FF. Sealing ability of MTA, Super EBA, Vitremer and amalgam as root-end lling materials. Braz Oral Res 2004;18:31721.

Conclusions
The results of this retrospective study on the survival rate of primary and permanent teeth after performing IPT are promising. An estimated 96% of primary molars and 93% of permanent teeth survived 3 years without showing adverse clinical symptoms or pathological signs on traditional radiographs. Although studies with high-quality research design are necessary to conrm the previously mentioned results, the traditional strategy of complete carious removal to manage deep carious lesions in primary and permanent teeth with asymptomatic pulpitis may be questioned.

References
1. Kidd EA. How clean must a cavity be before restoration? Caries Res 2004;38: 30513. 2. Reeves R, Stanley HR. The relationship of bacterial penetration and pulpal pathosis in carious teeth. Oral Surg Oral Med Oral Pathol 1966;22:5965. 3. Mertz-Fairhurst EJ, Curtis JW Jr, Ergle JW, et al. Ultraconservative and cariostatic sealed restorations: results at year 10. J Am Dent Assoc 1998;129:5566. 4. Ricketts DN, Kidd EA, Innes N, et al. Complete or ultraconservative removal of decayed tissue in unlled teeth. Cochrane Database Syst Rev 2006;3:CD003808. 5. Maltz M, Oliveira EF, Fontanella V, et al. Deep caries lesions after incomplete dentine caries removal: 40-month follow-up study. Caries Res 2007;41:4936. 6. Innes NP, Evans DJ, Stirrups DR. The Hall technique; a randomized controlled clinical trial of a novel method of managing carious primary molars in general dental practice: acceptability of the technique and outcomes at 23 months. BMC Oral Health 2007;7:18. 7. Fuks AB. Vital pulp therapy with new materials for primary teeth: new directions and treatment perspectives. J Endod 2008;34:S1824. 8. American Academy on Pediatric Dentistry Clinical Affairs Committee-Pulp Therapy subcommittee; American Academy on Pediatric Dentistry Council on Clinical Affairs. Guideline on pulp therapy for primary andyoung permanent teeth. Pediatr Dent 2008-2009;30(7 suppl):170174. 9. Whitworth JM, Myers PM, Smith J, et al. Endodontic complications after plastic restorations in general practice. Int Endod J 2005;38:40916. 10. Murray PE, Smith AJ, Garcia-Godoy F, et al. Comparison of operative procedure variables on pulpal viability in an ex vivo model. Int Endod J 2008;41:389400.

JOE Volume 36, Number 9, September 2010

Indirect Pulp Treatment in Teeth with Clinically Diagnosed Deep Carious Lesions

1493

Clinical Research

Frequency of Nonodontogenic Pain after Endodontic Therapy: A Systematic Review and Meta-Analysis
Donald R. Nixdorf, DDS, MS,* Estephan J. Moana-Filho, DDS, MS, Alan S. Law, DDS, PhD, Lisa A. McGuire, MLIS,k James S. Hodges, PhD, and Mike T. John, DDS, MPH, PhD**
Abstract
Introduction: Little is known about ill-dened pain that persists after endodontic procedures, including an estimate of the problems magnitude. We conducted a systematic review of prospective studies that reported the frequency of nonodontogenic pain in patients who had undergone endodontic procedures. Methods: Nonodontogenic pain was dened as dentoalveolar pain present for 6 months or more after endodontic treatment without evidence of dental pathology. Endodontic procedures reviewed were nonsurgical root canal treatment, retreatment, and surgical root canal treatment. Studies were searched in four databases electronically, complemented by hand searching. A summary estimate of nonodontogenic tooth pain frequency was derived using random-effects meta-analysis. Results: Of 770 articles retrieved and reviewed, 10 met inclusion criteria, and nine had data on both odontogenic and nonodontogenic causes of pain. A total of 3,343 teeth were enrolled within the included studies and 1,125 had follow-up information regarding pain status. We identied 48 teeth with nonodontogenic pain and estimated a 3.4% (95% condence interval, 1.4%-5.5%) frequency of occurrence. In nine articles containing data regarding both odontogenic and nonodontogenic causes of tooth pain, 56% (44/78) of all cases were thought to have a nonodontogenic cause. Conclusions: Nonodontogenic pain is not an uncommon outcome after root canal therapy and may represent half of all cases of persistent tooth pain. These ndings have implications for the diagnosis and treatment of painful teeth that were previously root canal treated because therapy directed at the tooth in question would not be expected to resolve nonodontogenic pain. (J Endod 2010;36:14941498)

ooth pain, meaning pain of known pulpal or periradicular etiology, is not the only reason for pain perceived in the dentoalveolar regions (1). Nonodontogenic causes comprise varying etiologies, such as referred myofascial pain (2), headache (3), neuropathic disorders (4), and pain stemming from various pathological conditions (5). Quantifying the frequency of nonodontogenic pain after root canal therapy is important for dentists and patients, so patients can make educated decisions by knowing the risks and benets associated with treatment. Determining the extent of this problem is the rst step toward the long-term goal of reducing diagnostic errors that often lead to irreversible dental procedures in an attempt to alleviate the pain, such as root canal retreatments, surgical root canal treatments, and tooth extractions (6). Several studies have investigated the component diagnoses, listed earlier, that comprise this group of nonodontogenic pain cases referred to tertiary care centers (5, 7, 8). Even though such pain is thought to be rare (9), the magnitude of this problem is not known to a degree that would allow for development of appropriate public health policy. Important subtypes of this pain are not quantied either, especially those pains thought to be neuropathic in nature. For patients and dentists alike, they represent a considerable challenge because they are known to respond less than favorably to treatment (10). Given the current situation (ie, multiple diagnoses comprising this group of nonodontogenic pain that have widely differing treatment needs), it is important to quantify this problem to inform clinicians so they can use this information in their daily practice. Therefore, we sought to estimate the frequency of nonodontogenic dentoalveolar pain present at 6 months or greater after root canal therapy by performing a meta-analysis, which is a robust method of synthesizing published information (11).

Materials and Methods

Inclusion Criteria Eligible for inclusion in this review were endodontic procedure articles published in any language before June 5, 2009, that reported on postoperative tooth pain after at least a 6-month follow-up. Qualifying endodontic procedures included initial root canal treatment or retreatment, surgical or nonsurgical, but not pulpotomy, partial pulpectomy, or pulp capping. The unit of observation considered was a human permanent tooth in vivo; primary teeth were excluded. The study outcome was the presence of dentoalveolar pain that explicitly did not have an odontogenic etiology, such as a cracked tooth, missed canal, or periapical pathosis. Pain could be spontaneous or provoked by biting, palpation, or percussion.

Key Words
Dentoalveolar, pain, root canal therapy, systematic review, tooth

From the *Division of TMD & Orofacial Pain, School of Dentistry, University of Minnesota, Minneapolis, MN; Department of Neurology, Medical School, University of Minnesota, Minneapolis, MN; Center for Neurosensory Disorders, School of Dentistry, University of North Carolina, Chapel Hill, NC, USA; Private Practice, The Dental Specialists, Lake Elmo, MN; kBio-Medical Library, University of Minnesota, Minneapolis, MN; Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN; and **Division of Epidemiology & Community Health, School of Public Health, University of Minnesota, Minneapolis, MN. Supported by National Institutes of Health grants: K12-RR023247, U01-DE016746 and U01-DE016747 (DR Nixdorf). Address requests for reprints to Dr Donald Nixdorf, University of Minnesota, 6-320 Moos Tower, 515 Delaware Street SE, Minneapolis, MN 55455. E-mail address: nixdorf@umn.edu. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.020

1494

Nixdorf et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
Another inclusion criterion for studies was that they reported baseline data of the population from which the follow-up sample was drawn. This requirement, allowing the frequency of occurrence to be calculated, limited study inclusion to case series, cohort, and clinical trial studies and excluded cross-sectional and case-control studies. Articles reporting randomized trials were included as a special type of prospective cohort study; however, the treatment arms were collapsed for our analysis. Unpublished research and studies reported only in abstract form were not considered.

Results
Study Identication and Characteristics We identied 770 articles (495 by electronic searching of databases and 275 by hand searching); the oldest was published in 1921. Twenty-eight were published in a language other than English (7 French; 6 Chinese; 5 Japanese; 2 each in Italian, Russian, and Spanish; and 1 each in Croatian, Danish, German, and Greek). Screening of titles and abstracts resulted in 306 articles being excluded. After full-text review, another 464 articles were excluded, so 10 articles were included in the meta-analysis (Fig. 1). All 10 articles were published in English, and 6 of them were identied by hand searching. The 10 included studies varied in the types of endodontic treatments provided, numbers of teeth treated (6-276), duration of follow-up (1-6 years), and the percentage of teeth followed up (20%-100%) (Table 1). From 3,343 teeth enrolled in the 10 studies, 1,125 teeth were followed up for at least 6 months. Among them, 48 teeth (4.3%) in seven studies were reported to have pain without an identiable odontogenic source. In these studies, teeth were determined to have tooth-related disease when the tooth was present and not properly restored, fractured, periapical radiolucency present (15, 16), sinus tract present (17), root fracture associated with severe bone loss (18), and failure of coronal restoration (19) could be identied with the root canaltreated teeth. Variation in the quality of reporting (STROBE criteria) was observed (interquartile range = 8.5-20.5, range = 5-20.5), with a median reporting quality score of 19.8 (possible scores ranging from 0 to 22). Furthermore, 9 studies contained details on both all-cause pain and nonodontogenic pain, making it possible to calculate the proportion of such pain outcomes. Summary Estimate of Nonodontogenic Pain Frequency The meta-analytic summary estimate of nonodontogenic pain frequency was 3.4% (95% condence interval, 1.4%-5.5%, Fig. 2). Moderate heterogeneity (ie, inconsistency) (20) among study estimates was observed (I2 = 65%, p = 0.002). When each study was eliminated in turn from the analysis and the meta-analysis was run with the nine remaining studies, the summary estimates ranged from 2.1% to 4.2%. Thus, individual studies did not unduly inuence the summary estimate even though one study identied 21 of the 48 cases (44%) of nonodontogenic pain. Exploration of Study Heterogeneity In meta-regression analyses (Table 2), follow-up duration was the factor that differentiated pain frequency the most; the three studies with a follow-up of 6 to 12 months had a frequency of persistent nonodontogenic pain higher by 4.5 percentage points than in the seven studies with a follow-up greater than 12 months. The study reporting quality affected pain frequency the least; the ve above-median STROBE criteria studies had a frequency of persistent nonodontogenic pain higher by 1.5 percentage points than in the ve below-median studies. However, even when differences were substantial in magnitude, all were statistically nonsignicant because of small numbers of studies. Proportion of All-Cause Pain That Is Nonodontogenic Nine studies had data for all-cause tooth pain, as previously reported (12), as well as for nonodontogenic pain, thus allowing an estimation of the proportion of such pain outcomes in each study population. In these 9 studies, 44 nonodontogenic pain cases (56%) of the 78 all-cause pain cases were identied.
Frequency of Nonodontogenic Pain after Endodontic Therapy

Information Sources and Search Strategy We conducted an initial search in MEDLINE via the PubMed interface, covering the period from 1949 to June 5, 2009, and using the search terms specied in Figure 1. This search was then adapted for use and run in the Cochrane Library, TRIP database, and Google Scholar. We also hand searched the references of prominent articles, literature reviews, and textbook chapters (source list available upon request). Our intent was to be broad in scope to ensure the inclusion of as much relevant existing data as reasonably possible. The training and reliability assessment of article selection, data abstraction of study variables, and the assessment of reported study quality have been previously reported (12). Statistical Methods We used random-effects meta-analysis (13) to determine a summary estimate of nonodontogenic pain frequency. In a sensitivity analysis, we examined whether the deletion of a single study substantially changed the meta-analysis summary estimate. To explore factors inuencing the estimate, we performed a meta-regression to investigate differences between studies with the following characteristics: (1) surgical versus nonsurgical treatment, (2) a follow-up rate of recall less than 50% versus greater than or equal to 50%, (3) follow-up at 6 to 12 months versus more than 12 months, (4) initial treatment versus retreatment, and (5) above-median quality reporting score versus below-median score according to the STROBE criteria (12, 14). We also estimated the proportion of nonodontogenic tooth pain among the subset of cases that had information on both all-cause and nonodontogenic pain. All analyses were performed using the STATA software package (Stata Statistical Software: Release 10.1; StataCorp LP, College Station, TX: StataCorp LP) and the user-written metan and metareg commands.

Figure 1. A owchart of the systematic review process.

JOE Volume 36, Number 9, September 2010

1495

Clinical Research
STROBE Rating
5 10.7 8 20.5 20.5 20 19.5 20.5 20.5 8.5 1-3.5 1-1 1-1 4-6 4-6 1-1 4-6 4-6 4-6 6-6 0 0 7 79 71 0 64 71 94 0 19.8 386 1-6

Multiple Procedures

Follow-up (y)

Nonodontogenic Tooth Pain

Figure 2. Random effects meta-analysis of the frequency of nonodontogenic tooth pain in 10 studies.

0 0 1 0 3 21 7 6 6 4

48

Discussion
This systematic review identied 10 prospective studies (3,343 enrolled teeth) and estimated the frequency of nonodontogenic pain at 6 months or more after root canal treatment to be 3.4% (95% condence interval, 1.4%-5.5%). At this rate, with more than 16.4 million root canal treatments performed annually in the United States (21), each year over half a million endodontic patients would be at risk for nonodontogenic pain. Nonodontogenic dentolalveolar pain is often difcult to diagnose (5, 8) because it is poorly understood (22). Even dening and categorizing such persistent pain is challenging, but conceptually nonodontogenic pain in the dentoalveolar region can arise from four potential processes: (1) referred musculoskeletal pain disorder, (2) neuropathic pain disorder, (3) headache disorders presenting in the dentoalveolar region, and (4) a pathological process outside the immediate dentoalveolar region that refers pain to that area, such as sinus disease, salivary gland disorders, brain tumors, angina, throat cancer, and craniofacial vascular disorders (1). In theory, our nonodontogenic pain frequency estimate is an estimate of the incidence of this condition. In practice, because the condition is challenging to diagnose, misclassication at baseline by failing to identify a nonodontogenic reason for pain results in the inclusion of such cases in the reported studies. Misclassied neuropathic pain cases at baseline would be expected to continue to be painful after endodontic treatment (9) or become more recalcitrant (23), whereas nonodontogenic cases of referred pain from distant tissues, such as musculoskeletal, pathological, and headache disorders, would likely not be adequately addressed with endodontic treatment. The amount of misclassication is not known because to our knowledge such research results have not been reported. Furthermore, because patients undergoing endodontic treatment commonly exhibit preexisting dentoalveolar pain of inammatory origin (24), this study cannot differentiate between patients whose nonodontogenic pain arose from preexisting pathosis and patients whose pain arose from the procedure. Therefore, our estimate represents a mixture of truly incident and remissive cases as well as maintenance of the condition in patients misclassied at baseline. For this reason, we call this estimate a frequency of occurrence, which quanties the burden of nonodontogenic pain, a condition with many clinical challenges.

Follow-up, n (%)

6 (100) 10 (100) 43 (86) 103 (20) 122 (28) 175 (44) 131 (25) 137 (24) 122 (26) 276 (86) 6 10 50 523 442 400 532 582 477 321 Initial NSRCT Periapical Surgery Periapical Surgery Retreatment NSRCT Initial NSRCT Combined Treatments Initial NSRCT Initial NSRCT Retreatment NSRCT Periapical Surgery

Teeth Enrolled

TABLE 1. Characteristics of the 10 studies included in the meta-analysis

3,343
NSRCT, nonsurgical root canal therapy. *Reference found by hand searching. Assumed that each patient contributed only one tooth because not explicitly stated.

Endodontic Procedure

Liu and Sidhu, 1995 Danin et al, 1999* von Arx and Kurt, 1999 Farzaneh et al, 2004a * Farzaneh et al, 2004b * Polycarpou et al, 2005 Marquis et al, 2006* de Chevigny et al, 2008a* de Chevigny et al, 2008b * Saunders et al, 2008

Authors, Year

1496

Aggregate values:

1,125 (37)

Nixdorf et al.

JOE Volume 36, Number 9, September 2010

Clinical Research
TABLE 2. The assessment of study heterogeneity by meta-regression Comparison of Subgroups
Surgical treatment vs nonsurgical treatment $50% follow-up rate vs <50% follow-up rate >12 months vs 6-12 months follow-up duration Retreatment vs initial treatment* Above-median vs belowmedian reported study quality (STROBE criteria)
*Four studies with missing data.

Coefcient (Standard Error)

2.7% (2.7) 2.8% (2.7) 4.5% (2.6) 2.3% (2.1) 1.5% (2.6)

p Value
0.36 0.34 0.13 0.35 0.57

The meta-analytic approach allows aggregation of data to produce a robust estimate (11) but has known limitations based on the quality of the studies included (25, 26). Therefore, we restricted our metaanalysis to prospective studies, which are thought to produce more accurate results in general (27) and to endodontic outcome studies in particular (12). Seven of the 10 studies identied were published in the last decade, suggesting an increased interest in reporting patient-oriented outcomes and more design rigor in recent endodontic studies. Only one study assessed nonodontogenic pain as its primary outcome. This study found a frequency of occurrence of 12% (21/ 175) (15), a number substantially higher than our meta-analysis summary, which may therefore suggest that our meta-analytic summary estimate is low. In exploratory analyses, studies with a shorter follow-up (6-12 months) had a greater frequency of persistent nonodontogenic pain than those with a longer follow-up, which is an important nding (28) and may suggest that such persistent pain improves with time. A reduced frequency of persistent postprocedural pain over time has been observed by other studies investigating nondental surgical models of human pain, such as cesarean sections (29), and has been suggested to occur with orofacial pains (23) but has not been explored in relation to endodontic procedures. A methodologic problem of our review was that the reporting unit was the tooth, whereas the outcome of persistent dentoalveolar pain is a patient-based measure. Teeth within the same individual do not represent statistically independent observations because they share the same environment, so condence intervals for our point estimates should be larger than presented. However, we believe that this is not likely a major problem because even though 6 studies reported multiple observations per patient, the difference between the number of patients and the total number of teeth was low (12%; 386/3,343). Another important issue in this review was the large proportion of patients that were not followed (67%; 2,218/3,343), which allows ample opportunity for missing cases of nonodontogenic pain. This is potentially troubling because endodontic patients have been found not to inform their endodontist when persistent pain is present (30). This is not supported by our meta-regression, which found that studies with less than 50% followup rates had higher pain frequencies than those with greater than or equal to 50% follow-up rates. This nding, although not statistically signicant, is contrary to the common view. Caution needs to be used when interpreting such results because these assessments are exploratory and do not take into account that of the 6 studies having greater than 50% follow-up rate, four had the lowest STROBE scores, and the above-median STROBE criteria studies had higher frequencies compared with the below-median half. Our meta-analysis provides some insight about the proportion of persistent pain after endodontic procedures that is nonodontogenic in

nature. Combining the present studys nding with our previous study that estimated the frequency of all-cause tooth pain to be 5.3% (12), nonodontogenic cases may account for 64% (3.4/5.3) of these teeth with pain. When comparing the proportion of patients exhibiting nonodontogenic pain among those determined to have all-cause pain in the 9 studies with available data, the proportion was 56% (44/78). When we use the best single study to assess this proportion (ie, the study that used nonodontogenic pain as its primary outcome [15]), this fraction was 57% (21/37). This suggests that at least half of all persistent tooth pain is of nonodontogenic nature, so these cases would best be managed without further endodontic therapy. This is contrary to current opinion (31-34) and practice (9) in dentistry, which advocates retreatment. Regardless of the recommended approach to treat pain after endodontic care, the large proportion of nonodontogenic pain has substantial implications for diagnoses and further treatment of these pain conditions. In conclusion, 3.4% of patients experienced persistent pain of nonodontogenic origin after root canal therapy, a number that likely represents about half of all persistent tooth pain. Therefore, the outcome of nonodontogenic tooth pain is not as rare as commonly assumed. Given that nonodontogenic pain has diverse etiologies and successful treatment is often difcult, further research is needed to diagnose nonodontogenic pain subtypes, quantify the burden on the individual experiencing it, provide adequate treatment, and assess longterm outcomes. Also needed is research that differentiates cases of nonodontogenic pain from those of a local etiology because tooth-based pathology is amenable to endodontic retreatment and nonodontogenic pain would be best treated if recognized.

Acknowledgments
The authors thank Estelle Arnaud-Battandier, David Bereiter, Dino Bilankov, Zheng Chang, Wenjung Kang, Sergey Khasabov, Thomas List, Keiichiro Okamoto, Akimasa Tashiro, and Ana Velly for translating articles.

References
1. Mattscheck D, Law AS, Nixdorf DR. Diagnosis of non-odonogentic toothache. In: Hargreaves KM, Cohen S, eds. Cohens pathways of the pulp. 10th ed. St. Louis, MO: Mosby Inc; 2011:4970. 2. Wright EF. Referred craniofacial pain patterns in patients with temporomandibular disorder. JAMA 2000;131:130715. 3. Alonso AA, Nixdorf DR. Case series of four different headache types presenting as tooth pain. J Endod 2006;32:11103. 4. Baad-Hansen L. Atypical odontalgiapathophysiology and clinical management. J Oral Rehabil 2008;35:111. 5. Israel HA, Ward JD, Horrell B, et al. Oral and maxillofacial surgery in patients with chronic orofacial pain. J Oral Maxillofac Surg 2003;61:6627. 6. Linn J, Trantor I, Teo N, et al. The differential diagnosis of toothache from other orofacial pains in clinical practice. Aust Dent J 2007;52:S1004. 7. Fricton JR. Critical commentary 1a unied concept of idiopathic pain: Clinical features. J Orofac Pain 1999;13:1859. 8. de Siqueira SRDT, Nobrega JCM, Valle LBS, et al. Idiopathic trigeminal neuralgia: clinical aspects and dental procedures. Oral Surg Oral Med Oral Path Oral Rad Endod 2004;98:3115. 9. Oshima K, Ishii T, Ogura Y, et al. Clinical investigation of patients who develop neuropathic tooth pain after endodontic procedures. J Endod 2009;35:95861. 10. Lewis MAO, Sankar V, De Laat A, et al. Management of neuropathic orofacial pain. Oral Surg Oral Med Oral Path Oral Rad and Endod 2007;103(suppl 1). S32.e1, S32.e24. 11. Borenstein M, Hedges LV, Higgins JPT, et al. Introduction to meta-analysis. Padstow, UK: John Wiley & Sons, Ltd; 2009. 12. Nixdorf DR, Moana-Filho EJ, Law AS, et al. Frequency of persistent tooth pain after root canal therapy: a systematic review and meta-analysis. J Endod 2010;36:22430. 13. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7: 17788.

JOE Volume 36, Number 9, September 2010

Frequency of Nonodontogenic Pain after Endodontic Therapy

1497

Clinical Research
14. Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Ann Intern Med 2007;147:W16394. 15. Polycarpou N, Ng YL, Canavan D, et al. Prevalence of persistent pain after endodontic treatment and factors affecting its occurrence in cases with complete radiographic healing. Int Endod J 2005;38:16978. 16. de Chevigny C, Dao TT, Basrani BR, et al. Treatment outcomes in endodontics: the Toronto studyphases 3 and 4: orthograde retreatment. J Endod 2008;34:1317. 17. Farzaneh M, Abitbol S, Lawrence HP, et al. Treatment outcome in endodonticsthe Toronto study. Phase II: initial treatment. J Endod 2004;30:3029. 18. Marquis VL, Dao T, Farzaneh M, et al. Treatment outcome in endodontics: the Toronto study. Phase III: initial treatment. J Endod 2006;32:299306. 19. Sanders A, Slade GD, Lim S, et al. Impact of oral disease on quality of life in the US and Australian populations. Community Dent Oral Epidemiol 2009;37:17181. 20. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:55760. 21. American Dental Association. Survey of dental services rendered and distribution of dentists in the United States by region and state, 1999. Chicago, IL: American Dental Association; 2002. 22. Quail G. Atypical facial paina diagnostic challenge. Aust Fam Physician 2005;34:6415. 23. Allerbring M, Haegerstam G. Chronic idiopathic orofacial pain. A long-term followup study. Acta Odontol Scand 2004;62:669. 24. Fouad A, Levin L. Pupal reactions to caries and dental procedures. In: Hargreaves KM, Cohen S, eds. Cohens Pathways of the Pulp. 10th ed. St. Louis, MO: Mosby, Inc; 2011:50428. 25. Moles DR, Needleman IG, Niederman R, et al. Introduction to cumulative metaanalysis in dentistry: lessons learned from undertaking a cumulative metaanalysis in periodontology. J Dent Res 2005;84:3459. 26. Spangberg LSW. Systematic reviews in endodonticsexamples of GIGO? Oral Surg Oral Med Oral Path Oral Rad Endod 2007;103:7234. 27. Cummings SR, Newman TB, Hulley SB. Designing a cohort study. In: Cummings SR, Newman TB, Hulley SB, eds. Designing Clinical Research. 3rd ed. Philadelphia, PA: Lippincott, Williams & Wilkins; 2007:97107. 28. Savitz DA. Integration of evidence across studies. In: Interpreting Epidemiological Evidence: Strategies for Study Design and Analysis. New York, NY: Oxford University Press; 2003:26183. 29. Nikolajsen L, Sorensen HC, Jensen TS, et al. Chronic pain following Caesarean section. Acta Anaesthesiol Scand 2004;48:1116. 30. Lobb WK, Zakariasen KL, McGrath PJ. Endodontic treatment outcomes: do patients perceive problems? J Am Dent Assoc 1996;127:597600. 31. Abbott PV. Factors associated with continuing pain in endodontics. Aust Dent J 1994; 39:15761. 32. Cohn SA. Clinical updatethe teeth and the maxillary sinus: the mutual impact of clinical procedures, disease conditions and their treatment implications. Part 1. The differential diagnosis of tooth sinus painthe dentists view. Aust Endod J 1999;25: 2931. 33. Boucher Y, Sobel M, Sauveur G. Persistent pain related to root canal lling and apical fenestration: a case report. J Endod 2000;26:2424. 34. Kim S, Kratchman S. Modern endodontic surgery concepts and practice: a review. J Endod 2006;32:60123.

1498

Nixdorf et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology

Regulation of the Stromal Cellderived Factor-1aCXCR4 Axis in Human Dental Pulp Cells
Qi-mei Gong, MS, Jing-jing Quan, MS, Hong-wei Jiang, PhD, and Jun-qi Ling, PhD
Abstract
Introduction: Although the presence of the stromal cellderived factor (SDF)-1aCXCR4 axis has been reported in dental pulp tissue, little has been known about the underlying regulation of this axis in dental pulp stem cells (DPSCs). The purpose of this study was to investigate whether inammation or hypoxia can regulate this axis in cultured human dental pulp cells (DPCs). Methods: Primary cultures of DPCs were stimulated by various concentrations of lipopolysaccharide (LPS) for 48 hours, and the production of SDF-1a or CXCR4 was assessed through the enzyme-linked immunosorbent assay and Western blotting, respectively. Additionally, DPCs were incubated in a hypoxic condition (1% O2) for 24 hours, and the cell proliferation ability was detected by methylthiazol tetrazolum assay. Quantitative reverse-transcription polymerase chain reaction (RTPCR) was used to observe messenger RNA level changes of hypoxia inducible factor-1a(HIF-a), SDF-1a, and CXCR4. The effects of hypoxia on cell migration ability were further conrmed by transmigration assay. Results: All concentrations of LPS inhibited SDF-1a production except that 1 mg/mL LPS increased the expression of CXCR4. Hypoxia promoted the proliferation of DPCs in a 24-hour culture period. Quantitative RT-PCR showed that messenger RNA levels of HIFa and CXCR4 increased, whereas SDF-1a decreased in hypoxic DPCs. Transmigration assay indicated that hypoxia increased the migration ability of DPCs. Conclusions: These results suggested that inammation and hypoxia might play an important role in regulating the SDF-1a-CXCR4 axis, which further recruits DPSCs to participate in reparative dentinogenesis. (J Endod 2010;36:14991503)

ental pulp tissue has the ability to regenerate dentin in response to tooth injuries caused by trauma or infection. This regeneration process is also known as reparative dentinogenesis, and its typical pathological feature is the formation of tertiary dentin, which results from the recruitment and proliferation of dental pulp stem cells (DPSCs) (1). After damage of the tooth, these progenitor cells will be activated to migrate to the injury sites and differentiate into odontoblast-like cells, which synthesize and excrete an extracellular matrix to form the reparative dentin. Until now, many types of signal molecules have been shown to be involved in this complex process. Stromal cellderived factor 1a (SDF-1a), a member of the chemokine family, with its receptor, CXC chemokine receptor 4 (CXCR4), forms the SDF-1a/CXCR4 axis. The primary role of this axis is to mobilize CD34+ hematopoietic stem cells to the bone marrow (2). It has also been reported to promote the migration and differentiation of tissue-committed stem cells (TCSCs) to injury sites under some pathological conditions, such as diabetes, liver cirrhosis, or myocardial infarction (35). Our previous study found that in inamed dental pulp positive staining of this axis could be detected in endothelial cells of the microblood vessels, which were surrounded by a large number of pulp cells (6). Another study also showed that CXCR4-positive dental pulp cells (DPCs) were localized in the perivascular area (7). Because CXCR4 has been reported as a new surface stem cell marker (8), these studies suggested that the SDF-1a/CXCR4 axis plays an important role in recruiting DPSCs from the stem cell niche to the injury sites. However, the detailed regulation mechanisms of this axis in DPSCs are currently unknown. This information is needed to further investigate the contribution of DPSCs to the secretion of SDF-1a and the expression of CXCR4 and analyze the factors that can inuence this axis in DPSCs. DPCs isolated from mammalian dental pulp are found to be a heterogeneous population of DPSCs and other progenitors, which possess mesenchymal stem cells like qualities similar to bone marrow mesenchymal cells (BMSCs) (9, 10). In order to further look through the regulation of the SDF-1a/CXCR4 axis in DPSCs, we established two injury models of inammation and hypoxia by using cultured human DPCs to nd whether these injuries could affect the activity of this axis in DPCs. We hypothesize that by mimicking the stress environment of dental pulp in vitro, the proliferation and migration ability of DPCs may be changed. Meanwhile, the expression levels of SDF-1a and CXCR4 may also be affected.

Key Words
CXC chemokine receptor 4, dental pulp cells, hypoxiainduced factor-1a, pulp regulation, stromal cellderived factor-1a

Materials and Methods

DPCs Culture Healthy teeth were collected from the extracted premolars or third molars for orthodontic reasons of patients (age, 13-25 years) in the Department of Oral and Maxillofacial Surgery, the Afliated Stomatology Hospital of Sun Yat-sen University, Guangzhou,

From the Department of Operative Dentistry and Endodontics, Guanghua School and Hospital of Stomatology & Institute of Stomatological Research, Sun Yat-sen University, Guangzhou, Guangdong, China. Qi-mei Gong and Jing-jing Quan contributed equally to this study. Supported by the Specialized Research Fund for the Doctoral Program of Higher Education of China (grant no. 20070558205). Address requests for reprints to Dr Jun-qi Ling, Department of Operative Dentistry and Endodontics, Guanghua School and Hospital of Stomatology, Sun Yat-sen University, Guangzhou 510055, PR China. E-mail address: lingjq@mail.sysu.edu.cn. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.011

JOE Volume 36, Number 9, September 2010

Regulation of SDF-1a-CXCR4 in HDPCs

1499

Basic ResearchBiology
China. Informed consent was obtained from each patient, and research protocols were approved by the Ethics Committee of the university. DPCs were isolated as previously reported (10). DPCs were cultured in alphamodied Eagle medium (aMEM; Gibco, Grand Island, NY) supplemented with 100 U/mL penicillin, 100 mg/mL streptomycin (Sigma, St. Louis, MO), and 10% fetal bovine serum (FBS, Gibco). Cultures were maintained at 37 C in an incubator (5% CO2/20% O2). hypoxia, 300 mL DPCs (3 105 cells/mL) were seeded in the upper chamber, and 600 mL of serum-free medium with or without 100 ng/ mL rhSDF-1a (R&D) was placed in the lower chamber. The chambers were incubated for 10 hours. Then, nonmigrating cells were scraped off from the top, and migrating cells were stained with Hoechst 33342 (Sigma, St. Louis, MO) and observed by using a uorescence microscopy. Five elds under a 400-fold magnication were randomly selected, and immunoreactive cells were counted.

Enzyme-Linked Immunosorbent Assay After being stimulated with different concentrations (0.1, 1, and 10 mg/mL) of Escherichia coli (E. coli 055:B5) lipopolysaccharide (LPS, Sigma) for 48 hours, supernatant of DPCs was collected, and the SDF1a production was measured in triplicate using an enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN) following the manufacturers instructions. Western Blotting DPCs were treated with LPS (0.1, 1, and 10 mg/mL) for 48 hours and then harvested in radioimmunoprecipitation lysis buffer (10 mmol/L Tris-HCL, 1 mmol/L EDTA, 1% sodium dodecyl sulfate, 1% Nonidet P-40, 1:100 proteinase inhibitor cocktail, 50 mmol/L b-glycerophosphate, and 50 mmol/L sodium uoride). Western blotting was performed as described (11). Proteins were detected with polyclonal antibodies against human CXCR4 (fusin (H-118):sc-9046, 1:200, Santa Cruz, CA) and monoclonal antibodies against human b-actin (1:5,000; Cell Signaling, Beverly, MA). Cell Proliferation Assay The hypoxic condition was achieved by reducing the oxygen concentration to 1% in a chamber by using a gas mixture (5% CO2 and balanced N2). DPCs were plated at a concentration of 5 103/ well in 96-well plates and incubated for 6, 12, 18, and 24 hours, respectively, in normoxia (20% O2) or hypoxia (1% O2). The cell proliferation ability was evaluated by using methylthiazol tetrazolium (MTT, Sigma) assay as described before (7). Quantitative RT-PCR After incubation for 6, 18, and 24 hours, respectively, in normoxia or hypoxia, the total RNA was extracted by using Trizol (Invitrogen, Carlsbad, CA). Quantications of HIF1-a, SDF-1a, CXCR4, and b-actin messenger RNA (mRNA) were performed based on the reported methods (6). Data of these targeted genes were normalized to the internal control b-actin to obtain DCt. The nal quantities of gene of interest relative to normoxic samples were reported by 2-DDCt method (12). The details of primers used were listed in Table 1. Transmigration Assay Transwell inserts (5-mm pore, Corning, New York) were used in this assay. After being cultured for 18 hours under normoxia or
TABLE 1. Primer Sequences Used in Quantitative RT-PCR Gene
HIF-1a SDF-1a CXCR4 b-actin

Statistical Analysis Data analysis was performed using SAS 8.1 (SAS, Raleigh, NC). A paired Student t test was used to compare two means. One-way analysis of variance was applied to compare two or more means followed by the Student-Newman-Keuls test. Differences in which p was less than 0.05 were considered to be statistically signicant.

Results
LPS Inhibited SDF-1a Expression The ELISA assay showed that nonstimulated DPCs secreted SDF-1a constitutively. After being stimulated with various concentrations of LPS (0.1, 1, and 10 mg/mL) for 48 hours, SDF-1a production of DPCs signicantly reduced (p < 0.05) in comparison with the control group (Fig. 1). LPS Increased CXCR4 Production In our preliminary study, we used the method of immunouorescence to show that CXCR4 staining was mostly found in the cytoplasm of cultured DPCs (data were not shown). In order to detect protein level changes before and after LPS stimulation, Western blotting was performed, and the analysis showed that 1 mg/mL LPS increased the expression of CXCR4 in DPCs after 48 hours of treatment (p < 0.05); the other two concentrations had no similar effect (Fig. 2). Hypoxia Promoted DPCs Proliferation After being cultured for 6, 12, 18, or 24 hours either in normoxia or hypoxia, the MTT assay was used to determine whether hypoxia had an effect on cell proliferation. MTT results showed an increased optical density value in all hypoxic groups during the 24-hour cultivation period (p < 0.05), indicating that DPCs proliferated at a faster rate in hypoxic culture condition (Fig. 3). Hypoxia Affected the mRNA Levels of HIF-1a, SDF-1a, and CXCR4 To investigate the effect of hypoxia on expression of these candidate genes, DPCs were exposed to normoxia or hypoxia for different periods (6, 18, and 24 hours). Quantitative RT-PCR was then performed to analyze HIF-1a, SDF-1a, and CXCR4 mRNA expression levels in DPCs under different culture conditions. Compared with normoxic DPCs,

Primers
Forward: 5-AAGGTATTGCACTGCACAGGC-3 Reverse: 5-CAGCACCAAGCAGGTCATAGG-3 Forward: 5-CCCGTCAGCCTGAGCTACAG-3 Reverse: 5-CGTTGGCTCTGGCAACATG-3 Forward: 5-GACCGCTTCTACCCCAATGA-3 Reverse: 5-GCCAACCATGATGTGCTGAA-3 Forward: 5-GCATGGGTCAGAAGGATTCCT-3 Reverse: 5-TCGTCCCAGTTGGTGACGAT-3

Length
99bp 64bp 63bp 106bp

1500

Gong et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology

Figure 1. ELISA analysis of SDF-1a levels in DPCs. Normal DPCs expressed SDF-1a constitutively. After being stimulated with LPS, SDF-1a production signicantly reduced. Data were shown as means standard deviation of three independent experiments. *p < 0.05 as compared with the control group.

hypoxic DPCs expressed high levels of HIF-1a and CXCR4 but low levels of SDF-1a in the 6-, 18-, and 24-hour groups, respectively (Fig. 4).

Figure 3. The effects of hypoxia on DPCs proliferation. An increased optical density value was observed at each time point during the 24-hour culture period. Data were shown as means standard deviation of three independent experiments. *p < 0.05 as compared with the normoxic group.

Hypoxia Enhanced SDF-1aDependent Cell Migration The transmigration assay was performed to determine the effect of hypoxia on DPCs migration ability. Compared with normoxic DPCs, hypoxic DPCs displayed more migratory properties. Because 100 ng/ mL rhSDF-1a has been proven to have a maximum ability to migrate DPCs in vitro (7), we added this concentration of rhSDF-1a in the lower chamber to further observe effects of hypoxia. If hypoxia is able to increase CXCR4 amount in DPCs, more cells would cross micropores of the transwell inserts and migrate to the other side. Our results found that in the presence of rhSDF-1a, more hypoxic DPCs crossed the 5-mm micropores (p < 0.05) in comparison with the normoxic group (Fig. 5).

Discussion
Dental pulp is frequently inamed because of bacterial infection of dental caries. One of the histological features in inamed dental pulp is the adhesion and migration of various leukocytes. In our previous research, we showed that SDF-1 and its receptor CXCR4 were mainly identied in inammatory cells, indicating this axis had been involved in dental pulp infection. However, the regulation of this axis in DPCs

under inammatory stimulation in vitro was less reported. In the present study, we used different concentrations of LPS to stimulate DPCs, and we found that the secretion of SDF-1a was largely suppressed, whereas 1 mg/mL LPS increased the production of CXCR4. It is surprising that the SDF-1a amount decreased because in the previous study expression of SDF-1 mRNA in inamed dental pulp was signicantly increased. Additionally, our results seem contrary to other researchers work (eg, Hosokawa et al found (13) that 10 mg/mL LPS signicantly enhanced the production of SDF-1 in human gingival broblasts). Another group also showed that heat-killed Enterococcus faecalis was able to increase SDF-1 production in human dental pulp broblasts (14). The expression of SDF-1 has been reported to be complexly regulated. It is known that LPS of gram-negative bacterial cells could induce synthesis of proinammatory cytokines, such as interleukin (IL)-1 or tumor necrosis factor a (TNF-a) (15). Fedyk et al (16) showed a novel down-regulation of SDF-1 in response to inammation of human gingival broblasts, which were probably affected by monocyte-derived IL-1a and TNF-a (16). A recent report showed that LPS from Porphyromonas gingivalis could alter the balance of cytokines, which led to a decreased SDF-1 production in human gingival and periodontal ligament broblasts (17). Taking these into account, we speculated that SDF-1a expression changes in DPCs might be caused either directly by LPS or indirectly by inammatory cytokines. On one hand, LPS may escape from the immune response

Figure 2. Western blotting results of CXCR4 production in DPCs; 1 mg/mL LPS increased the expression of CXCR4 in DPCs after 48 hours of treatment; the other two concentrations did not affect CXCR4 production. Data were shown as means standard deviation of three independent experiments. *p < 0.05 as compared with the control group.

Figure 4. The expression levels of the mRNAs encoding HIF-1a, SDF-1a, and CXCR4 in hypoxic DPCs relative to normoxic DPCs as determined by quantitative RT-PCR. The HIF-1a and CXCR4 expression level increased, but SDF-1a expression decreased in the 24-hour culture period. Data were shown as means standard deviation of three independent experiments.

JOE Volume 36, Number 9, September 2010

Regulation of SDF-1a-CXCR4 in HDPCs

1501

Basic ResearchBiology
the 24-hour culture period, indicating that hypoxia promoted the expression of HIF1-a in DPCs. As a transcriptional factor, HIF can regulate both ligands and receptors of the SDF-1a-CXCR4 axis. Ceradini et al (27) reported that the 5 anking region of the human SDF-1 gene had two potential HIF-1 binding sites. A putative HIF-1 binding site was also shown in the CXCR4 promoter by Schioppas group (28). To the best of our knowledge, this is the rst study showing the regulation of the SDFa-1-CXCR4 in hypoxic DPCs. Unexpectedly, we found that in the 24 hours the mRNA level of SDF-1a reduced, which was inconsistent with Ceradinis report. The mechanisms leading to this down-regulation are currently unknown because many transcriptional pathways have been involved in hypoxic or ischemic responses (29). Gross et al (30) found that the SDF-1 mRNA level of inner ear cells decreased after 5 hours of hypoxic culture, and they suggested that the down-regulation of SDF1 might generate signals to regulate the migration of local precursor cells. A recent report also detected a decreased pattern of SDF-1 mRNA expression in the rat hippocampus after transient global ischemia. They mentioned that a short period of hypoxia (1 day) might be sufcient to provoke neuronal death but insufcient for SDF-1 gene induction (31). Because the hypoxic condition used in our study is a transient period of 24 hours, it is possible that this time interval would not be efcient for HIF to act as a promoter to increase SDF-1a production. Nevertheless, whether SDF-1a has a protein level stabilization or other posttranslational mechanism needs to be investigated further. The increase in the level of CXCR4 mRNA suggests that HIF-1a might bind to the promoter of CXCR4 and regulate its transcription. This is consistent with the transmigration assay results of our study. When using the concentration of 100 ng/mL rhSDF-1a, hypoxic DPCs migrated efciently through the transwell inserts, indicating that hypoxia might increase the distribution of CXCR4. We also found that without the rhSDF-1a, hypoxia also improved DPCs migration ability. The SDFa-1-CXCR4 axis has been reported in DPSCs research. For example, CXCR4 has been used as a stem cell marker in side population cells sorting from porcine dental pulp (32). Furthermore, an in vivo experiment found that transplanted DPSCs could induce endogenous axon guidance via SDFa-1-CXCR4 signaling interactions (33). Because the expression of this axis has also been found in leukocytes, lymphocytes, and dendritic cells (34), it indicates that communication between different cell types in dental pulp may be particularly crucial in the homing of DPSCs. Considering the fact that the recruitment of CXCR4-positive stem/progenitor cells to a SDF-1 gradient may be regulated by several molecules related to inammation or hypoxia (27, 35), the downregulation of SDF-1a in two stress models of our study indicate that DPSCs may follow the gradient signal to efciently mobilize to the injury sites. To sum up, our present study suggests that both inammation and hypoxia might regulate the SDF-1a-CXCR4 axis in DPCs. Further studies will focus on the specic signal pathways and in vivo animal models of dental pulp injury. All these studies will deeply look through the whole mechanism of SDF-1a-CXCR4 axis in pulp regeneration, which might give some clues to the regenerative endodontic therapies.

Figure 5. Transmigration assay showed hypoxia improved DPCss migration ability, and hypoxic DPCs displayed increased responsiveness to 100 ng/mL rhSDF-1a. Data were shown as means standard deviation of three independent experiments. *p < 0.05 as compared with the normoxia group; **p < 0.05 as compared with the group of normoxia plus rhSDF-1a.

to inhibit SDF-1a production. On the other hand, these inammatory cytokines such as IL-1 and/or TNF-a may contribute to the suppression of SDF-1a production during the disease process. The results of CXCR4 production agree with Takabayashi et als report (18), which stated that 100 ng/mL LPS increased the mRNA expression of CXCR4 in human oral carcinoma cells in a 24-hour period. Petty et al also found that, during the development of LPS-induced pneumonitis in mice, a signicant increase in CXCR4 was found on circulating neutrophils (19). Recent ndings show that CXCR4 is not only involved in LPS binding but also responsible for triggering signals in response to LPS (20). The oxygen concentration of human organisms varies between the tissues; in the eyes, it varies from 1% to 5% and in the bone marrow from 0% to 4% (21). In some pathological processes such as cerebral ischemia, the oxygen tension in the brain can be even lower (22). Dental pulp is in the center of a tooth and has limited or no collateral circulation except from the small root apical end. This structure makes dental pulp susceptible to injury in which inammation can increase the internal pressure and lead to hypoxia. Normally, cell culture is performed at an atmospheric oxygen concentration (20% O2). However, many studies on stem cells indicated that the hypoxic condition can amplify progenitor cell numbers (23). The effect of hypoxia on DPCs, especially in the human species, is less reported. Sakdee et al (24) showed that 3% O2 signicantly promoted the proliferation of human DPCs from day 3 to 14. Consistent with Sakdees results, we found that DPCs proliferated at a faster rate in 1% O2. It might be explained that hypoxia reduces oxidative damage to DNA (21). Additionally, limited energy only allows the expression of highly evolutionary conserved signals, which control the self-renewal abilities of stem cells. Recently, Wang et al (25) found that simulated ischemia (serum deprivation and hypoxia) increased side population stem cells of the dental pulp. Furthermore, these stressful conditions also promoted the expression of OCT4, which plays an important role in maintaining the selfrenewal capacity and pluripotency of these dental pulp stem cells (25). Hypoxia can induce the expression of hypoxia inducible factor (HIF), which consists of an a (HIF-a) and a b (HIF-b, or ARNT) subunit. Under normoxia, HIF-a is translated but rapidly degraded. When oxygen levels decrease, the stabilized HIF-a proteins can dimerize HIF-b in the nucleus and then bind to a core sequence of (A/G CGTG) in hypoxia-response elements of target genes (26). Hypoxia-response elements have been found in nearly 200 genes, which are related to metabolism, angiogenesis, or cell differentiation. In our study, real-time PCR showed that the HIF1-a mRNA level increased in 1502

Acknowledgments
We would like to thank Siyu Cao for critical review of this article.

References
1. Goldberg M, Farges JC, Lacerda-Pinheiro S, et al. Inammatory and immunological aspects of dental pulp repair. Pharmacol Res 2008;58:13747. 2. Wang J, Loberg R, Taichman RS. The pivotal role of CXCL12 (SDF-1)/CXCR4 axis in bone metastasis. Cancer Metastasis Rev 2006;25:57387.

Gong et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology
3. Yano T, Liu Z, Donovan J, et al. Stromal cell derived factor-1 (SDF-1)/CXCL12 attenuates diabetes in mice and promotes pancreatic beta-cell survival by activation of the prosurvival kinase Akt. Diabetes 2007;56:294657. 4. Asano Y, Iimuro Y, Son G, et al. Hepatocyte growth factor promotes remodeling of murine liver brosis, accelerating recruitment of bone marrow-derived cells into the liver. Hepatol Res 2007;37:108094. 5. Saxena A, Fish JE, White MD, et al. Stromal cell-derived factor-1alpha is cardioprotective after myocardial infarction. Circulation 2008;117:222431. 6. Jiang HW, Ling JQ, Gong QM. The expression of stromal cell-derived factor 1 (SDF1) in inamed human dental pulp. J Endod 2008;34:13514. 7. Jiang L, Zhu YQ, Du R, et al. The expression and role of stromal cell-derived factor1alpha-CXCR4 axis in human dental pulp. J Endod 2008;34:93944. 8. Miller RJ, Banisadr G, Bhattacharyya BJ. CXCR4 signaling in the regulation of stem cell migration and development. J Neuroimmunol 2008;198:318. 9. Liu L, Ling J, Wei X, et al. Stem cell regulatory gene expression in human adult dental pulp and periodontal ligament cells undergoing odontogenic/osteogenic differentiation. J Endod 2009;35:136876. 10. Huang GT, Gronthos S, Shi S. Mesenchymal stem cells derived from dental tissues vs. those from other sources: their biology and role in regenerative medicine. J Dent Res 2009;88:792806. 11. Jiang HW, Zhang W, Ren BP, et al. Expression of toll like receptor 4 in normal human odontoblasts and dental pulp tissue. J Endod 2006;32:74751. 12. Mani SA, Guo W, Liao MJ, et al. The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 2008;133:70415. 13. Hosokawa Y, Hosokawa I, Ozaki K, et al. CXCL12 and CXCR4 expression by human gingival broblasts in periodontal disease. Clin Exp Immunol 2005;141:46774. 14. Sipert CR, Moraes IG, Bernardinelli N, et al. Heat-killed Enterococcus faecalis alters nitric oxide and CXCL12 production but not CXCL8 and CCL3 production by cultured human dental pulp broblasts. J Endod 2010;36:914. 15. Hahn CL, Liewehr FR. Relationships between caries bacteria, host responses, and clinical signs and symptoms of pulpitis. J Endod 2007;33:2139. 16. Fedyk ER, Jones D, Critchley HO, et al. Expression of stromal-derived factor-1 is decreased by IL-1 and TNF and in dermal wound healing. J Immunol 2001;166: 574954. 17. Morandini AC, Sipert CR, Gasparoto TH, et al. Differential production of macrophage inammatory protein-1alpha, stromal-derived factor-1, and IL-6 by human cultured periodontal ligament and gingival broblasts challenged with lipopolysaccharide from P. gingivalis. J Periodontol 2010;81:3107. 18. Takabayashi T, Takahashi N, Okamoto M, et al. Lipopolysaccharides increase the amount of CXCR4, and modulate the morphology and invasive activity of oral cancer cells in a CXCL12-dependent manner. Oral Oncol 2009;45:96873. 19. Petty JM, Sueblinvong V, Lenox CC, et al. Pulmonary stromal-derived factor-1 expression and effect on neutrophil recruitment during acute lung injury. J Immunol 2007;178:814857. 20. Triantalou M, Lepper PM, Briault CD, et al. Chemokine receptor 4 (CXCR4) is part of the lipopolysaccharide "sensing apparatus. Eur J Immunol 2008;38: 192203. 21. Ivanovic Z. Hypoxia or in situ normoxia: The stem cell paradigm. J Cell Physiol 2009; 219:2715. 22. Erecinska M, Silver IA. Tissue oxygen tension and brain sensitivity to hypoxia. Respir Physiol 2001;128:26376. 23. Keith B, Simon MC. Hypoxia-inducible factors, stem cells, and cancer. Cell 2007; 129:46572. 24. Sakdee JB, White RR, Pagonis TC, et al. Hypoxia-amplied proliferation of human dental pulp cells. J Endod 2009;35:81823. 25. Wang J, Wei X, Ling J, et al. Side population increase after simulated transient ischemia in human dental pulp cell. J Endod 2010;36:4538. 26. Weidemann A, Johnson RS. Biology of HIF-1alpha. Cell Death Differ 2008;15: 6217. 27. Ceradini DJ, Kulkarni AR, Callaghan MJ, et al. Progenitor cell trafcking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med 2004;10: 85864. 28. Schioppa T, Uranchimeg B, Saccani A, et al. Regulation of the chemokine receptor CXCR4 by hypoxia. J Exp Med 2003;198:1391402. 29. Safronova O, Morita I. Transcriptome remodeling in hypoxic inammation. J Dent Res 2010;89:43044. 30. Gross J, Moller R, Amarjargal N, et al. Expression of erythropoietin and angiogenic growth factors following inner ear injury of newborn rats. Prague Med Rep 2009; 110:31031. 31. Riek-Burchardt M, Kolodziej A, Henrich-Noack P, et al. Differential regulation of CXCL12 and PACAP mRNA expression after focal and global ischemia. Neuropharmacology 2010;58:199207. 32. Iohara K, Zheng L, Wake H, et al. A novel stem cell source for vasculogenesis in ischemia: subfraction of side population cells from dental pulp. Stem Cells 2008; 26:240818. 33. Arthur A, Shi S, Zannettino AC, et al. Implanted adult human dental pulp stem cells induce endogenous axon guidance. Stem Cells 2009;27:222937. 34. Farges JC, Keller JF, Carrouel F, et al. Odontoblasts in the dental pulp immune response. J Exp Zool B Mol Dev Evol 2009;312B:42536. 35. Kucia M, Reca R, Miekus K, et al. Trafcking of normal stem cells and metastasis of cancer stem cells involve similar mechanisms: pivotal role of the SDF-1-CXCR4 axis. Stem Cells 2005;23:87994.

JOE Volume 36, Number 9, September 2010

Regulation of SDF-1a-CXCR4 in HDPCs

1503

Basic ResearchBiology

Inherent Differential Propensity of Dental Pulp Stem Cells Derived from Human Deciduous and Permanent Teeth
Vijayendran Govindasamy, PhD,* Aimi Naim Abdullah, BSc, Veronica Sainik Ronald, BSc, Sabri Musa, BDS, MSc, Zeti Adura Che Ab. Aziz, BDS, MSc, Rosnah Binti Zain, BDS, MS, Satish Totey, PhD,* Ramesh R. Bhonde, PhD,* and Noor Hayaty Abu Kasim, PhD
Abstract
Introduction: Lately, several new stem cell sources and their effective isolation have been reported that claim to have potential for therapeutic applications. However, it is not yet clear which type of stem cell sources are most potent and best for targeted therapy. Lack of understanding of nature of these cells and their lineagespecic propensity might hinder their full potential. Therefore, understanding the gene expression prole that indicates their lineage-specic proclivity is fundamental to the development of successful cell-based therapies. Methods: We compared proliferation rate, gene expression prole, and lineage-specic propensity of stem cells derived from human deciduous (SCD) and permanent teeth (DPSCs) over 5 passages. Results: The proliferation rate of SCD was higher (cell number, 25 106 cells/mL; percent colony-forming units [CFUs], 151.67 10.5; percent cells in S/G2 phase, 12.4 1.48) than that of DPSCs (cell number, 21 106 cells/mL; percent CFUs, 133 17.62; percent cells in S/G2 phase, 10.4 1.18). It was observed that fold expression of several pluripotent markers such as OCT4, SOX2, NANOG, and REX1 were higher (>2) in SCD as compared with DPSCs. However, DPSCs showed higher expression of neuroectodermal markers PAX6, GBX2, and nestin (fold expression >100). Similarly, higher neurosphere formation and neuronal marker expression (NF, GFAP) were found in the differentiated DPSCs into neuron-like cells as compared with SCD. Conclusions: This study thus demonstrates that both SCD and DPSCs exhibit specic gene expression prole, with clear-cut inclination of DPSCs toward neuronal lineage. (J Endod 2010;36:15041515)

Key Words
Deciduous teeth, dental pulp stem cells, inherent propensity, permanent teeth

he therapeutic potential of stem cells derived from human dental pulp since its discovery (1) in regenerative medicine has been extensively studied at several preclinical (2) and clinical levels (3). The dental pulp tissue (DPT) appears to be an excellent source for stem cells because it can be obtained from the deciduous dentition requiring extraction as part of a planned serial extraction for management of occlusion and is originated from migrating neural crest cells during early development of embryos (4). The DPT can be isolated from various age groups and teeth; for example, cells isolated from dental tissue of human impacted tooth germ are known as tooth germ progenitor cells (TGPCs) (5); stem cells from human exfoliated deciduous teeth are known as SHED (6), and stem cells can also be isolated from human permanent teeth (impacted molar) (DPSCs) (7) or from apical papilla (SCAP) (8). Our present work is specically focused on stem cells from extracted deciduous (SCD) and permanent teeth (DPSCs). Past studies showed that both these groups of cells are able to differentiate into osteogenic, chondrogenic, adipogenic, and myogenic cells in culture (9, 10). Moreover, recent studies have shown that both SHED and permanent teeth are able to break germ layer commitment and differentiate into cells expressing neurons (11) and hepatocytes (12). Furthermore, this group of cells has been reported to have potential for use in cell-based therapy for neurodegenerative and cardiac diseases (13, 14). It has been established that gene expression and growth factor prole reect the source of tissue from which the stem cells have been obtained. This indicates that stem cell heterogeneity is biologically relevant. We have demonstrated previously that gene expression prole, growth pattern, and propensity of human embryonic stem cells, bone marrowderived mesenchymal stromal cells, and umbilical cord are different, and hence mesenchymal stem cells derived from various tissue sources are different from each other and indicate their propensity toward a specic lineage (15, 16). Therefore, it is logical that different tissue sources might generate stem cell products producing different cytokines and growth factors that might be more suited for specic clinical applications. Similarly, we hypothesized that gene expression varies within various sources of the same group such as in the case of stem cells of dental origin, which determines the development pathway of these cells. Although during the past few years the interest in dental stem cells has risen markedly and several reports on characterization and differentiation of dental stem cells from different sources and age groups have been published (17), the gene expression proledependent propensity toward lineage specicity remains poorly understood. Therefore, we undertook this present study comparing proliferation rate, gene

From the *Stempeutics Research Malaysia Sdn Bhd, Kuala Lumpur, Malaysia; and Department of Childrens Dentistry and Orthodontics, Department of Conservative Dentistry, and Department of Oral Pathology, Oral Medicine & Periodontology/Oral Cancer Research and Coordinating Centre(OCRCC), Faculty of Dentistry, University of Malaya, Kuala Lumpur, Malaysia. Address requests for reprints to Noor Hayaty Abu Kasim, BDS, PhD, Department of Conservative Dentistry, Faculty of Dentistry, University of Malaya, 50603 Kuala Lumpur, Malaysia. E-mail address: nhayaty@um.edu.my. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.006

1504

Govindasamy et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology
expression prole, and their lineage propensity of SCD and DPSCs to better understand their inherent therapeutic potential for specic clinical indications.

Materials and Methods

Isolation and Culture of SCD and DPSCs Sound intact human third molars from adults (2435 years of age) and deciduous teeth (58 years of age) were collected with informed consent from patients undergoing extraction at the Department of Children Dentistry and Orthodontics and Department of Oral and Maxillofacial Surgery, University of Malaya, respectively, under approved guidelines set by the Medical Ethics Committee, Faculty of Dentistry, University of Malaya (Medical Ethics Clearance Number DF CD0907/0042(L). Root surfaces were cleaned with povidone-iodine (Sigma Aldrich, St Louis, MO), and the pulps were extirpated within 2 hours after extraction and processed. The pulp tissue was minced into small fragments before digestion in a solution of 3 mg/mL collagenase type I (Gibco, Grand Island, NY) for 40 minutes at 37 C. After neutralization with 10% of fetal bovine serum (FBS) (Hyclone; Thermo Fisher Scientic Inc, Waltham, MA), the cells were centrifuged and were seeded in culture asks with culture medium containing a-MEM (Invitrogen, Carlsbad, CA), 0.5% 10,000 mg/mL penicillin/streptomycin (Invitrogen); 1% 1 Glutamax (Invitrogen) and 10% FBS, with humidied atmosphere of 95% of air and 5% of CO2 at 37 C. Nonadherent cells were removed 48 hours after initial plating. The medium was replaced every 3 days. When primary culture became subconuent, after 1014 days, cells were collected by trypsinization and processed for subsequent passages. All the experiments were done with pool of 5 dental pulp tissues. Therefore, 3 different pooled samples were used in 3 replicates. Human bone marrow samples were taken after written consent by using guidelines approved by the Ministry of Health, Malaysia. Bone marrowderived mesenchymal stem cells (BM-MSCs) cultures were established from 3 donors (age range, 1825 years) as previously described (18).
TABLE 1. List of Primers Used in this Study Gene symbol / name
18s RNA REX1 OCT 3/4 (POU5F1) SOX2 NANOG Osteocalcin Osterix ABCG2 AFP PAX6 GATA2 b-III Tubulin MSX1 NF BMP 4 hTERT Nestin HNF-3 Beta HAND 1 GFAP KRT1-5 NUUR1 TH KRT-8 TGF 1

Colony-forming Units The number of colony-forming units (CFUs) was determined by plating 100 cells in 35-mm dish. After 14 days in culture, the cells were xed in 100% methanol for 20 minutes and stained by 3% crystal violet stain. Colonies more than 2 mm in diameter were counted. The CFU equals the total number of colonies divided by the initial number of cells multiplied by 100%. Growth Kinetics The proliferation rate was determined by plating 5000 cells/cm2 from SCD and DPSCs per T25 cm2 culture ask (BD Pharmingen, San Diego, CA). Three replicates were performed for each passage and time point. Cells were detached by trypsinization after reaching conuency of 90%. Cells were counted and assessed for viability by means of trypan blue dye exclusion before the next passage. Cells were replated for subsequent passages, and total of 5 passages were studied in this experiment. Growth kinetics was analyzed by calculating population doubling (PD) time. The PD time was obtained by the formula: TD = tplg2/(lgNH lgNI), where NI is the inoculum cell number, NH is the cell harvest number, and t is the time of the culture (in hours). Cell Cycle Analysis Cells were seeded at 5000 cells/cm2 on a 35-mm tissue culture dish (BD Pharmingen) and cultured until reaching a conuence of 90%. The cells were detached, xed, and permeabilized in 70% ethanol overnight in the dark at 4 C. DNA was stained with propidium iodide/ RNase staining buffer (BD Pharmingen) in a volume of 500 mL (containing 1 106 cells) for 15 minutes at room temperature and then washed in DPBS (Invitrogen). DNA content was analyzed on Guava Technologies (Millipore, Billerica, MA) ow cytometer by using Cytosoft, Version 5.2, Guava Technologies software. Flow Cytometric Analysis The immunophenotyping was done by using ow cytometry at passage 5. On reaching 90% conuency, the cells were harvested

Forward
CGGCTACCATCCAAGGAA GCGTACGCAAATTAAAGTCCAGA CGACCATCTGCCGCTTTGAG CCCCCGGCGGCAATAGCA CCTCCTCCATGGATCTGCTTATTCA CATGAGAGCCCTCACA GCAGCTAGAAGGGAGTGGTG GTTTATCCGTGGTGTGTCTGG AGAACCTGTCACAAGCTGTG ATGAACAGTCAGCCAATGGG AGCCGGCACCTGTTGTGCAA AACAGCACGGCCATCCAGG CCTTCCCTTTAACCCTCACAC ACGCCTGAGGAATGGTTCACG GTCCTGCTAGGAGGCGCGAG AGCTATGCCCGGACCTCCAT CAGCGTTGGAACAGAGGTTGG GACAAGTGAGAGAGCAAGTG TGCCTGAGAAAGAGAACCAG CGATCAACTCACCGCCAACA CACAGTCTGCTGAGGTTGGA CGGACAGCAGTCCTCCATTAAGGT TCATCACCTGGTCACCAAGTT TGAGGTCAAGGCACAGTACG GCCCGCTTCTCTTACAGTGTGATT

Reverse
GCTGGAATTACCGCGGCT CAGCATCCTAAACAGCTCGCAGAAT CCCCCTGTCCCCCATTCCTA TCGGCGCCGGGGAGATACAT TCGGCGCCGGGGAGATACAT AGAGCGACACCCTAGAC GCAGGCAGGTGAATTCTTCC CTGAGCTATAGAGGCCTGGG GACAGCAAGCTGAGGATGTC CACACCAGGGGAAATGAGTC TGACTTCTCCTGCATGCACT CTTGGGGCCCTGGGCCTCCGA CCGATTTCTCTGCGCTTTTC GCCTCAATGGTTTCC GTTCTCCAGATGTTCTTCG GCCTGCAGCAGGAGGATCTT TGGCACAGGTGTCTCAAGGGTAG ACAGTAGTGGAAACCGGAG AGGATGAACAAACAC GTGGCTTCATCTGCTTCCTGTC GAGCTGCTCCATCTGTAGGG CTGAAATCGGCAGTACTGACAGCG GGTCGCCGTGCCTGTACT TGATGTTCCGGTTCATCTCA AGTACGTGCAGACGGTGGTAGTTT

Base pairs
186 282 572 447 259 314 358 651 675 625 243 242 284 555 338 184 388 234 254 158 155 711 124 160 497

JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

1505

Basic ResearchBiology

Figure 1. Morphology, CFUs, DNA content, and growth kinetics of SCD and DP-MSC. (A, B) Phase contrast microscope; original magnication, 10 of SCD and DPSCs, respectively, at passage 5; (C, D) CFUs of SCD and DPSCs, respectively, at passage 5; (E) long-term growth curves of SCD and DPSCs; (F) PD time of SCD and DPSCs at passages 1; and (G, H) assessment of DNA content in SCD and DPSCs, respectively, at passage 5. In all experiments, the results represent average of 5 culture replicates with SD, and a representative photomicrograph was given for each experiment. SD, standard deviation. Scale bar, 100 mmol/L. (This gure is available in color online at www.aae.org/joe/.)

with 0.05% trypsin (Invitrogen) and resuspended in phosphatebuffered saline (PBS) at a cell density of 1.5 106 cells/mL. Two hundred microliters of the cell suspension (1 105 cells) was incubated with the labeled antibodies in the dark for 1 hour at 37 C. The following antibodies were used to mark the cell surface epitopes: CD90-phycoerythrin (PE), CD44-PE, CD73-PE, CD166-PE and CD34PE, CD45-uoroisothiocyanate (FITC), and HLA-DR-FITC (all from BD Pharmingen). All analyses were standardized against negative control cells incubated with isotype-specic immunoglobulin (Ig) G1-PE and IgG1-FITC (BD Pharmingen). At least 10,000 events were acquired on Guava Technologies ow cytometer, and the results were analyzed by using Cytosoft, Version 5.2, Guava Technologies.

Differentiation of DPSCs The cultures were initiated at a density of 1000 cells/cm2 in 6-well plates and were grown to conuence and subjected to differentiation into adipogenic, chondrogenic, and osteogenic lineages as per the method described earlier (18).
1506
Govindasamy et al.

Adipogenic lineage was stimulated by inducing the cells with 10% FBS, 200 mmol/L indomethacin, 0.5 mmol/L 3-isobutyl-1-methyxanthine (IBMX), 10 mg/mL insulin, and 1 mmol/L dexamethasone (all reagents from Sigma Aldrich). Lipid droplets were visualized by staining with oil red O staining (Sigma Aldrich). The percent of adipocytes was estimated by counting 500 total cells in multiple elds. For chondrogenesis differentiation, briey cells were cultured in media supplemented with ITS+1 (Sigma Aldrich), 50 mmol/L L-ascorbic acid-2 phosphates (Sigma Aldrich), 55 mmol/L sodium pyruvate (Invitrogen), 25 mmol/L L-proline (Sigma Aldrich), and 10 ng/mL of transforming growth factor-beta (TGF-b) (Sigma Aldrich). Assessment of proteoglycan accumulation was visualized by alcian blue staining (Sigma Adrich). The osteogenic differentiation was stimulated in a 3-week culture in media supplemented with 10% FBS, 107 mol/L dexamethasone, 10 mmol/L b-glycerol phosphate (Fluka, Buchs, Switzerland), and 100 mmol/L of L-ascorbic acid-2 phosphates. Assessment of calcium

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology
TABLE 2. Phenotype Characterization, Colony-forming Ability, DNA Content and Differentiation Potential of SCD and DPSCs Cultured at Passage 5 Parameters
Phenotypic characterization (%)* CD34 CD44 CD45 CD73 CD90 CD166 HLA-DR Colonies (CFU) (% cells seeded) Cell cycle analysis % of G1/G0 phase % of S/G2 phase In vitro adipogenic differentiationk In vitro osteogenic differentiation

SCD

DPSCs

0 94.21 2.9 0 99.88 3.1 93.71 0.9 98.11 0.8 0 151.67 10.5

0 95.83 1.8 0 99.45 4.1 99.49 0.8 79.85 6.8 0 133 17.62

87.6 1.48 12.4 1.48 58 8 59 5

89.6 1.18 10.4 1.18 45 6 64 8

SCD, stem cells derived from human deciduous teeth; DPSCs, stem cells derived from human permanent teeth. *Fluorescence-activated cell sorter analysis (FACS) of SCD and DPSCs: The result shows the average value of % positive cells standard deviation to the total number of cells analyzed (n $ 5). P < .05. Percent SCD and DPSCs colonies grown at passage 5. Data are expressed as average value of % of colonies standard deviation. Percent DNA content in SCD and DPSCs observed in G1/GO and S/G2 phase at passage 5. k Percent neutral oil droplet formation stained with red O cells out of 500 total cell counted. Percent mineralized area.

Reverse Transcription Polymerase Chain Reaction and Real-time Reverse Transcription Polymerase Chain Reaction Total RNA was extracted by using Trizol (Invitrogen) according to the manufacturers instructions. The RNA was reverse-transcribed into cDNA by using Superscript II reverse transcriptase (Invitrogen) according to the manufacturers instructions. cDNA amplication was performed in a thermocycler by using Taq polymerase supplied with KCl buffer and 1.5 mmol/L MgCl2 (Invitrogen) at 94 C/1 min, 58 C/30 sec, 72 C/1 min. Polymerase chain reaction (PCR) products were resolved on 1.5% agarose (Invitrogen) gel run in 1 Tris borateethylenediaminetetraacetic acid buffer. The primer sequences are tabulated in Table 1. For the real-time PCR, the amplication reaction was performed by using Taqman Universal Master Mix and Assay-on-Demand Taqman primer/probe sets (Applied Biosystems) according to manufacturers protocol using the ABI 7900HT RT-PCR system (Applied Biosystems). The assays for OCT 4, SOX 2, NANOG, and DNMT1 were Hs00742896_s1 (130 base pairs [bp]), Hs02387400_g1 (122 bp), Hs02387400_g1 (109 bp), and Hs00945900_g1 (100 bp). Eukaryotic 18S rRNA (assay ID Hs99999901_s1 [188 bp]) was used as an internal control in all assays. The relative quantication of gene expression was assessed by DDCT method. All PCRs were performed in duplicates. The expressions of some primers in the semiquantitative reverse transcription (RT)-PCR analysis were quantied in duplicate by using SYBR green master mix. PCR reactions were run on an ABI 7900HT RTPCR system, and SDS v2.1 software was used to analyze the results. All measurements were normalized by 18s rRNA. Neurogenic Differentiation For neuronal differentiation, SCD and DPSCs were cultured at the rate of 10,000 cells/mL in a non-coated 35-mm dish containing Neurobasal-A medium (Invitrogen) supplemented with B-27 supplement (Invitrogen), penicillin-streptomycin, L-glutamine, epidermal growth factor (EGF), and basic broblast growth factor (bFGF) (all from Invitrogen) for 15 days. Neurosphere-like bodies generated after 15 days were then counted. Briey, a xed area (10 mm2) of the center of each well was converted into a digital image with a digital still camera (DSC-S70; Sony, Tokyo, Japan), and the number of neurospheres with a diameter of more than 60 mm was counted with a Scion Image Beta 4.02 (Scion Corporation, Frederick, MD). Neurospheres that were generated after 15 days of incubation were triturated by using poliSCD glass pipettes, and single-cell suspension was obtained and seeded on gelatin-coated 35-mm dishes in neurodifferentiation medium one (Neurobasal A) containing 1 mg/mL laminin (Invitrogen), 5 mg/mL bronectin (Nitta Gelatin, Osaka, Japan), 2 mmol/L L-glutamine, 10 mg/mL N2 supplement (Invitrogen), 20 ng/ mL bFGF, and 40 ng/mL EGF. The medium was changed to neurodifferentiation medium 2 (Neurobasal A) containing 1 mg/mL laminin, 5 mg/ mL bronectin, 2 mmol/L L-glutamine, 10 mg/mL N2 supplement, 20 ng/mL neurotrophin-3 (Peprotech, Rocky Hill, NJ) after 24 hours of cultivation. The medium was changed every 3 days. Immunocytochemical analysis was performed 21 days after cultivation. The cells were xed for 30 minutes in cold 4% paraformaldehyde in PBS, treated with 0.1% Triton-X for optimal penetration of cell membrane, and incubated at room temperature in a blocking solution (0.5% bovine serum albumin; Sigma Aldrich) for 30 minutes. The primary antibodies used were goat anti-human-OCT4 (1:400; ABCAM Inc, San Francisco, CA), mouse anti-human-GFAP (1:400; Chemicon, Temecula, CA), and mouse anti-human-b-III tubulin (1:400; Chemicon). Secondary antibodies used were FITC-conjugated rabbit anti-goat IgG (1:700;
1507

accumulation was visualized by von Kossa staining. Percentage of calcium accumulation was analyzed by using Image ProPlus software (Media Cybernetics, San Diego, CA). Osteogenic differentiation is presented as percent of the mineralized area in the total culture area. Quantitative amplications of osteogenic markers (osteocalcin and osterix) were carried out in duplicate by using SYBR green master mix (Applied Biosystems, Foster City, CA). Polymerase chain reactions were run on an ABI 7900HT RT-PCR system (Applied Biosystems), and SDS v2.1 software was used to analyze the results. All measurements were normalized by 18s rRNA. The sense and antisense primers used for each gene are shown in Table 1.

Human Taqman Low Density Array Human Taqman Low Density Array (TLDA) (Applied Biosystems) containing a well-dened set of validated gene expression markers to characterize embryonic stem cell identity was used for analyzing the expression of a focused panel of pluripotent and stem cell markers. The 384 wells of each TLDA card were preloaded with uorogenic probes and primers (Applied Bioystems). The pooled cDNAs were loaded on the microuidic cards for thermal cycling on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). Expression values for target genes were normalized to the expression of 18s rRNA. Transcriptional analysis was performed for BM-MSCs (early passage), SCD and DPSCs (both early and late passages). For data analysis, the ABI PRISM 7900HT sequence detection system software (SDS) calculated the levels of target (SCD and DPSCs) gene expression in samples relative to the level of expression in the calibrator (BM-MSCs) with comparative CT method (DDCT). For estimation of the fold change by TLDA when the initial transcript levels were undetectable, the initial cycle threshold (CT) value was assigned to be 39, which would lead to a possible underestimation of the actual fold change.
JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

Basic ResearchBiology

Figure 2. Immunophenotype analysis of SCD and DPSCs at passage 5. Cells were tested against human antigens CD34, CD44, CD45, CD73, CD90, CD166, and HLADR. 7-Amino-actinomycin D (7-AAD) was used to check the viability of the cells. SD, standard deviation; CD, cluster of differentiations. Scale bar, 100 mmol/L. (This gure is available in color online at www.aae.org/joe/.)

Chemicon) and rhodamine-conjugated anti-mouse IgG (1:500; Chemicon). Slides were counterstained with 4, 6-diamidino-2-phenylindole dihydrochloride (DAPI; Chemicon) for 5 minutes. Fluorescent images were captured by using a Nikon Eclipse 90i microscope (Nikon, Tokyo, Japan) and Image-Pro Express software (Media Cybernetics, Inc, Silver Spring, MD).

Karyotype Analysis Cultures were treated with colcemid 2 hours before harvest, detached by trysinization, and treated with 0.5 mol/L hypotonic solution (KCl/water) before xing with Conroys solution (3:1 methanol/glacial acetic acid). The spreads were treated with 0.005% trypsin, stained with Giemsa (Sigma Aldrich), and 2030 separate metaphase spreads were examined for each culture. Data and Statistical Analysis The descriptive statistical analyses were performed by using the software SPSS for Windows (Version 11.0; SPSS Predictive Analytics, Chicago, IL). The data were analyzed by using two-way analysis of variance (ANOVA). The signicance level was set at P = .05. Tukey post hoc multiple comparison tests were carried out to determine the differences between groups. To visualize the differences between SCD and DPSCs, we applied a novel approach based on principle component analysis (PCA) (Plymouth Routines In Multivariate Ecological Research version
1508
Govindasamy et al.

6 [PRIMER 6] software [http://www.zen87707.zen.co.uk/primer-e]) on our pluripotent array results. PCA is a mathematical algorithm that describes the data on the basis of their dissimilarity, so that a greater distance corresponds with a greater dissimilarity. The main aim of this technique is to reduce the dimensionality and to display the nature of the variation present in the data. This is achieved by creating new variables, the principle components that are linear combinations of the observations (19). The rst principle component explains as much of the variation as possible with a single statistic. The second principal component, which is uncorrelated with the rst, accounts for as much as possible of the remaining variation, and so on. If there are P variables, then it is possible to calculate P different principle components, but the rst few will normally account for most of the variation, and these might therefore be used to describe the data without undue loss of information (20).

Results
Isolation and Characterization of SCD and DPSCs Morphologic characteristics of SCD and DPSCs displayed indistinguishable broblastic morphology resembling that of BM-MSCs (Fig. 1A, B). The colony-forming properties of SCD and DPSCs were assessed. The CFUs were higher in SCD as compared with those of DPSCs (P < .05) (Table 2). The number and size of colonies were more in SCD than in DPSCs, indicating that SCD has higher
JOE Volume 36, Number 9, September 2010

Basic ResearchBiology

Figure 3. In vitro mesoderm differential potentiality of SCD and DPSCs. Osteogenesis was conrmed by mineralized matrix deposition stained with von Kossa staining at day 21 in SCD (A) and DPSCs (B) followed by conrmation of osteogenesis at mRNA level; mRNA expression of osterix and osteocalcin by RT-PCR by using SYBR green reagent, and values were normalized to 18sRNA. Adipogenesis was detected by neutral oil droplet formation stained with oil red O at day 21 in SCD (C) and DPSCs (D). Chondrogenesis was detected by the presence of proteoglycans stained with alcian blue at day 21 in SCD (E) and DPSCs (F). All experiments were conducted at passage 5. Results represent average of 5 culture replicates. (This gure is available in color online at www.aae.org/joe/.)

proliferation rate than DPSCs (Fig. 1C, D). This result reected the growth kinetics of the cells. After the end of passage 5, overall cell yield was signicantly higher in SCD (25 106 cells) in T25 cm2 ask as compared with DPSCs (21 106 cells) (Fig. 1E) (P < .05). The time required for PD varied between approximately 27 hours (P1) and 29 hours (P5) for SCD, whereas approximately 28 (P1) and 29 hours (P5) were observed in DPSCs (Fig. 1F). We further analyzed cell cycle analysis of SCD and DPSCs at passage 5. Flow cytometry analysis revealed that the percentage of proliferation rate was higher in SCD (S + G2 + M = 12.4 1.48) (P < .05) with approximately 85% of cells in phase G1/GO, as compared with that of DPSCs (S + G2 + M = 10.4

1.18) with approximately 90% of the cells in phase G1/GO (Table 2; Fig. 1G, H).

Cell Surface Antigen Prole of SCD and DPSCs Immunophenotyping of stem cells derived from SCD or DPSCs showed that the cells were negative for hematopoietic markers (21) CD34 and CD45 and HLA-DR, whereas more than 90% of cells were positive for mesenchymal stem cell markers (22) CD44, CD73, CD90, and CD166. Cells from SCD expressed higher percentage of CD166 (98.11 0.8) (P < .05) than cells derived from DPSCs (79.85 6.8) (Table 2; Fig. 2).
1509

JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

Basic ResearchBiology

Figure 4. Karyotyping analysis of SCD and DPSCs. (A) Karyotype expanded in SCD and DPSCs at passage 5 revealed chromosomal stability. Results are representative of 3 experiments.

Differentiation of SCD and DPSCs Into Classic Lineages Osteogenic differentiation was conrmed in both SCD and DPSCs by the deposition of a silver-stained mineralized matrix (Fig. 3A, B). Percent mineralization was 59 5 for SCD and 64 8 for DPSCs, respectively (Table 2). The mRNA expression of 2 osteoblast markers, osteocalcin and osterix (22, 23), was found to be higher in SCD as compared with DPSCs. Observations from the real-time PCR analyses were in line with the results from the staining assays. Adipogenic differentiation was conrmed in both SCD and DPSCs by the accumulation of neutral lipid vacuoles (Fig. 3C, D). After adipogenic induction, 58% 8% of SCD and 45% 6% of DPSCs possessed cells with an adipogenic phenotype (Table 2). Thus, a higher number of SCD (P < .05) than that of DPSCs featured an adipogenic differentiation capacity. Chondrogenic differentiation was conrmed by the formation of spheres in the micromass culture and the secretion of cartilage-specic proteoglycans stainable. Both SCD and DPSCs demonstrated a cartilage-like phenotype with chondrocyte-like lacunae (Fig. 3E, F). Cytogenetic Stability of SCD and DPSCs Both SCD and DPSCs showed normal karyotypes at passage 5. A representative ideogram is illustrated in Fig. 4. Pluripotent Gene Array Analysis between SCD and DPSCs Pluripotent marker proles of SCD and DPSCs were compared with BM-MSCs as a calibrator to evaluate the stability of the gene expression prole over the course of culture. Forty-one genes were upregulated, whereas 12 genes were down-regulated in either SCD or DPSCs as compared with BM-MSCs (Fig. 5AC). On the whole, the pluripotent was highly maintained in SCD as compared with DPSCs and BM-MSCs. The expression proles (ratio [DPSCs were kept at 1]) of some of the pluripotent markers (16, 23) between DPSCs and SCD at passage 5 were as follows: POU5F1 (OCT3/4) (18.54); TDGF1 (17.98); SOX2 (10.14); GABR3 (21.73); GAL (32.4); IFTM1 (28.40); and LIN28 (9.72). We also found that SCD expressed some
1510
Govindasamy et al.

of the endoderm and mesoderm markers such as GATA6, GATA4, SOX17, CDH5, FLT1, and DES (22, 23) as compared with DPSCs. Interestingly, DPSCs expressed higher neuron/ectoderm markers (24). The expression proles (ratio [SCD was kept at 1]) at passage 5 were as follows: nestin (10.08); GBX2 (1.14); PAX6 (1.50); and TH (2.09). PCA (Fig. 5D), on the basis of up-regulated genes, showed 98.7% of all the variance between the experimental groups (SCD and DPSCs at passages 1 and 5) could be described by using the rst (#PC1) and second principle (#PC2) component analysis. The remaining principle component (#PC3) had minor contributions to the total gene expression variance and produced no signicant shifts between the experimental groups.

Conrmation of Pluripotent Array Analysis The expression of OCT4, SOX2, and NANOG was signicantly upregulated (P < .05) in SCD as compared with DPSCs (Fig. 6A, B). On the contrary, the expression of some of the neuron markers (25) (PAX6, NUUR1, nestin, b-III tubulin) was up-regulated in DPSCs compared with in SCD. These results were consistent with the pluripotent array results (Fig. 6A, C). Neuronal Differentiation in SCD and DPSCs As shown in Fig. 7A, B, both SCD and DPSCs are capable of forming 3 distinct neurospheres. The neurospheres counted at day 15 revealed that DPSCs had a signicantly increased number of neurites as compared with SCD. The total neurospheres counted per 10 mm2 between SCD and DPSCs were as follows: 30 7 and 39 9, respectively, (P < .001) in the range of neurosphere sizes from 6080 mmol/ L; 29 1 and 25 5, respectively, in the range of neurosphere sizes from 81100 mmol/L; and 9 1 and 12 2, respectively, in neurosphere sizes greater than 100 mmol/L (Fig. 7C). Once attached in coated dish at day 16, these neurospheres spontaneously show outgrowth and dendrite-like structure by day 5 (Fig. 7DI). Real-time PCR showed a higher expression of neuronal markers (23) (bIII-tubulin, NF,
JOE Volume 36, Number 9, September 2010

Basic ResearchBiology

Figure 5. Expression prole of pluripotent/stem cells and lineage markers of SCD and DPSCs at passages 1 and 5. Expression levels of set of pluripotent/lineage marker genes. Gene levels were normalized against 18s, and mRNA from BM-MSCs was used as calibrator. (A) The genes were represented in the heat map (low expression in black and high expression in red); (B) percentage of up-regulated genes in SCD or DPSCs as compared with BM-MSCs; (C) percentage of downregulated genes in SCD or DPSCs as compared with BM-MSCs; (D) PCA of pluripotent/stem cells and lineage markers of SCD and DPSCs. Cells were plotted according to their coordinates on PC1 and PC2. (This gure is available in color online at www.aae.org/joe/.)

GFAP; P < .05; P < .001) with the down-regulation of OCT4 in neuronal cells differentiated from DPSCs, as compared with neuronal cells differentiated from SCD (Fig. 7J). This was also established at protein level (Fig. 7KN).

Discussion
The aim of this study was to characterize and to assess the propensity toward specic lineage of SCD and DPSCs. The growth kinetics results revealed that SCD possessed higher proliferation rate than DPSCs. Our results are concurrent with previous reports on the comparison between SCD and DPSCs (26). The use of SCD and DPSCs is easy for several reasons. The rst and foremost reason is the ease of isolation, noninvasive collection with less or no ethical issues compared with BM-MSCs (26). We further demonstrated that both SCD and DPSCs were able to differentiate into osteoblasts, adipocytes, and chondrocytes, thus qualifying the minimum requirement of MSCs (27). However, quantication

results of osteogenic and adipogenic indicated that SCD exhibited better differentiation capability than DPSCs. SCD was regarded as a novel population of stem cells that is capable of differentiating into various cell types (at both in vitro and in vivo) into osteoblasts, odontoblasts, adipocytes, chondrocytes, and even hepatocytes (12, 2830). On the other hand, DPSCs were considered to be more appropriate candidates for dental tissue regeneration (31) and neurodegenerative diseases (32). Nonetheless, most of the stem cells might differentiate in vitro into the desirable cells through well-dened transcriptional cascade, which can be initiated experimentally with xenobiotics such as IBMX, dexamethasone, and insulin (3335). In the current study we have investigated the inherent propensity of the stem cells to differentiate into their natural destiny on the basis of their origin. We found that SCD highly expressed many pluripotent markers as compared with DPSCs. Similar ndings have been reported previously by our group (16) by using Whartons jelly mesenchymal stem cells (WJ-MSCs), whereby higher expression of pluripotent markers was found in WJ-MSCs as compared with BM-MSCs. Hence, 1511

JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

Basic ResearchBiology

Figure 6. Expression prole of pluripotent and lineage-specic stem cell markers of SCD and DPSCs. (A) Semiquantitative RT-PCR of selected pluripotent/stem cells and lineage markers. (B) Relative levels of selected pluripotent/stem cells and lineage markers were performed by Taqman-based assay qRT-PCR. (C) Relative levels of selected pluripotent/stem cells and lineage markers were performed by SYBR greenbased qRT-PCR. The lower the CT value, the more copies are present in the specic sample. Values are presented after normalized to 18s RNA level. The average of 2 replicates is displayed.

these data support the notion that SCD is more primitive or pluripotent population of cells similar to WJ-MSCs. One of the striking features of this study is the overexpression of many transcription factors such as POU5F1 (also known as OCT3/4), SOX2, NANOG, and LIN28 that are responsible for the maintenance of pluripotency in early embryos and embryonic stem cells (36, 37). The abundant persistence of these markers leading to stemness status was very recently shown where limbal progenitor cells could be induced to pluripotent stem cells with the characteristics of embryonic stem cells under different culture conditions and without induction of exogenous transcription factors (38). Given that cells from the same organ or tissue will share some commonalities in gene expression and share the same microenvironmental niche (39), we hypothesize that the generation of ectoderm-specic cell type would be highly efcient as compared with BM-MSCs, because DPSCs are of neural crest origin (4). On the basis of our results, we predicted that SCD could also be forced to form desired cell type under the inuence of appropriate microenvironment. PCA performed on the dataset by using the regulated genes showed that SCD and DPSCs exhibited a distinct expression pattern. 1512

Van de Waterbeemd et al (40) used this PCA method to study the optimal harvest point in bacterial vaccine production, and Liu et al (41) described the detection of endogenous signaling activation pathways by an oncogenic stimulus. Here we show yet another novel facet of PCA method to distinguish between gene expression pattern of SCD and DPSCs. Dental cells are neural crest derived (4), and because neural crest stem cells can differentiate into neural cells in vivo (42, 43), dental pulp stem cells are likely to have a greater potential for neural differentiation than any other stem cells. However, our data revealed differential admixture of lineage proclivity between DPSCs and SCD, although both came from the same origin. The ectoderm lineage commitment of DPSCs was reected by their ability to express higher neuronal mRNA in undifferentiated state. To test whether the variation of neuronal mRNA found in undifferentiated DPSCs and SCD will reect their end point differentiation, we detoured undifferentiated DPSCs and SCD cells into neuronal cells and characterized them. A generally accepted experiment for the presence of neuronal cells is the induction of neurospheres (4446)

Govindasamy et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology

Figure 7. Schematic overview of the generation of neurospheres in SCD and DPSCs and gene expression prole of selected neuron markers. (A, B) Formation of neurosphere in non-coated dish for 15 days under neuron media in SCD and DPSCs, respectively; (C) the neurospheres formed under neuron-condition media at day 15. The number of neurosphere was divided with small (6180 mmol/L), medium (81100 mmol/L), and large diameters (>100 mmol/L). (D, E) Neurospheres were transferred into a coated dish at day 16 and were migrated radially out of the sphere in SCD and DPSCs, respectively. (F, G) After 2 days of maturation, cells had morphologic features typical of neuron in SCD and DPSCs, respectively. (H, I) After 5 days of maturation, dendrite-like outgrowth from SCD and DPSCs, respectively, showing complex neuronal processess (arrow and insert); (J) gene expression prole of OCT4, nestin, b-III tubulin, GFAP, and NF at day 20 in SCD and DPSCs. (KN) Specic co-immunocytochemical staining of neurospheres indicated the presence of b-III tubulin and GFAP at day 20 in SCD and DPSCs, respectively. In both co-immunocytochemical pictures, nuclei were stained with DAPI (blue) and OCT4 (green); )P < .05, ))P < .01, and )))P < .001. (This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

1513

Basic ResearchBiology
in which cells aggregate to oating spheres. We found higher neurospheres in DPSCs as compared with SCD. One of the probable reasons could be the tremendous fold expression of nestin that was found in DPSCs as compared with SCD. Nestin, the marker for neuroepithelial stem cells (47, 48), seems to be essential for the induction of neurospheres (49). Wislet-Genedebien et al (50) generated neurospheres only from rat BM-MSCs after induction of nestin expression by supplementation with N2 and B27. This abundant expression of nestin could potentially enable DPSCs to differentiate more efciently than SCD into neural cells and/or act neuroprotective after transplantation. Whether DPSCs indeed possess the ability to differentiate into neuronal committed cells is currently under investigation. We are aware that the expression of 1 or 2 markers like GFAP, NF, or bIII-tubulin is not sufcient proof for neuronal functionality. Nevertheless, our nding is a step forward in evaluating higher propensity of DPSCs toward neuronal lineage as compared with SCD, which is novel and unexpected. In conclusion, our study showed that gene variations occurred within the different sources of the same stem cells, and these variations determine their lineage propensity toward a specic destination. From this study we infer that SCD retained their plasticity over the passages, whereas DPSCs lost their plasticity and were shown to be more committed toward neuronal lineage. Our results clearly demonstrated that both SCD and DPSCs could act as useful candidates for regenerative medicine in various diseases, emphasizing the usage of DPSCs for neurologic diseases.
13. Arthur A, Rychkov G, Shi S, et al. Adult human dental pulp stem cells differentiate toward functionally active neurons under appropriate environmental cues. Stem Cells 2008;26:178795. 14. Gandia C, Arminan A, Garcia-Verdugo JM, et al. Human dental pulp stem cells improve left ventricular function, induce angiogenesis and reduce infarct size in rats with acute myocardial infarction. Stem Cells 2008;26:63845. 15. Pal R, Totey S, Mamidi MK, et al. Propensity of human embryonic stem cell lines during early stage of lineage specication controls their terminal differentiation into mature cell types. Exp Biol Med 2009;234:123043. 16. Nekanti U, Rao VB, Bahirvani AG, et al. Long-term expansion and pluripotent marker array analysis of Whartons Jelly-derived mesenchymal stem cells. Stem Cells Dev 2010;19:11730. 17. Lindroos B, Maenpaa K, Ylikomi T, et al. Characterization of human dental stem cells and buccal mucosa broblasts. Biochem Biophys Res Commun 2008;368: 32935. 18. Pal R, Hanwate M, Jan M, et al. Phenotypic and functional comparison of culture condition for upscaling of bone marrow derived mesenchymal stem cells. J Tissue Eng Regen Med 2009;3:16374. 19. A AA, Azen SP. Statistical analysis: a computer oriented approach. 2nd ed. New York: Academic Press; 1979. 20. Jollifte IT. Principle component analysis. 2nd ed. Aberdeen, UK: Springer Series in Statistics; 2002. 21. Nakamura S, Yamada Y, Baba S, et al. Culture medium study of human mesenchymal stem cells for pratical use of tissue engineering and regenerative medicine. J BioMedical Materials and Engineering 2008;18:12936. 22. Karaoz E, Dogan BN, Aksov A, et al. Isolation and in vitro characterization of dental pulp stem cells from natal teeth. Histochem Cell Biol 2010;133:95112. 23. Morsceck C, Vollner F, Saugspier M, et al. Comparison of human dental follicle cells (DFCs) and stem cells from human exfoliated deciduous teeth (SCD) after neural differentiation in vitro [published online ahead of print July 10, 2009]. Clin Oral Invest doi:10.1007/s00784-009-0310-4. 24. Vescovi AL, Reynolds BA, Fraser DD, et al. bFGF regulates the proliferative fate of unipotent (neuronal) and bipotent (neuronal/astroglial) EFG-generated CNS progenitor cells. Neuron 1993;11:95166. 25. Vollner F, Ernst W, Driemel O, et al. A two-step strategy for neuronal differentiation in vitro of human dental follicle cells. Differentiation 2009;77:43341. 26. Nakamura S, Yamada Y, Katagiri W, et al. Stem cell proliferation pathways comparison between human exfoliated deciduous teeth and dental pulp stem cells by gene expression prole from promising dental pulp. J Endod 2009;35:153642. 27. Dominici M, Blanc KL, Mueller I, et al. Minimal criteria for dening multipotent mesenchymal stromal cells: the international society for cellular therapy position statement. Cytotherapy 2006;8:3157. 28. Koyama N, Okubo Y, Nakao, et al. Evaluation of pluripotency in human dental pulp cells. J Oral Maxillofac Surg 2009;67:5016. 29. Kerkis I, Kerkis A, Dozortsev, et al. Isolation and characterization of a population of immature dental pulp stem cells expressing Oct-4 and other embryonic stem cell markers. Cells Tissues Organs 2006;184:10516. 30. Miura M, Gronthos S, Zhao M, et al. SCD: stem cells from human exfoliated deciduous teeth. Proc Natl Acad Sci U S A 2003;100:580712. 31. Batouli S, Miura M, Brahim J, et al. Comparison of stem-cell-mediated osteogenesis and dentinogenesis. J Dent Res 2003;82:97681. 32. Sasaki R, Aoki S, Yamato M, et al. Neurosphere generation from dental pulp of an adult rat incisor. Eur J Neurosci 2008;27:53848. 33. Owen M, Friedenstein AJ. Stromal stem cells: marrow-derived osteogenic precursors. Ciba Found Symp 1988;136:4260. 34. Toma C, Pittenger MF, Cahill KS, et al. Human mesenchymal stem cells differentiate to a cardiomycte phenotype in the adult murine heart. Circulation 2002;105:938. 35. Dennis JE, Merriam A, Awadallah A, et al. A quadripotential mesenchymal progenitor cell isolated from the marrow of an adult mouse. J Bone Miner Res 1999;14: 7009. 36. Niwa H. How is pluripotency determined and maintained? Development 2007;134: 63546. 37. Pesce M, Anastassiadis K, Scholer HR. Oct-4: lessons of totipotency from embryonic stem cells. Cells Tissues Organs 1999;165:14452. 38. Balasubramanian S, Babai N, Chaundhuri A, et al. Non cell-autonomous reprogramming of adult ocular progenitors: generation of pluripotent stem cells without exogenous transcription factors. Stem Cells 2009;12:305362. 39. Ma D, Ma Z, Zhang X, et al. Effect of age and extrinsic microenvironment on the proliferation and osteogenic differentiation of rat dental pulp stem cell in vitro. J Endod 2009;35:154653. 40. Van de Waterbeemd B, Streeand M, Peenings J, et al. Gene-expression-based quality indicates optimal harvest point in Bordetella pertussis cultivation for vaccine production. Biotechnol Bioeng 2009;103:9008. 41. Liu Z, Wang M, Alvarez JV, et al. Singular value decomposition-based regression activation of endogenous signaling pathways in vivo. Genome Biol 2008;9:R180.

Acknowledgments
This work was funded by University Malaya Research Grant (UMRG/RG073/09HTM) under Stempeutics Research Malaysia Sdn Bhd and University Malaya joint research collaboration. We are grateful to Ms Saratha Thevi Thrichelvam, Stempeutics Research Malaysia Sdn. Bhd, Kuala Lumpur, Malaysia for her assistance in FACS, karyotyping, cell cycle data analysis and Ms Srinda Nathan from Analisa Resources for helping with TLDA pluripotent microarray.

References
1. Gronthos S, Mankani M, Brahim J, et al. Postnatal human dental pulp stem cells (DPSCs) in vitro and in vivo. Proc Natl Acad Sci U S A 2000;97:1362530. 2. Ikeda E, Yagi K, Kojima M, et al. Multipotent cells from the human third molar: feasibility of cell-based therapy for liver disease. Differentiation 2008;76:495505. 3. DAquino R, De Rosa Alfredo, Lanza V. Human mandible bone defect repair by the grafting of dental pulp stem/progenitor cells and collagen sponge biocomplexes. Eur Cell Mater 2009;18:7583. 4. Lumsden AG. Spatial organization of the epithelium and the role of neural crest cells in the initation of the mammalian tooth germ. Development 1988;103:15569. 5. Seo BM, Miura M, Gronthos S, et al. Investigation of multipotent postnatal stem cells from human periodontal ligament. Lancet 2004;364:14955. 6. Sonoyama W, Liu Y, Fang D, et al. Mesenchymal stem cell-mediated functional tooth regeneration in swine. PLoS ONE 2006;1:e79. 7. Gronthos S, Brahim J, Li W, et al. Stem cell properties of human dental pulp stem cells. J Dent Res 2002;81:5315. 8. Laino G, Carinci F, Graziano A, et al. In vitro bone production using stem cells derived from human dental pulp. J Craniofac Surg 2006;17:5115. 9. Iohara K, Zheng L, Ito M, et al. Side population cells isolated from porcine dental pulp tissue with self-renewal and multipotency for dentinogenesis, chondrogenesis, adipogenesis, and neurogenesis. Stem Cells 2006;24:2493503. 10. Zhang W, Walboomers XF, Van Kuppevelt TH, et al. In vivo evaluation of human dental pulp stem cells differentiated towards multiple lineages. J Tissues Eng Regen Med 2008;2:11725. 11. Takeyasu M, Nozaki T, Daito M. Differentiations of dental pulp stem cells into neural lineage. Pediatric Den J 2006;16:15462. 12. Ishkitiev N, Yaegaki K, Calenic B, et al. Deciduous and permanent dental pulp mesenchymal cells acquire hepatic morphologic and functional features in vitro. J Endod 2010;36:46974.

1514

Govindasamy et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology
42. Lasky JL, Wu H. Notch signaling, brain development and human disease. Pediatr Res 2005;57:104R9. 43. Pardal R, Ortega-Saenz P, Duran R, et al. Glial-like stem cells sustain physiologic neurogenesis in the adult mammalian carotid body. Cells 2007;131:36477. 44. Cattaneo E, McKay R. Proliferation and differentiation of neuronal stem cells regulated by nerve growth factor. Nature 1990;347:7625. 45. Reynolds BA, Weiss S. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 1992;255:170710. 46. Stroch A, Paul G, Csete M, et al. Long term proliferation and dopaminergic differentiation of human mesencephalic neural precursor cells. Exp Neurol 2001;170: 31725. 47. Dahlstrand J, Lardelli M, Lendahl U. Nestin mRNA expression correlates with the central nervous system progenitor cell state in many, but not all, regions of developing central nervous system. Brain Res Dev 1995;84:10929. 48. Lendahl U, Zimmerman LB, McKay RD. CNS stem cells express a new class of intermediate lament protein. Cells 1990;60:58595. 49. Tropepe V, Hitoshi S, Sirard C, et al. Direct neural fate specication from embryonic stem cells: a primitive mammalian neural stem cell stage acquired through a default mechanism. Neuron 2001;30:6578. 50. Wislet-Gendebien S, Leprince P, Moonen G, et al. Regulation of neural markers nestin and GFAP expression by cultivated bone marrow stromal cells. J Cell Sci 2003; 116:3295302.

JOE Volume 36, Number 9, September 2010

Inherent Propensity Lineage-specic Potential of Various Sources of Dental Stem Cells

1515

Basic ResearchBiology

Comparison of Tetraacetylethylendiamine + Sodium Perborate and Sodium Hypochlorite Cytotoxicity on L929 Fibroblasts
Gabriella Simbula, PhD,* Claudia Dettori, DDS, Tania Camboni, PhD,* and Elisabetta Cotti, DDS, MS
Abstract
Introduction: Tetraacetylethylenediamine in association with sodium perborate (TAED+P) can be suggested for its use as an endodontic disinfectant because of its antimicrobial activity against different bacterial species when used at low concentrations. The purpose of this study was to measure the cytotoxicity of TAED+P on L929 broblasts and to compare it with that of sodium hypochlorite (NaOCl). Methods: L929 broblasts were grown in Dulbecco Modied Eagle Medium containing 10% fetal calf serum (FCS) at 37 C and 5% CO2. At conuence, cells were split, plated in a 96-well plate, and incubated for 24 hours to allow attachment. The two disinfectants TAED+P and NaOCl were tested at various concentrations. The neutral red uptake and the 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyl tetrazolium bromide assays were used to evaluate the cell viability. The 50% inhibitory dose values for both disinfectants were calculated and statistically analyzed. The effect of both disinfectants on broblast viability was also determined in the presence of various concentrations of FCS. One-way analysis of variance with post hoc analysis using Tukey multiple comparison test was used for parametric data. Results: Both disinfectants induced a dose-related loss of cell viability; TAED+P resulted less cytotoxic than NaOCl in all the examined experimental conditions. Conclusions: These data support the possible use of TAED+P as an endodontic irrigant. Further studies are required to analyze its antibacterial activity against endodontic pathogens. (J Endod 2010;36:15161520)

Key Words
Cytotoxicity, endodontic irrigation, sodium perborate, tetraacetylethylenediamine

t is well recognized that a variety of anaerobic and facultative aerobic bacterial species represent the main etiologic agent of pulp necrosis and apical periodontitis (1). The growth of microbial species in the endodontic environment form dense bacterial biolms (14) that may be associated with refractory periapical periodontitis (5, 6) because it offers an increased resistance to antimicrobial agents (7, 8). The primary aim for the prevention and treatment of apical periodontitis consists on the eradication of microorganisms from the endodontic system. Because mechanical instrumentation of the root canal alone is not sufcient to reduce endodontic infection and to promote healing, copious irrigation with antibacterial products in combination with the endodontic instrumentation and the application of antimicrobial medications between two successive appointments is necessary to achieve a considerable reduction in bacterial counts (1, 9). The ideal endodontic irrigating solution should have a broad antibacterial spectrum, dissolve necrotic tissue debris, and prevent the formation of smear layer or remove it when it is formed (9). Furthermore, it should be biocompatible with live host tissues (9). Sodium hypoclorite (NaOCl), at different concentrations, represents the most commonly used endodontic irrigating solution because of its broad spectrum of antimicrobial activity and its properties to dissolve necrotic tissues (8, 9). Because irrigation with NaOCl is still not predictably effective, recent studies have focused on searching for an alternative solution (9, 10, 11). Tetraacetylethylenediamine in association with sodium perborate (TAED+P) can be suggested for its use as an endodontic disinfectant because of its antimicrobial activity against different bacterial species when used at low concentrations (12). Spratt et al (12) investigated in vitro the decontamination efcacy of a TAED+P 3% solution (pH = 8), which is equivalent to 0.26% peracetic acid, against waterborne species of a biolm model. The contaminating oral species tested were Streptococcus oralis, Enterococcus faecalis, and Staphylococcus aureus. The authors concluded that TAED+P provided effective control on the dental unit water line biolms (12). Because of its capacity to release oxygen, TAED+P has also been tested as a bleaching activator and compared with sodiumperboratehydrogen peroxide and sodium perborate distilled water mixtures. The data obtained showed that TAED+P may increase the effectiveness of the bleaching process in virtue to its ability to accelerate oxygen release from sodium perboratedistilled water mixtures (13). On these bases, the aim of this study is to test the cytotoxicity TAED+P on L929 broblasts and to compare it with that of NaOCl, which is still the gold standard for endodontic irrigation (9) despite its possible adverse effects (14).

From the * Department of Toxicology, Oncology and Molecular Pathology Unit, University of Cagliari, Cagliari, Italy; and Department of Conservative Dentistry and Endodontics University of Cagliari, Cagliari, Italy. Address requests for reprints to Dr Elisabetta Cotti, via Roma #149, Cagliari, 09124, Italy. E-mail address: cottiend@ tin.it. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.010

Materials and Methods

Chemicals TAED+P (Perasafe; Antec International, Sufolk, UK) and 5% NaOCl (Niclor; Ogna, Muggio, Milano, Italy) were tested. Neutral red and (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) methyl-thiazolyl-diphenil-tetrazolium bromide (MTT) (Sigma Chemical Company, St Louis, MO) were used for the cytotoxicity tests.

1516

Simbula et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchBiology
140

2 hours

140

4 hours
*

% Viability vs Control

100 80 60 40 20 0

* ***
NaOCl TAED+P
0. 00 75 0. 00 25 0. 01 25 0. 02 5 0. 05 0. 00 5 0. 5

% Viability vs Control

120

120 100 80 60 40 20 0
0. 01 25 0. 00 75 0. 00 25 0. 02 5 0. 00 5 0. 05 0. 05 0. 5

*** ***

NaOCl TAED+P

Dose (%)

Dose (%) 140 % Viability vs Control 120 100 80 60 40 20 0

0. 00 25

140

6 hours

24 hours

% Viability vs Control

120 100 80 60 40 20 0
0. 00 75 0. 00 25 0. 00 5 0. 05 0. 01 25 0. 02 5 0. 5

***

***

***

***

**

NaOCl TAED+P

NaOCl TAED+P
0. 00 5 0. 00 75 0. 01 25 0. 02 5 0. 5

Dose (%)

Dose (%)

Figure 1. The effect of increasing concentrations of NaOCl and TAED+P on L929 cell viability determined by MTT assay. Results are the means standard deviation from three separate experiments. Signicantly different from control cells for *p < 0.05, **p < 0.01, and ***p < 0.001.

140

2 hours
% Viability vs Control

140 120 100 80 60 40 20 0

4 hours

% Viability vs Control

120 100 80 60 40 20 0
0 0. 00 25 0. 00 5 0. 00 75 0. 01 25 0. 02 5 0. 05 0. 5

** * ***

***

NaOCl TAED+P

NaOCl TAED+P
0. 00 25 0. 00 5 0. 00 75 0. 01 25 0. 02 5 0. 05 0. 05 0. 5

Dose (%)

Dose (%)

140

6 hours
% Viability vs Control

140 120 100 80 60 40 20 0

24 hours

% Viability vs Control

120 100 80 60 40 20 0
0 0. 00 25 0. 00 5 0. 00 75

*** * ***

*** *** *** *** ***

NaOCl TAED+P
0. 00 25 0. 00 5 0. 00 75 0. 01 25 0. 02 5 0. 5

NaOCl TAED+P
0. 01 25 0. 02 5 0. 05 0. 5

Dose (%)

Dose (%)

Figure 2. The effect of increasing concentrations of NaOCl and TAED+P on L929 cell viability determined by NRU assay. Results are the means standard deviation from three separate experiments. Signicantly different from control cells for *p < 0.05, **p < 0.01, and ***p < 0.001.

JOE Volume 36, Number 9, September 2010

Comparison of TAED+P and NaOCl Cytotoxicity on L929 Fibroblasts

1517

Basic ResearchBiology
Cell Line Culture The L929 broblast cell line was supplied by Interlab Cell Line Collection (Servizio Biotecnologie, IST, Genova, Italy). L929 cells were maintained in Dulbecco medium (DMEM; Invitrogen Srl, Milan, Italy) supplemented with 2 mmol/L L-Glutamine (Invitrogen), 100 IU/mL to 100 mg/mL of penicillin-streptomycin (Invitrogen), and 10% fetal calf serum (FCS) (Hyclone Laboratories Inc, Logan, UT) in a humidied atmosphere of 5% CO2/95%O2 at 37 C. At conuence, cells were washed with Hanks balanced salt solution (Invitrogen), treated with 0.05% trypsin/0.5 mmol/L EDTA, and collected in culture media. Cell suspension was counted by microscope with a Neubauer chamber and seeded in 96-well microtrite plates at a concentration of 1 104 cells/well. Each experiment was conducted using six cultures for each group; 2% TAED+P and 5% NaOCl were dissolved in distilled water and added to the culture medium to the nal concentrations specied in the text in the presence or in the absence of FCS. The cytotoxicity of the two disinfectants was evaluated after a 2-, 4-, 6-, and 24-hour incubation period. Cell Viability The cytotoxity of the test materials was examined at doses ranging from 0.0025% to 0.5%. Cell viability was determined by the neutral red uptake (NRU) and MTT assays. Determination of the viability of the adherent cells by NRU assay was performed according to Borefreund and Puerner (15). The percentage of cell viability at each dose was calculated by comparing the absorbance of each dose with the control (untreated cells). For the MTT assay, 200 mL of 1:10 diluted MTT stock (5 mg/mL of MTT in Hanks balanced salt solution) was added to each culture. A cell-free blank was included for each experiment. After incu100%

bation for 4 hours, 200 mL of destaining solution (isopropanol 10%NP40-0.4N HCl) was added to each well. The absorbance of each well was determined using the Packard Spectra Count microplate reader (Packard Instrument Co, Downers Grove, IL) at the wavelength of 570 nm with the background subtraction at 690 nm. To make data comparable between NaOCl and TAED+P, the percentage of cell viability at each tested dose was calculated by comparing the absorbance of each dose with the control (untreated cells).

Statistical Analysis Instant software Prism 5 (GraphPad Software, San Diego, CA) was used to analyze the data. One-way analysis of variance with the post hoc analysis using the Tukey multiple comparison test was used for parametric data. A p value of <0.05 was considered statistically signicant.

Results
The effect of NaOCl and TAED+P solutions at concentrations ranging from 0.0025% to 0.5% on cell viability was assessed by NRU and MTT assays. The viability of untreated cells remained unchanged all throughout the experimental period. Both disinfectants NaOCl and TAED+P induced a dose-related inhibitory effect on the cell viability of L929 broblasts (Figs. 1 and 2). Supporting the known cytotoxic action of NaOCl on cultured broblasts, the earliest toxic effect was detected by MTT assay already 2 hours after treatment in the presence of NaOCl concentrations $0.025%, where a 60% of cell loss was observed. Moreover, a progressive decrease in cell viability was observed at 4, 6, and 24 hours with all the NaOCl tested concentrations except for the 0.0025% dose, which did not affect cell viability as compared with viability of the untreated
NaOCl TAED+P

Percentage Inhibition (ID%)

75%

*** ***

***

***

50%

25%

ID50

0%

0. 05

0. 00 25

0. 00 5

0. 00 75

0. 01 25

Dose (%)

120

NaOCl 24h
% Viability vs Control

0. 02 5

0. 5

120 100 80 60 40 20 0

TAED+P 24h

% Viability vs Control

100 80 60 40 20 0

0% FCS 2% FCS 5% FCS 10% FCS

0% FCS 2% FCS 5% FCS 10% FCS

0 0. 00 25

0 0. 00 25

0. 00 5 0. 00 75

0. 01 25

0. 02 5

0. 05

0. 5

0. 00 5

0. 00 75

0. 01 25

0. 02 5

0. 05

Dose (%)

Dose (%)

Figure 3. (A) The dose-related inhibition of cell proliferation by TAED+P compared with NaOCl at concentrations 0.0025% up to 0.5% using the MTT assay at 24 hours. (B) The effect of increasing concentrations of TAED+P or NaOCl on L929 cell viability in the presence or absence of FCS. Signicantly different from control cells for *p < 0.05, **p < 0.01, and ***p < 0.001. (This gure is available in color online at www.aae.org/joe/.)

1518

Simbula et al.

JOE Volume 36, Number 9, September 2010

0. 5

Basic ResearchBiology
cells. Similar results were obtained with the NRU method (Fig. 2) even if the effect observed at early time points (2, 4, and 6 hours) was less dramatic because of the reduced sensitivity of NRU assay compared with MTT in detecting viability of the cells. Data obtained by testing the effect of various concentrations of TAED+P on cell viability showed that this disinfectant is less toxic than NaOCl at concentrations #0.025 at each examined time point; the difference in the activity of the two disinfectant was more pronounced at 24 hours. From examination with the optic microscope (data not shown), it appeared that TAED+P treatment with doses ranging from 0.0025% to 0.025% induced less changes in the morphology of broblasts compared with NaOCl at each time point evaluated. Figure 3A presents the 50% inhibition dose values for both disinfectants. The 50% inhibition dose of NaOCl and TAED+P was estimated at 50 mg/mL (0.005%) and at 350 mg/mL (0.035%), respectively. Analysis of the statistical data showed that TAED+P is less cytotoxic than NaOCl. Because the MTT assay was shown to be more sensitive than the NRU assay in detecting variation in cell viability, the following experiments were performed by MTT assay. The presence of different FCS concentrations, ranging from 0% to 10%, may mimic what happens in vivo in the presence of tissue damage. Therefore, we determined the effect on broblasts viability of both disinfectants also in the presence of increasing FCS concentrations.
30 min.

As indicated in Figure 3B, the presence or absence of FCS did not interfere with the efcacy of NaOCl and TAED+P at any of the tested concentrations. Finally, we treated broblasts with increasing concentrations of both TAED+P and NaOCl for 30 minutes, and then the medium was replaced with fresh medium without disinfectants to evaluate the recovery of the cells 24 hours after the treatment. Data obtained showed that although after 30 minutes the viability of cells treated with NaOCl and TAED+P were comparable at each tested concentration, after 24 hours the cells treated with TAED+P (and not the cells treated with NaOCl) appeared to be able to recover viability at low doses up to 0.025% (Fig. 4).

Discussion
In order to reduce the number of microorganisms from endodontic infection and to increase the chances for successful treatment and prevention of apical periodontitis, complete instrumentation of the root canal system should be accomplished in conjunction with the use of an antibacterial irrigating solution (1, 9). During endodontic treatment, the irrigants can come in contact with the periradicular tissues, and in some instance they can create complications such as tissue damage and subsequent degrees of discomfort to the patient with possible interferences with the healing process (14). Therefore, it is important to select an irrigant with therapeutic benets that could be weighed against its potential cytotoxic effects and to use a diluted concentration of the irrigant that still exhibits adequate antibacterial properties. The purpose of our investigation was to test the cytotoxicity of TAED+P compared with that of NaOCl on L929 broblasts using the NRU and MTT assays. The MTT method was shown to be more sensitive than the NRU assay in its ability to reveal cell loss (Figs. 1 and 2) in accordance with previous studies (16). Our results regarding the cytotoxicity of NaOCl are in agreement with those obtained by other researchers (17, 18) and support the suggestion to use this product at the lowest possible concentration that is still effective (1%) (9). In this study, TAED+P exhibited a cytotoxicity that was lower than that of NaOCl at concentrations #0.025 at each examined time point. Furthermore, only the broblasts treated with increasing concentrations of TAED+P for 30 minutes and then maintained in fresh medium for 24 hours were able to recover viability (Fig. 4). Finally, we did not observe any interference of different calf serum (FCS) concentrations with the cytotoxic effect of both disinfectants, indicating a stability of these compounds even in the presence of simulated tissue damage. In conclusion, both TAED+P and NaOCl induce a dose-related loss of cell viability, and from our results it appears that TAED+P is less cytotoxic than NaOCl. These data support the possible use of TAED+P as an irrigating solution if the microbiological studies in progress conrm its antibacterial activity against endodontic pathogens.

200

* ** *** ***
NaOCl TAED+P

% Viability vs Control

150

100

50

0
0. 00 75 0. 00 25 0. 01 25 0. 00 5 0. 02 5 0. 05 0. 5 0 1

Dose (%)

200

30 min.+24hours
NaOCl

% Viaility vs Control

150

TAED+P

100

***

***

***

***

***

50

References
1. Baumgartner JC, Siqueira JF, Sedgeley CM, et al. Microbiology of endodontic disease. In: Ingle JI, Bakland LK, Baumgartner JC, eds. Ingles Endodontics. 6th ed. Hamilton, Ontario: BC Decker Inc; 2008:221308. 2. Spratt DA, Pratten J, Wilson M, et al. An in vitro evaluation of the antimicrobial efcacy on biolms of root canal isolates. Int Endod J 2001;37:3007. 3. Debicka P, Lipski M, Buckowska-Ralinska J, et al. Biolm formation on root canalreview. Ann Acad Med Stetin 2008;54:1526. 4. Leonardo MR, Rossi MA, Silva LAB, et al. EM evaluation of bacterial biolm and microorganisms on the external root surface of human teeth. J Endod 2002;28: 8158. 5. Noguchi N, Noiri Y, Narimatsu M, et al. Identication and localization of extraradicular biolm-forming bacteria associated with refractory endodontic pathogens. Appl Environ Microbiol 2005;71:873843.

0
0. 00 25 0. 00 75 0. 01 25 0. 00 5 0. 02 5 0. 05 0. 5 0 1

Dose (%)

Figure 4. The effect of increasing concentrations of TAED+P and NaOCl on L929 cell viability after 30 minutes of pretreatment and 24-hour release determined by MTT assay. Results are the means standard deviation from three separate experiments. Signicantly different from control cells for *p < 0.05, **p < 0.01, and ***p < 0.001.

JOE Volume 36, Number 9, September 2010

Comparison of TAED+P and NaOCl Cytotoxicity on L929 Fibroblasts

1519

Basic ResearchBiology
6. Chavez de Paz LE. Redening the persistent infection in root canals: possible role of biolm communities. J Endod 2007;33:65262. 7. Estrela C, Sydney BG, Figueiredo JAP, et al. A model system to study antimicrobial strategies in endodontic biolms. J Appl Oral Sci 2009;17:8791. 8. Estrela C, Sydney BG, Figueiredo JAP, et al. Antibacterial efcacy of intracanal medicaments on bacterial biolm. J Appl Oral Sci 2009;17:17. 9. Zehnder M. Root canal irrigants. J Endod 2006;32:38998. 10. Giardino L, Ambu E, Savoldi E, et al. Comparative evaluation of antimicrobial efcacy of sodium hypochlorite and tetraclean against Enterococcus faecalis biolm. J Endod 2007;33:8525. 11. Torabinejad M, Shabahang S, Aprecio RM, et al. The antimicrobial effect of MTAD: an in vitro investigation. J Endod 2003;29:4003. 12. Spratt DA, Latif J, Montebugnoli LL, et al. In vitro modeling of dental water line contamination and decontamination. FEMS Microbiol Lett 2004;235:3637. 13. Celik EU, Turkun M, Yapar GD. Oxygen release of tetra acetyl ethylene diamine (TAED) and sodium perborate combination. Int Endod J 2008;41:5716. 14. Hulsmann M. R}dig T, Nordmayer S. Complications during root canal irrigation. En o dod Topics 2009;16:2763. 15. Borefreund E, Puerner JA. Cytotoxicity of metals, metal-metal and metal-chelator combinations assayed in vitro. Toxicology 1986;39:12134. 16. Hidalgo E, Bartolome R, Barroso C, et al. Silver nitrate: antimicrobial activity related to cytotoxicity in human culture broblasts. Skin Pharmacol Appl Skin Physiol 1998; 11:14051. 17. Hildago E, Bartolome R, Dominguez C. Cytotoxicity mechanisms of sodium hypochlorite in cultured human dermal broblasts and its bactericidal effectiveness. Chem Biol Interact 2002;139:26582. 18. Zhang W, Torabinejad M, Li Y. Evaluation of cytotoxicity of MTAD using the MTTTetrazolium method. J Endod 2003;29:6547.

1520

Simbula et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Effects of Different Curing Units and Luting Agents on Push-out Bond Strength of Translucent Posts
Yahya Orcun Zorba, PhD,* Ali Erdemir, PhD, Ali Turkyilmaz, and Ayce Unverdi Eldeniz, PhD
Abstract
Introduction: The aim of this study was to evaluate the effects of different curing units and 2 luting cements on the push-out bond strength of a translucent ber post. Methods: Thirty maxillary incisor roots were endodontically treated. Post spaces were prepared, and the smear layers were removed. Posts (FRC Postec Plus) were luted with either a self-etch cement (Panavia F 2.0) or a self-adhesive cement (Maxcem). Luting agents were then light-activated with a quartz-tungstenhalogen, a blue lightemitting diode, or a plasma-arc curing unit. Roots/cemented posts were transversally sectioned from coronal to apical. Push-out tests were performed, and data were analyzed by using threeway analysis of variance and Tukey tests. Results: Push-out bond strengths were signicantly affected by the type of luting agent (P < .05) and root region (P < .05). The type of light source used in curing did not affect push-out bond strengths (P > .05). Conclusions: Selfadhesive resin cement provided higher bond strength than the self-etch cement when smear layer was removed before the post cementation. The push-out bond strength in the apical portion of the root was signicantly lower than in the coronal region. The use of different curing units in the photoirradiation of dual-cured resin cement did not affect the retention of the ber post as a result of the limited light transmission capability of this post. (J Endod 2010;36:15211525)

iber posts, in combination with resin cements, have been widely used to restore endodontically treated teeth (1). The resin cements can be categorized according to adhesion technique as etch-and-rinse, self-etch, or self-adhesive. Although all 3 classes of cements have been investigated, the correct clinical choice among them is not always clear (2). In terms of polymerization, resin cements can be self-curing, light-curing, or dualcuring. Self-curing cement is difcult to manipulate and place because of its limited working time. Light-curing resin is not recommended for ber-post cementation because of inadequate depth of cure in the apical portion of the root, even when translucent posts are used (3). Dual-curing cement sets mainly through radical polymerization that is initiated by either light exposure or self-cures. Although dual-cured resin cements have been recommended for the cementation of ber posts, some dualcured cements might not reach to an adequate degree of conversion in the absence of light (4); therefore, light-curing is recommended (5). Adequate curing is considered to be especially important in the apical portion of the root canal (6). Various types of curing units are routinely available for the light activation of composite resins, including quartz tungsten halogen (QTH), light-emitting diodes (LED), plasma arcs (PAC), and lasers (7). However, limited data are available about the effects of different light sources on post bond strength. The purpose of this in vitro study was to evaluate the effects of different lightcuring techniques and adhesives on the retention of ber posts in different root regions. The null hypothesis is that the push-out strength of root canal ber posts is not significantly affected by curing technique, type of adhesive system, or root region.

Materials and Methods

Thirty extracted human maxillary incisor teeth with straight root canals were used in this study. Crowns were removed to obtain a root height of 17 mm. A working length of 1 mm above the radiographic apex was established. The root canals were instrumented by using stainless steel K-les (#08-10-15; Dentsply Maillefer, Ballaigues, Switzerland), Mtwo rotary nickel-titanium instruments (#10-15-20-25; Sweden & Martina, Due Carrare, Padova, Italy), and .06 taper proles (#30-35-40; Dentsply Maillefer) mounted in a 16:1 gear-reduction handpiece driven by an electric motor (X-Smart; Dentsply DeTrey GmbH, Konstanz, Germany). Instrumentation was performed under a loupe (Heine, Herrsching, Germany) at 2.5 magnication. The root canals were irrigated with 2.5% sodium hypochlorite after each instrument. The root canals were dried with paper points and obturated with gutta-percha cones and AH Plus sealer (Dentsply) by using a lateral condensation technique with a nger spreader (Mani Inc, Tochigi, Japan). Immediately after obturation, the post spaces were prepared by removing 10 mm of gutta-percha by using a #2 Peeso reamer (Mani Inc., Tochigi-Ken, Japan), retaining at least 5 mm of root lling at the apical level (8). The canal walls were enlarged by using FRC Postec Plus low-speed post drill (#1; Ivoclar-Vivadent, Schaan, Liechtenstein), and the smear layer was removed by using ultrasonic agitation (ART P1; Bonart Co, Taiwan), with a #15 hand-le and 5 mL of 17% ethylenediaminetetraacetic acid (EDTA) followed by 5 mL of 5.25% NaOCl as an irrigant. Final irrigation was accomplished with 10 mL of distilled water. The root canals were dried by using paper points. Before cementation, posts (FRC Postec Plus) were cleaned with 95% ethanol, and a single layer of silane coupling agent (Monobond-S; Ivoclar-Vivadent) was applied to

Key Words
Curing units, luting agents, push out bond strength

From the *Department of Restorative Dentistry and Endodontics, Faculty of Dentistry, Erciyes University, Kayseri, Turkey; Department of Endodontics, Faculty of Dentistry, Kirikkale University, Kirikkale, Turkey; and Department of Endodontics, Faculty of Dentistry, Selcuk University, Konya, Turkey. Address requests for reprints to Dr Yahya Orcun Zorba, Department of Restorative Dentistry and Endodontics, Faculty of Dentistry, Erciyes University, 38039, Melikgazi, Kayseri, Turkey. E-mail address: orcunzorba@kku.edu.tr. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.04.026

JOE Volume 36, Number 9, September 2010

Effects of Curing and Luting Agents on Push-out Bond Strength

1521

Basic ResearchTechnology
the post surface. Roots were randomly divided into 2 experimental groups according to luting material, and each group was further divided into 3 subgroups (n = 5) according to the curing unit used, as follows: (1) Panavia F 2.0 (Kuraray Dental, Tokyo, Japan) + QTH (Blue Swan; Dentanet, Istanbul, Turkey); (2) Panavia F 2.0 + LED (Blue Swan; Dentanet); (3) Panavia F 2.0 + PAC (Remecure; Remedent, Deurle, Belgium); (4) Maxcem (KerrHawe, Bioggio, Switzerland) + QTH; (5) Maxcem + LED; (6) Maxcem + PAC. The luting systems were applied according to the manufacturers instructions. The specimens were polymerized with the tip of the light unit in direct contact with the coronal end of the post. Exposure times were 90 seconds with the QTH and LED units and 12 seconds with the PAC unit. After polymerization, specimens were stored for 24 hours at 37C in light-proof boxes.
33.3/20/26.7/20 40/26.7/6.7/26.7 33.3/13.3/26.7/26.7 33.3/26.7/6.7/33.3 Maxcem) Panavia F 2.0

PAC (PA/LA/C/M)

TABLE 1. Effect of Luting Systems, Curing Technique, and Root Region on Post Retention

Scanning Electron Microscope Analysis A scanning electron microscope (SEM) (JSM-5600; JEOL Ltd, Tokyo, Japan) evaluation of the post space was also performed after smear layer removal and the primer application of Panavia. Each root was cut in a mesiodistal direction with a low-speed diamond saw for these evaluations. Also, 2 specimens tested in push-out mode were processed for observation under the SEM, with the aim of assessing the site of bond failure from each subgroup. The root sections were split along the buccolingual axis with a hammer and chisel. Exposed root canals were gold-sputtered (Polaron Range SC7620; Quorum Technology, Newhaven, UK), and SEM photomicrographs at 100 and 1500 original magnication were taken of the canal walls at 1-mm and 8-mm levels.

Root region

Results
Table 1 shows the push-out bond strengths and distribution of failures for each of the experimental groups. Push-out bond strength was found to vary signicantly according to type of luting system and root region (P < .05); however, curing unit used had no effect on push1522

Zorba et al.

Luting system

JOE Volume 36, Number 9, September 2010

Coronal Apical Coronal Apical

16.12 1.90 8.35 1.82c 11.48 1.99b 6.66 2.04c,d

Statistical Analysis Statistical analysis was performed by using SPSS, Version 11.0 (SPSS, Chicago, IL), with a signicance level set at P <.05. Three-way analysis of variance and Tukey post hoc tests were used to analyze push-out bond strength, and c2 test was used to analyze failure modes.

Light source

QTH

15.27 1.96 7.29 1.99c,d 12.28 2.45b 4.41 2.18c,d

Push-out Test Posts/cemented roots were sectioned perpendicularly to their long axes into series of 1-mm slices by using a low-speed diamond saw (Micracut; Metkon, Bursa, Turkey) under a water coolant. Six slices were obtained from each root, 3 representing the coronal region and 3 representing the apical region of the post space. Push-out bond strengths were carried out by securing each slice to the metal support of a universal testing machine (Model 500; Testometric, Rochdale, Lancashire, UK), with the apical aspect facing a cylindrical plunger of 0.65-mm diameter. Loading was performed at a crosshead speed of 0.5 mm/min until bond failure occurred. Push-out strength (MPa) was calculated by dividing the load at debonding (N) by the area (mm2) (9, 10). In total, push-out tests were conducted on 30 slices for each experimental subgroup (15 coronal, 15 apical). After bond-strength testing, failure type was determined by analyzing each sample under a stereomicroscope (Olympus SZ 6045 TR Zoomstereomicroscope; Olympus Optical Co, Tokyo, Japan) at 40 magnication. Type of failure was classied into 4 categories as follows: (1) adhesive failure between post and luting material; (2) adhesive failure between dentin and luting material; (3) cohesive failure of the post system; (4) mixed failure (a combination of 2 of the above).

Values are mean standard deviation in MPa. Symbols represent differences among luting systems (P < .05). Lower case letters represent signicant differences among root regions (P < .05). No signicant differences were found among failure modes. C, cohesive within post system; LA, adhesive between luting material and dentin; LED, light-emitting diode; M, mixed adhesive/cohesive (combination of any 2 of PA/LA/C); PA, adhesive between post and resin cement; PAC, plasma arc; QTH, quartz tungsten halogen.

Failure mode (%)

LED

16.03 1.85 7.05 1.14c,d 11.55 2.34b 6.57 1.69c,d

PAC

40/13.3/26.7/20 33.3/26.7/6.7/33.3 40/13.3/20/26.7 33.3/26.7/6.7/33.3

QTH (PA/LA/C/M)

33.3/13.3/20/33.3 33.3/33.3/6.7/26.7 33.3/13.3/26.7/26.7 26.7/33.3/6.7/33.3

LED (PA/LA/C/M)

Basic ResearchTechnology

Figure 1. Morphologic aspects of intraradicular coronal dentin at different magnications (A, with smear layer; B, without smear layer; C, after primer of Panavia F 2.0 was applied; D, after push-out test in specimens luted with Panavia F 2.0; E, after push-out test in specimens luted with Maxcem; F, mixed-type failure).

out bond strength of the tested post at either the coronal or apical level (P > .05). The bond strength of the self-adhesive (Maxcem) was higher than that of the self-etch resin cement (Panavia F) (P < .05). Apical push-out bond strength was signicantly lower than coronal push-out bond strength for all groups (P < .05). The c2 test revealed no significant difference in failure modes among the groups tested (P > .05). SEM analysis revealed more open dentin tubules in the coronal region than in the apical region. Short resin tags were observed in the open tubules with both cements tested. However, Maxcem showed more resin tags than Panavia F in both coronal and apical regions after push-out test. Figs. 1 and 2 show the morphologic changes happened in intraradicular dentin of coronal and apical root sections, respectively, before and after the smear layer removal, Panavia Fs primer, and after the testing.

Discussion
The FRC Postec post used in the present study has elasticity similar to dentin and is translucent, radiopaque, and designed to be tapered for better preservation of tooth structure (11, 12). In the present study a Peeso reamer was used to remove the coronal part of root lling and the reamer of this post system (FRC Postec Plus reamer) was used to prepare the canal dentin according to the manufacturers suggestions to achieve optimal matching between the post and root canal walls.

In the endodontic literature, there is no consensus on the time interval between the endodontic treatment and the post space preparation. Posts can be inserted immediately after completion of the root canal lling or at a later time after full setting of the sealer. Some previous studies reported that immediate post space preparation was less time-consuming procedure and preferable as a result of its demonstrated less apical leakage (13). Therefore, post spaces were immediately prepared after the root llings in this study, as we routinely apply this method in our clinics, although there is recently mentioned detrimental effect of unset epoxy resin sealer to ber post retention (14). The thin slice (1-mm) push-out strength test is considered to be a valid method for evaluating ber post adhesion to root canal walls (15). It has been reported to be more reliable than the microtensile method for measuring the adhesion of ber posts (16). This method is easy and allows fabrication of several specimens from 1 root as well as testing regional differences between root sections (15). In push-out tests, to ensure that the shear force is concentrated on the adhesive interface is important. If this force is not completely centralized, friction with part of the dentin wall can be increased, modifying the result (17). The risk of friction is higher when discs of greater thickness and with cylindrical versus conical posts were evaluated (18, 19). The tapered post used in the present study has a constant angle along its length, and friction is minimized by directing the axial force 1523

JOE Volume 36, Number 9, September 2010

Effects of Curing and Luting Agents on Push-out Bond Strength

Basic ResearchTechnology

Figure 2. Morphologic aspects of intraradicular apical dentin at different magnications (A, with smear layer; B, without smear layer; C, after primer of Panavia F 2.0 was applied; D, after push-out test in specimens luted with Panavia F 2.0; E, after push-out test in specimens luted with Maxcem; F, mixed-type failure).

from the smallest to the largest diameter, concentrating the push-out force at the adhesive interface. Hypothesis studied in the present study was partially accepted as push-out bond strength of tested ber post and was found to vary according to the type of adhesive system used and region of the root, but it was not affected when different curing units were used. When both self-etch and self-adhesive luting systems were tested to lute ber posts to root dentin, better bond strength values were obtained with self-adhesive resin cement. This result is in accordance with the ndings of Radovic et al (20) and Bitter et al (21). This could be attributable to the removal of the smear layer before the cementation of the post to dentin with alternative irrigation of 17% EDTA and 5.25% NaOCl. Self-adhesive cements contain multifunctional phosphoric acid methacrylates that are claimed to react with the hydroxyapatite of the hard tissue. To ensure a correct inltration pattern, these cements should be able to etch the substrate in short time and require optimal wetting properties to achieve a fast interaction with dentin (22). It has been reported that when smear layer is present, despite their initial acidic pH, self-adhesive cements did not produce any evidence of dentin demineralization and/or hybridization (23, 24). Therefore, the presence of the smear layer has been recognized as the weak link in bonding of selfadhesive cements. Phosphoric acid etching before self-adhesive cement application has been shown to be detrimental to effective dentin 1524

bonding (25), and usage of a milder acidic agent has been proposed (24). Most likely, use of EDTA (26) together with NaOCl effectively removed the supercial loosely bound fraction of the smear layer and enhanced the adhesion of self-adhesive cement in the present study (Fig. 1). Distilled water irrigation before post cementation probably eliminated the negative effect of NaOCl on the adhesive bonding to dentin (27). Both adhesive cements tested demonstrated measurable adhesion to root dentin, with the highest values for the coronal region and lowest for the apical region in accordance with the previous reports (28, 29). The explanation for this result could be attributed to different factors such as regional differences in the quantity, volume, and orientation of the tubules toward the apical portion of the root canal, apical sclerosis, the high cavity conguration factor (28), the difculty in visualization and access to the apical part, restrictions in the ow and distribution of the material in this region that make more bubbles and voids within the cement (30), and thick smear layer formation during the post space preparation (Figs. 1A and 2A) that could not be modied by the adhesive cement or removed by EDTA/NaOCl irrigation (Figs. 1B and 2B) for optimum bonding. The mechanical properties of dual-cure adhesives were also found to be affected by different curing strategies (31) because these cements have been shown to reach inadequate degree of conversion in the

Zorba et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
absence of light. However, within the post space, high attenuation of light passing through the canal might jeopardize bonding quality, especially in the apical region. To circumvent this problem, previous studies have recommended using high-intensity light-curing units to increase the curing efciency of adhesive resins especially in deep cavities (32). Therefore, 3 different curing units, QTH, LED, and PAC, were used to start the polymerization of dual-cure resin cements in the present study. However, no signicant difference was found in bond strength values, although high-intensity light-curing units were also evaluated. This result could be explained by reduced effectiveness of these light sources as a result of light scattering within the resin cement, shadowing produced by both the tooth structure and the post, and limited light transmission capability of translucent FRC Postec Plus post used in the present study (3336). Evaluation of the failure modes in the present study demonstrated no cohesive failure inside the root canal dentin, which indicates that bond strengths to root canal dentin were weaker compared with the shear bond strength of the dentin itself (21). Both Maxcem and Panavia F cements demonstrated more adhesive failures between post/resin cement and between cement/dentin when specimens from apical region were evaluated and compared with the coronal region. Cohesive and mixed-type failure modes were detected more in specimens obtained from coronal root region with both cements. Therefore, the bond strength between post/cement and cement/dentin could be considered as the limiting factor of the measured bond strength values.
8. Abramovitz I, Tagger M, Tamse A, Metzger Z. The effect of immediate vs. delayed post space preparation on the apical seal of a root canal lling: a study in an increased-sensitivity pressure-driven system. J Endod 2000;26:4359. 9. Hashem AA, Ghoneim AG, Lutfy RA, Fouda MY. The effect of different irrigating solutions on bond strength of two root canal-lling systems. J Endod 2009;35: 53740. 10. Babb BR, Loushine RJ, Bryan TE, et al. Bonding of self-adhesive (self-etching) root canal sealers to radicular dentin. J Endod 2009;35:57882. 11. Demiryurek EO, Kulunk S, Yuksel G, Sarac D, Bulucu B. Effects of three sealers on bond strength of a ber post. J Endod 2010;36:497501. 12. Baldissara P, Zicari F, Valandro LF, et al. Effect of root canal treatments on quartz ber posts bonding to root dentin. J Endod 2006;32:9858. 13. Solano F, Hartwell G, Appelstein C. Comparison of apical leakage between immediate versus delayed post space preparation using AH Plus sealer. J Endod 2005; 31:7524. 14. Vano M, Cury AH, Goracci C, et al. Retention of ber posts cemented at different time intervals in canals obturated using an epoxy resin sealer. J Dent 2008;36:8017. 15. Goracci C, Grandini S, Bossu M, Bertelli E, Ferrari M. Laboratory assessment of the retentive potential of adhesive posts: a review. J Dent 2007;35:82735. 16. Goracci C, Tavares AU, Fabianelli A, et al. The adhesion between microtensile and push out bond strength measurements. Eur J Oral Sci 2004;112:35361. 17. de Durao Mauricio PJ, Gonzalez-Lopez S, Aguilar-Mendoza JA, Felix S, Gonzalez Rodrguez MP. Comparison of regional bond strength in root thirds among berreinforced posts luted with different cements. J Biomed Mater Res B Appl Biomater. 2007;83:36472. 18. Patierno JM, Rueggeberg FA, Anderson RW, Weller RN, Pashley D. Push-out strength and SEM evaluation of resin composite bonded internal cervical dentin. Endod Dent Traumatol 1996;12:22736. 19. Wakeeld C, Draughn R, Sneed W, Davis T. Shear bond strengths of six bonding systems using the push-out method of in vitro testing. Oper Dent 1998;23:6976. 20. Radovic I, Mazzitelli C, Chief N, Ferrari M. Evaluation of the adhesion of ber posts cemented using different adhesive approaches. Eur J Oral Sci 2008;116:55763. 21. Bitter K, Meyer-Lueckel H, Priehn K, Kanjuparambil JP, Neumann K, Kielbassa AM. Effects of luting agent and thermocycling on bond strengths to root canal dentine. Int Endod J 2006;39:80918. 22. Moszner N, Salz U, Zimmermann J. Chemical aspects of self-etching enamel-dentin adhesives: a systematic review. Dent Mater 2005;21:895910. 23. Yang B, Ludwig K, Adelung R, Kern M. Micro-tensile bond strength of three luting resins to human regional dentin. Dent Mater 2006;22:4556. 24. Monticelli F, Osorio R, Mazzitelli C, Ferrari M, Toledano M. Limited decalcication/ diffusion of self-adhesive cements into dentin. J Dent Res 2008;87:9749. 25. De Munck J, Vargas M, Van Landuyt K, Hikita K, Lambrechts P, Van Meerbeek B. Bonding of an auto-adhesive luting material to enamel and dentin. Dent Mater 2004;20:96371. 26. Gu XH, Mao CY, Kern M. Effect of different irrigation on smear layer removal after post space preparation. J Endod 2009;35:5836. 27. Dimitrouli M, Gunay H, Geurtsen W, Luhrs AK. Push-out strength of ber posts de pending on the type of root canal lling and resin cement. Clin Oral Investig 2010 [Epub ahead of print]. 28. Onay EO, Korkmaz Y, Kiremitci A. Effect of adhesive system type and root region on the push-out bond strength of glass-bre posts to radicular dentin. Int End J 2010; 43:25968. 29. de Durao Mauricio PJ, Gonzalez-Lopez S, Aguilar-Mendoza JA, Felix S, Gonzalez Rodrguez MP. Comparison of regional bond strength in root thirds among berreinforced posts luted with different cements. J Biomed Mater Res B Appl Biomater 2007;83:36472. 30. Bolhuis P, de Gee A, Feilzer A. The inuence of fatigue loading on the quality of the cement layer and retention strength of carbon ber post-resin composite core restorations. Oper Dent 2005;30:2207. 31. Tanoue N, Koishi Y, Atsuta M, Matsumura H. Properties of dual-curable luting composites polymerized with single and dual curing modes. J Oral Rehabil 2003; 30:101521. 32. Santos GC, El-Mowafy O, Rubo JH, Santos MJ. Hardening of dual-cure resin cements and a resin composite restorative cured with QTH and LED curing units. J Can Dent Assoc 2004;70:3238. 33. Musanje L, Darvell BW. Curing light attenuation in lled-resin restorative materials. Dent Mater 2006;22:80417. 34. Faria e Silva AL, Arias VG, Soares LE, Martin AA, Martins LR. Inuence of ber post translucency on the degree of conversion of a dual-cured resin cement. J Endod 2007;33:3035. 35. Kim YK, Kim SK, Kim KH, Kwon TY. Degree of conversion of dual-cured resin cement light-cured through three bre posts within human root canals: an ex vivo study. Int Endod J 2009;42:66774. 36. Goracci C, Corciolani G, Vichi A, Ferrari M. Light-transmitting ability of marketed ber posts. J Dent Res 2008;87:11226.

Conclusion
Within the limits of this in vitro study, self-adhesive cement, Maxcem, demonstrated better bonding to dentin than self-etch cement, Panavia F, when smear layer was removed before post cementation with EDTA/NaOCl. The use of different curing units for the photoirradiation of dual-cured resin cement had no effect on the retention of FRC Postec post because of the limited light transmission capability of this post. Push-out bond strength was signicantly lower in the apical third than in the coronal region, regardless of the type of cement or curing technique used.

Acknowledgments
This study was supported by a grant-in-aid (2007-17) from the Krkkale University Scientic Project Research Foundation, Turkey. This study was also partially supported by a grant given by Selcuk University, Scientic Research Project Coordination Center and conducted in the Research Laboratories of Faculty of Dentistry.

References
1. Zhang L, Huang L, Xiong Y, Fang M, Chen JH, Ferrari M. Effect of post-space treatment on retention of ber posts in different root regions using two self-etching systems. Eur J Oral Sci 2008;116:2806. 2. Rosenstiel SF, Land MF, Crispin BJ. Dental luting agents: a review of the current literature. J Prosthet Dent 1998;80:280301. 3. Roberts HW, Leonard DL, Vandewalle KS, Cohen ME, Charlton DG. The effect of a translucent post on resin composite depth of cure. Dent Mater 2004;20:61722. 4. Kumbuloglu O, Lassila LV, User A, Vallittu PK. A study of the physical and chemical properties of four resin composite luting cements. Int J Prosthodont 2004;17: 35763. 5. Radovic I, Corciolani G, Magni E, et al. Light transmission through ber post: the effect on adhesion, elastic modulus and hardness of dual-cure resin cement. Dent Mater 2009;25:83744. 6. Goracci C, Grandini S, Bossu M, Bertelli E, Ferrari M. Laboratory assessment of the retentive potential of adhesive posts: a review. J Dent 2007;35:82735. 7. Zorba YO, Bayndr YZ, Turgut H, Yildiz M. Quality of curing in relation to different light sources by measuring hardness, degree of conversion and depth of cure. Mat Res Innov 2009;13:4647.

JOE Volume 36, Number 9, September 2010

Effects of Curing and Luting Agents on Push-out Bond Strength

1525

Basic ResearchTechnology

An In Vitro Evaluation of Performance of Two Electronic Root Canal Length Measurement Devices during Retreatment of Different Obturating Materials
Vivek Aggarwal, MDS,* Mamta Singla, MDS, and Debipada Kabi, MDS*
Abstract
Introduction: The purpose of this in vitro evaluation was to study the accuracy of Root ZX and ProPex system during retreatment of canals obturated with different obturating materials (gutta-percha + zinc oxideeugenol sealer, gutta-percha + AH plus sealer, and Resilon + Epiphany sealer). Methods: Sixty human mandibular premolar roots were instrumented with Rotary ProTaper until size F4 under copious irrigation. After instrumentation, actual working length was directly measured by using #15 FlexoFile. Electronic working length was taken with the help of Root ZX and ProPex. Teeth were divided into 3 groups and obturated by using lateral condensation of guttapercha + zinc oxideeugenol sealer (group I), guttapercha + AH plus sealer (group II), and Resilon + Epiphany sealer (group III). After 7 days, the obturating materials were removed by using Gates-Glidden burs and K-les. A new electronic working length was taken. Original and post-treatment electronic working lengths were compared with actual working length. Accuracy of both electronic root canal length measurement devices was calculated for tolerance limits of 0.5 mm and 1.0 mm. Accuracy was compared by using c2 tests, and difference between actual working length and original and post-treatment electronic working lengths was compared by using Student t test. Results: The accuracy of electronic root canal length measurement devices before root canal obturation was 83.3% and 93.3% for Root ZX and ProPex, respectively, for a tolerance limit of 0.5 mm and 100% for a tolerance limit of 1.0 mm. There was no statistically signicant effect of different obturating materials tested in this study on the accuracy of Root ZX and ProPex for tolerance limits of 0.5 mm and 1.0 mm. Student t test revealed insignicant differences between the mean of actual working length and original and posttreatment electronic working lengths. Conclusions: Root ZX and ProPex can be a useful adjunct with a high accuracy rate during root canal preparation and retreatment. (J Endod 2010;36:15261530)

Key Words
Electronic root canal length measurement devices, endodontic retreatment, endodontic working length, root canal obturation

he endodontic literature has established the role of bacteria and their by-products in initiation and progression of periapical pathosis (13). Once infected, root canal space harbors a plethora of bacteria, which grow mostly in sessile biolms, aggregates, co-aggregates, and planktonic cells (35). Reduction in bacterial count is accomplished by a triad of mechanical shaping, cleaning (with various irrigating solutions), and disinfection with intracanal medicaments (6, 7). The endodontic therapy does not always provide a desirable healing outcome. Presence of clinical signs and symptoms, along with radiographic evidence of periapical bone destruction, indicates the need for retreatment (79). The most common cause of retreatment is the incomplete debridement of the root canal space along with defective root canal space obturation (1012). Therefore, the aim of retreatment is similar to the primary treatment, ie, thorough debridement of the root canal space and to shape it, so as to receive an obturation that provides a seal in all 3 dimensions (13, 14). For this purpose, orthograde nonsurgical retreatment is considered as the rst line of treatment (13, 14). The extent of the chemomechanical debridement should be within the root canal and should extend up to minor apical constriction (15). Minor apical constriction is a histologic entity and is generally present 0.51.0 mm short of the radiographic apex (15, 16). It is very important to accurately determine the length of the root canal to limit the instrumentation and root canal obturation within the connes of canal space. For this purpose, electronic root canal length measurement devices are used as an adjunct to the well-established radiographic techniques (17). The electronic root canal length measurement devices have gained popularity worldwide. They have been classied in the literature as generations of apex locators, but the generations do not provide the details of actual working mechanism of these devices (17). Root ZX (J Morita Corp, Tokyo, Japan) is one of the most commonly used electronic root canal length measurement devices This device uses 2 different alternating currents with different frequencies. It calculates the ratio of impedances of 2 different frequencies. At the apical terminus, there is a sharp change in the impedance, and Root ZX recognizes it as the apex (17). Recently some multi-frequency electronic root canal length measurement devices have been introduced to overcome some of the drawbacks of Root ZX. However, they work on a principle similar to impedance that is based on electronic root canal length measurement devices (17). Various factors affect the impedance of the root canal walls including the smear layer and size of the

From the *Department of Dental Surgery, Safdarjung Hospital, New Delhi, India; Department of Conservative Dentistry & Endodontics, Institute of Dental Sciences & Technology, Modinagar, India; and Vivek Aggarwal is presently Assistant Professor, Department of Conservative Dentistry & Endodontics, Faculty of Dentistry, Jamia Milia Islamia, New Delhi, India. Address requests for reprints to Dr Vivek Aggarwal, Department of Dental Surgery, Safdarjung Hospital, New Delhi, India. E-mail address: drvivekaggarwal@gmail. com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.04.016

1526

Aggarwal et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Figure 1. (a) Plastic box; (b) le clip; (c) endodontic le; (d) sample tooth, (e) cold cure resin to stabilize sample; (f) saline-soaked sponge; (g) lip clip; (h) freshly mixed alginate; (i) radiographic lm.

dentinal tubules (18, 19). Recently it has been shown that root canal preparation and residual root canal lling material also affect the root canal impedance (20). Newer obturating materials, which form a hybrid layer with dentin and provide a monoblock effect, can affect the root impedance as compared with conventional gutta-percha system (20). Various studies have evaluated the accuracy of different electronic root canal length measurement devices and effect of various factors (root canal pre-enlargement, different endodontic solutions, le size, stage of apex formation) (21, 22). Very limited studies have been performed to check the accuracy of different electronic root canal length measurement devices during retreatment with various obturating materials (23). The purpose of this in vitro evaluation was to study the accuracy of Root ZX and ProPex system during retreatment of canals obturated with different obturating materials (gutta-percha + zinc oxideeugenol sealer; gutta-percha + AH plus sealer, and Resilon + Epiphany sealer).

Methods and Materials

Sixty freshly extracted, caries-free, human permanent mandibular premolars with radiographically matured apex were collected. The teeth were radiographed buccolingually and mesiodistally to rule out any aberrant canal morphology and to conrm a single canal. Teeth were placed in 5.25% NaOCl for 2 hours. Remaining soft tissue remnants were removed with the help of a hand scaler. Teeth were stored in normal saline (0.9% NaCl) for a maximum period of 1 month. Teeth having approximately same width were selected for the study as far as possible. All the samples were decoronated at the level of cementoenamel junction with help of diamond disks (Horico H557F220; Pngst & Company, South Plaineld, NJ) in a straight air motor handpiece to obtain a at reference point for future working length measurements. After location of root canal system, the canal was gently irrigated with 5.25% NaOCl. A size 15 FlexoFile (Dentsply Maillefer, Tulsa, OK) was inserted into the canal until it was observed at the apical foramen

and its tip was tangential to the foramen. The apical tip and the le were visualized with the help of EyeMag Loupes (EyeMag Smart system; Carl Zeiss, Oberkochen, Germany) with a magnication of 2.5. The stopper was gently adjusted to correspond to the at reference surface. The length between the stopper and the tip of the le was measured with the help of a digital caliper and recorded to the nearest 0.01 mm. The working length was measured by subtracting 0.5 mm from this length. This was repeated 3 times, and the average value was recorded as the initial working length of the root. Canals were instrumented with ProTaper rotary (Dentsply Maillefer) instruments in sequence with S1, followed by Sx, up to coronal two thirds. A sequential apical instrumentation was done with S2, F1, F2, F3, and nally F4 les until working length. The canals were irrigated between change of each le, alternatively delivering 10 mL of sodium hypochlorite (5.25%) and ethylenediaminetetraacetic acid (EDTA) solution (17%) with a 27-gauge needle. After root canal preparation, the canal length was again directly measured (by visualizing the apex and le), and this length was recorded as actual working length (AWL).

Measurement of Working Length with Electronic Root Canal Length Measurement Devices Root ZX and ProPex (Dentsply Maillefer, Ballaigues, Switzerland) devices were used according to the manufacturers recommendations.
TABLE 1. Accuracy (%) of Electronic Root Canal Length Measurement Device before Root Canal Obturation (tolerance limit, 0.5 mm and 1.0 mm) Root ZX TL, 0.5 mm
83.3%
TL, tolerance limit.

ProPex TL, 0.5 mm

93.3%

TL, 1.0 mm
100%

TL, 1.0 mm
100%

JOE Volume 36, Number 9, September 2010

Performance of 2 Electronic Root Canal Length Measurement Devices

1527

Basic ResearchTechnology
TABLE 2. Accuracy (%) of Electronic Root Canal Length Measurement Device during Retreatment of Various Root Canal Obturation Materials (tolerance limit, 0.5 mm and 1.0 mm) Root ZX TL, 0.5 mm
Gutta-percha and zinc oxideeugenol Gutta-percha and AH plus Resilon 70% 80% 70%

ProPex TL, 1.0 mm

90% 100% 100%

TL, 0.5 mm
90% 90% 70%

TL, 1.0 mm
90% 100% 100%

Teeth were embedded in freshly mixed alginate and placed in a plastic cylinder. The walls of the cylinder were perforated before placement of alginate. Small amount of cold cure resin was applied to rmly hold the root in place in the cylinder. The cylinder was placed in plastic box lled with saline-soaked sponge. A slot was prepared between the box and sponge to hold and standardize the x-ray lm (Fig. 1). The canals were irrigated with 5 mL of 5.25% NaOCl, and excess solution was removed with absorbent paper points. The labial clip was rmly secured to the plastic box, with part of it embedded in the saline-soaked sponge. The le clip was attached to #15 FlexoFile. For using Root ZX, the le was inserted until the apex reading was reached. The le was withdrawn to ashing bar between 1 and apex on display. For using ProPex, the le was advanced until foramen was reached, as indicated by red signal, and then withdrawn to 0.0 signal. The reading was noted as electronic working length 1 (REWL1 for Root ZX and PEWL1 for ProPex). The samples were divided into 6 groups on the basis of type of obturation and different electronic root canal length measurement devices used during retreatment. Groups I and IV were obturated by using gutta-percha and zinc oxideeugenol, groups II and V were obturated by using gutta-percha and AH plus sealer, and groups III and VI were obturated by using Resilon and Epiphany sealer. The canals were irrigated with 5 mL of NaOCl and dried with absorbent paper points. Groups I and IV were obturated by using lateral compaction of gutta-percha and zinc oxideeugenol sealer. Groups II and V were obturated by using lateral compaction of gutta-percha and AH plus sealer. In groups III and VI, Epiphany self-etching primer was applied in the canals with the help of a suitable micro brush, and excess was removed with the help of absorbent paper points. Epiphany sealer was mixed according to manufacturers recommendations and placed in the canal with the help of Resilon master point. The selected master point was coated with sealer and gently seated at the working length. Lateral compaction was performed by using digital spreader (NTD11T; Brasseler, Lemgo, Germany). Accessory medium-ne Resilon points coated with sealer were laterally compacted into the canal until they could not be introduced more than 3 mm into the root canal. The excess material was seared off and compacted with a plugger (Premier Dental Products, Plymouth Meeting, PA) 1 mm below the canal opening. The coronal surface was light-cured for 40 seconds. A postoperative radiograph was obtained. The canal entrances in all the groups were sealed with Cavit-G (3M ESPE, Seefeld, Germany). The specimens were stored at 100% humidity at 37 C for 7 days. After 7 days, the temporary restoration was removed. Before initiating retreatment, a tentative working length was obtained with the help of radiograph. The coronal and middle thirds of root canal obturation were removed by using Gates-Glidden burs along with alternate irrigation with sodium hypochlorite (5.25%) and EDTA solution (17%). The samples were then again placed in plastic cylinder with freshly mixed alginate. After setting of alginate, the remaining apical third of the root canal obturation was removed by using size 25, 30, and 35 K-les until working length 1 mm short of radiographic length was achieved. The canals were again rinsed with NaOCl and dried with absorbent paper points. A size 25 FlexoFile was taken, and root canal length 1528

was measured by using the electronic root canal length measurement devices as described previously. Root ZX was used in groups I, II, and III, and ProPex was used in groups IV, V, and VI. If an apex reading was not achieved, the le was advanced more apically. The reading was noted as EWL 2 (REWL2 for Root ZX and PEWL2 for ProPex). A new radiograph was obtained to determine the extent of root canal lling present in the root. The AWL was compared with original and post-treatment EWL. Differences were calculated (AWL EWL1, AWL EWL2), and tolerance limits of 0.5 mm and 1.0 mm were taken. A negative difference indicated that EWL was larger and le tip had crossed the foramen. A positive difference indicated that tip was short of foramen. The performance of electronic root canal length measurement devices was evaluated in terms of percentages of acceptable measurements (tolerance limit of 0.5 mm and 1.0 mm). The percentages were compared by using c2 tests. Also the mean of difference between AWL and original and post-treatment EWLs was calculated and compared by using Student t test.

Results
In 2 samples, canals were blocked with obturating material (1 each of zinc oxideeugenol and Resilon & Epiphany). Two fresh samples were prepared and replaced the original blocked samples, keeping the nal sample size of 60. The accuracy of electronic root canal length measurement devices before root canal obturation was 83.3% and 93.3% for Root ZX and ProPex, respectively, when tolerance limit of 0.5 mm was taken (Table 1). The electronic root canal length measurement devices gave 100% accurate readings for a tolerance limit of 1.0 mm. The accuracy of 2 electronic root canal length measurement devices after retreatment of various root canal obturating materials is depicted in Table 2. There was no statistically signicant effect of different obturating materials tested in this study on the accuracy of Root ZX and ProPex for tolerance limits of 0.5 mm and 1.0 mm (c2 test, P > .05). Mean differences between AWL and original and post-treatment EWLs were calculated (Tables 3 and 4). There was no statistically signicant difference between all the groups.

Discussion
The aim of the endodontic treatment/retreatment is to shape, clean, and ll the root canal system until the minor apical foramen or the cementodentinal junction (15, 16, 24). Electronic root canal length measurement devices are a useful help to a clinician to locate the root canal length. Recent electronic root canal length measurement devices provide certain advantages over the 2dimensional x-rays: no radiation exposure, convenient to use, and
TABLE 3. Mean Difference between AWL and Root ZX
0.31 mm

jAWLEWL1j EWL1 n ProPex

0.22 mm

Aggarwal et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 4. Mean Difference between AWL and EWL2
Gutta-percha and zinc oxideeugenol Gutta-percha and AH plus Resilon

jAWLEWL2j during Retreatment of Various Root Canal Obturation Materials n Root ZX

0.42 mm 0.36 mm 0.43 mm

ProPex
0.37 mm 0.29 mm 0.39 mm

a high accuracy rate (17). The electronic root canal length measurement devices work on the basis of impedance and capacitance of the root canal system. As the apical constriction is reached, there is a drastic change in the impedance and capacitance (17). The recent electronic root canal length measurement devices recognize this change as a ratio or difference of impedance between the 2 or more sine wave frequencies (17). It has been shown that root canal contents can have an inuence on the accuracy of electronic root canal length measurement devices (25, 26), although the recent electronic root canal length measurement devices give a predictable success rate even in the presence of irrigating solution and/or pus (27, 28). Al-Bulushi et al (20) showed that root canal preparation and root canal obturation affected the impedance of the root canal system, which might also affect the reading of electronic root canal length measurement devices during retreatment. The present study evaluated the inuence of different root canal obturating materials on the accuracy of 2 electronic root canal length measurement devices. It has been shown in the literature that the distance between the apical constriction and the apex ranges from 0.51.0 mm (15). Therefore in the present study, the AWL was recorded by subtracting 0.5 mm from the measurement obtained when le appeared at the foramen. The methodology used during EWL determination was modied from Alves et al (23) and Kaufman et al (28). The inner cylinder was lled with

Figure 2. Radiograph showing extrusion of root canal obturating material.

freshly mixed alginate. Cold cure resin was used to stabilize the samples. The outer plastic box was lled with 0.9% NaClsoaked sponge to obtain a conductive contact between the lip clip and the le clipKle. During EWL determination, the le was advanced to penetrate the apex and get a warning signal and then retract the le to get a consistent 0.5 (Root ZX) and 0.0 (ProPex) signal (28). During retreatment, 2 samples were blocked by root canal obturating materials. This might be due to the fact that no solvent was used during the retreatment. Kaufman et al (28) have stated that guttapercha solvents like xylol can give shorter results with electronic root canal length measurement devices and thus lead to incomplete debridement. The working length was electronically obtained before and after root canal obturation (EWL1 and EWL2). The success of both electronic root canal length measurement devices was obtained for tolerance limits of 0.5 mm and 1.0 mm. The apical zone can have a wide range of shapes, and the distance between the apical constriction and the apex ranges from 0.51.0 mm (15, 16). Various authors have taken an error range of 0.5 mm to assess the accuracy of the electronic root canal length measurement devices (2931), whereas some have relaxed the limit to 1.0 mm (32, 33). Root ZX and ProPex gave 83.3% and 93.3% success rates, respectively, for a tolerance limit of 0.5 mm. These values were comparable with results conducted on canals with/without root canal instrumentation (2123, 2528, 30). Difference between AWL and EWL was obtained, and accuracy of electronic root canal length measurement devices was taken on tolerance limits of 0.5 mm and 1.0 mm. Negative values indicated that EWL was larger than AWL, and the le had crossed the apical constriction. Post-treatment EWLs showed a trend toward longer values than the AWL. In 2 of the samples, les exceeded the tolerance limit of 1.0 mm beyond the apex (1 each of Root ZX and ProPex) during retreatment of gutta-percha and zinc oxideeugenol sealer. Radiographic examination revealed that the obturating material was pushed out of the apex, and the le had to penetrate into the obturating material to touch the conductive alginate and give an apex reading (Fig. 2). If these readings were eliminated, both the electronic root canal length measurement devices gave100% success rate for a tolerance limit of 1.0 mm during retreatment. Root ZX calculates the ratio of impedance values at 2 frequencies (8 KHz and 0.4 KHz) (17). Root ZX has a built-in processor that relates the value of ratio with the position of le. The value of ratio is independent of the canal contents. ProPex is multi-frequencybased electronic root canal length measurement device. It differs from Root ZX in terms of number of sine wave frequencies used. Also the calculation of the impedance is based on the energy of the signal in contrast to amplitude of the signal, which is used by Root ZX (30). Both of the electronic root canal length measurement devices work independently of the canal contents (17, 30). The results of the present study showed that there was no statistically signicant inuence of different obturating materials on the accuracy of both electronic root canal length measurement devices. Also there was no difference between Root ZX and ProPex during retreatment. The mean difference between the AWL and EWL before obturation was 0.31 mm for Root ZX and 0.22 mm for ProPex. During retreatment, the mean difference between the AWL and post-treatment EWL 1529

JOE Volume 36, Number 9, September 2010

Performance of 2 Electronic Root Canal Length Measurement Devices

Basic ResearchTechnology
remained less than 0.5 mm, irrespective of the obturating material or the electronic root canal length measurement devices used. The results are comparable with those of Alves et al (23) and Goldberg et al (26). In the present study, it was found that the 2 tested electronic root canal length measurement devices (Root ZX and ProPex) gave acceptable readings in determination of root canal length in single-rooted teeth after root canal preparation and during retreatment. But it should not be concluded that electronic root canal length measurement devices could replace radiographs. Electronic root canal length measurement devices should be used with caution during retreatment, because there are chances of overinstrumentation if the obturating material is extruded beyond the apex. Nevertheless, the electronic root canal length measurement devices can be a useful adjunct with a high accuracy rate during retreatment.
13. Taintor J, Ingle J, Fahid A. Retreatment versus further treatment. Clin Prevent Dent 1983;5:814. 14. Lovdahl P, Gutmann J. Problems in nonsurgical root canal retreatment. In: Gutmann J, Dumsha T, Lovdahl P, eds. Problem solving in endodontics. 3rd ed. St Louis, MO: Mosby; 1993:157202. 15. Ricucci D. Apical limit of root canal instrumentation and obturation, part 1: literature review. Int Endod J 1998;31:38493. 16. Ricucci D, Langeland K. Apical limit of root canal instrumentation and obturation, part 2: a histological study. Int Endod J 1998;31:394409. 17. Nekoofar MH, Ghandi MM, Hayes SJ, Dummer PMH. The fundamental operating principles of electronic root canal length measurement devices. Int Endod J 2006;39:595609. 18. Meredith N, Gulabivala K. Electrical impedance measurements of root canal length. Endod Dent Traumatol 1997;13:12631. 19. Levinkind M, Vandernoot TJ, Elliott JC. Electrochemical impedance characterisation of human and bovine enamel. J Dent Res 1990;69:180611. 20. Al-Bulushi A, Levinkind M, Flanagan M, Ng Y-L, Gulabivala K. Effect of canal preparation and residual root lling material on root impedance. Int Endod J 2008;41: 892904. 21. Ibarrola JL, Chapman BL, Howard JH, Knowles KI, Ludlow MO. Effect of prearing on Root ZX apex locators. J Endod 1999;25:6256. 22. Ounsi HF, Naaman A. In vitro evaluation of the reliability of the Root ZX electronic apex locator. Int Endod J 1999;32:1203. 23. Alves AMH, Felippe MCS, Felippe WT, Rocha MJC. Ex vivo evaluation of the capacity of the Tri Auto ZX to locate the apical foramen during root canal retreatment. Int Endod J 2005;38:71824. 24. Katz A, Tamse A, Kaufman AY. Tooth length determination: a review. Oral Surg Oral Med Oral Pathol 1991;72:23842. 25. Jenkins J, Walker W, Schindler W, Flores C. An in vitro evaluation of the accuracy of the root ZX in the presence of various irrigants. J Endod 2001;27:20911. 26. Goldberg F, De Silvio A, Manfre S, Nastri N. In vitro measurement accuracy of an electronic apex locator in teeth with simulated apical root resorption. J Endod 2002;28:4613. 27. Haffner C, Folwaczny M, Galler K, Hickel R. Accuracy of electronic apex locators in comparison to actual length: an in vivo study. J Dent 2005;33:61925. 28. Kaufman AY, Jella S, Yoshpe M. Accuracy of a new apex locator: an in vitro study. Int Endod J 2002;35:18692. 29. Pagavino G, Pace R, Baccetti TA. A SEM study of in vivo accuracy of the Root ZX electronic apex locator. J Endod 1998;24:43841. 30. Plotino G, Grande NM, Brigante L, Lesti B, Somma F. Ex vivo accuracy of three electronic apex locators: Root ZX, Elements Diagnostic Unit and Apex Locator and ProPex. Int Endod J 2006;39:40814. 31. Saito T, Yamashita Y. Electronic determination of root canal length by newly developed measuring device: inuences of the diameter of apical foramen, the size of Kle and the root canal irrigants. Dentist Japan (Tokyo) 1990;27:6572. 32. Fan W, Fan B, Gutmann JL, Bian Z, Fan MW. Evaluation of the accuracy of three electronic apex locators using glass tubules. Int Endod J 2006;39:12735. 33. Keller ME, Brown CE Jr, Newton CW. A clinical evaluation of the Endocater: an electronic apex locator. J Endod 1991;17:2714.

References
1. Block RM, Bushell A, Rodrigues H, Langeland K. A histopathologic, histobacteriologic, and radiographic study of periapical endodontic surgical specimens. Oral Surg Oral Med Oral Pathol 1976;42:65678. 2. Abou-Rass M, Bogen G. Microorganisms in closed periapical lesions. Int Endod J 1998;31:3947. 3. Baumgartner JC, Falkler WA. Bacteria in the apical 5 mm of infected root canals. J Endod 1991;17:3803. 4. Chugal NM, Clive JM, Spangberg LS. Endodontic infection: some biologic and treatment factors associated with outcome. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:8190. 5. Seltzer S, Soltanoff W, Sinai I, Smith J. Biologic aspects of endodontics: part IV periapical tissue reactions to root lled teeth whose canals had been instrumented short of their apices. Oral Surg Oral Med Oral Pathol 1969;28:72438. 6. Paque F, Musch U, Hulsmann M. Comparison of root canal preparation using RaCe and ProTaper rotary Ni-Ti instruments. Int Endod J 2005;8:816. 7. Barbizam JVB, Fariniuk LF, Marchesan MA, Pecora JD, Sousa-Neto MD. Effectiveness of manual and rotary instrumentation techniques for cleaning attened root canals. J Endod 2002;28:3656. 8. Abou-Rass M. Evaluation and clinical management of previous endodontic therapy. J Pros Dent 1982;47:52834. 9. De Cleen M, Schuurs A, Wesselin KP, Wu M. Periapical status and prevalence of endodontic treatment in an adult Dutch population. Int Endod J 1993;26:1129. 10. Friedman S, Stabholz A, Tamse A. Endodontic retreatment: case selection and techniquepart 3: retreatment techniques. J Endod 1990;16:5439. 11. Saunders W, Saunders E, Sadiq J, Cruickshank E. Technical standard of root canal treatment in an adult Scottish sub-population. Br Dent J 1997;10:3826. 12. Weiger R, Hitzler S, Hermle G, Lost C. Periapical status, quality of root canal llings and estimated endodontic treatment need in an urban German population. Endod Dent Traumatol 1997;13:6974.

1530

Aggarwal et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Physicochemical Properties of Methacrylate Resinbased Root Canal Sealers

Gabriela Alessandra Marin-Bauza, MS,* Fuad Jacob Abi Rached-Junior, MS,* Aline Evangelista Souza-Gabriel, PhD,* Manoel Damiao Sousa-Neto, PhD, Carlos Eduardo Saraiva Miranda, PhD,* and Yara Teresinha Correa Silva-Sousa, PhD*
Abstract
Introduction: This study assessed in vitro the physicochemical properties of 2 methacrylate resin-based sealers (Epiphany SE and Hybrid Root SEAL), comparing the results with a well-established epoxy resin-based sealer (AH Plus). Methods: Five samples of each material were used for each test (setting time, ow, radiopacity, dimensional change after setting, and solubility) according to American National Standards Institute/American Dental Association (ANSI/ADA) Specication 57. The samples were assigned to 3 groups: I, AH Plus; II, Epiphany SE; and III, Hybrid Root SEAL. The distilled and deionized water used at the solubility test was submitted to atomic absorption spectrometry to observe the presence of Ca2+, K+, Ni2+, and Zn2+ ions. In addition, the surface morphology of the specimens was analyzed by means of scanning electron microscopy (SEM). Statistical analysis was performed by using one-way analysis of variance and Tukey-Kramer test (P < .05). Results: Flow, radiopacity, and solubility of all sealers were in accordance with ANSI/ADA. The setting time of Hybrid Root SEAL did not agree with ANSI/ADA requirements. The dimensional change of all sealers was greater than the values considered acceptable by ANSI/ADA. The spectrometry analysis showed signicant Ca2+ ions release for AH Plus. In SEM analysis, Hybrid Root SEAL presented spherical monomers with inferior size than AH Plus and Epiphany SE. Conclusions: It might be concluded that physicochemical properties of the tested sealers conformed to ANSI/ADA (2000) standardization, except for the setting time of Hybrid Root SEAL and the dimensional change of all sealers, which did not fulll the ANSI/ADA requirements. (J Endod 2010;36:15311536)

Key Words
Epoxy-resin and sealer, methacrylate, physicochemical

From the *School of Dentistry, University of Ribeirao Preto, Ribeirao Preto, Sao Paulo, Brazil; and School of Dentistry, Department of Restorative Dentistry, University of Sao Paulo, Ribeirao Preto, Sao Paulo, Brazil. Address requests for reprints to Prof. Yara T. Correa Silva Sousa, Rua Celia de Oliveira Meireles, 350 Bairro Jardim Canada, Ribeirao Preto, Sao Paulo, CEP 14024-070, Brazil. E-mail address: ysousa@unaerp.br. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.002

oot canal sealers should present good physicochemical properties and appropriate biological performance (1). Research accomplished in the endodontic eld aims to obtain a sealer that meets all the ideal criteria to achieve the best prognosis of root canal therapy (2, 3). Synthetic resins have been discussed and tested as endodontic core lling materials for many decades. Because of the success of the resin-based sealers, the AH series was developed more than 50 years ago (4). AH Plus (Dentsply De Trey Gmbh, Konstanz, Germany), a 2-component paste/paste sealer, is the result of this product development and is frequently used in this kind of research because of its well-studied and good physicochemical properties (58). A new lling material commercially named as Real Seal (SybronEndo, Orange, CA) or Epiphany (Pentron Clinical Technologies, Wallingford, CT) has been introduced to replace conventional sealers, with the promise of improving clinical performance (1, 9). Epiphany is a dual-cure methacrylate resin-based cement (9, 10). The resin matrix consists of a mixture of urethane dimethacrylate (UDMA), polyethylene glycol dimethacrylate (PEGDMA), ethoxylated bisphenol A dimethacrylate (EBPADMA), bisphenol A glycidyl dimethacrylate (BisGMA), silane-treated glass barium sulfate, silica, calcium hydroxide, bismuth oxychloride, peroxide, photo initiator, stabilizers, and pigments (7). This system uses a self-etching primer and also comprises Resilon (Resilon Research LLC, Madison, CT), which is a radiopaque thermoplastic synthetic material that contains bioactive glass, bismuth oxychloride, and barium sulfate (11, 12). In essence, a monoblock effect is produced, in which the core material (Resilon), sealer (Epiphany), and dentinal tubules become a single solid structure (13, 14). The second generation of methacrylate sealer replaced the original by the introduction of the Epiphany SE Self-Etch Sealer (Pentron Clinical Technologies), a self-etch dual-cure, hydrophilic resin sealer that bonds to both Resilon and dentin, without a separate priming step (15). According to the manufacturer, the 2 main bases of the Epiphany resinous matrix (EBPADMA and BisGMA) were kept the same in this new system. Acidic methacrylate resins and 2-hydroxyethyl methacrylate (HEMA) were added to Epiphany SE and were responsible for the self-etching characteristics of the sealer (16). Hybrid Root SEAL sealer (Sun Medical, Tokyo, Japan), also commercialized as MetaSEAL (Parkell Inc, Farmington, NY), has been recently introduced as another option for root canal obturation. This methacrylate resin-based sealer also has adhesion to the radicular dentin and to solid lling materials (17, 18). This sealer is a self-etch hydrophilic material as a result of the inclusion of the acidic monomer 4methacryloxyethyl trimellitate anhydride (4-META), which is capable of diffusing through the demineralized surface to form the hybrid layer after polymerization (19, 20). However, because of the presence of this acid radical in its composition, the cytotoxicity of Hybrid Root SEAL is high in the rst week after placement, decreasing over time (21). The combination of resin-based sealers and adhesive systems has been demonstrated to improve performance regarding the sealing capacity (22), increase the bond strength to dentin (13, 14), and form hybrid layer (23). Therefore, a number of tests have been developed to assess the physical properties of the endodontic lling

JOE Volume 36, Number 9, September 2010

Physicochemical Properties of Methacrylate Resinbased Root Canal Sealers

1531

Basic ResearchTechnology
materials to ensure workability when used in clinical situations (5). To date, literature is limited to provide information regarding physical properties of Epiphany SE and Hybrid Root SEAL. The purpose of this in vitro study was to assess the setting time, ow, radiopacity, dimensional change after setting, and solubility of the 2 methacrylate resin-based sealers (Epiphany SE Self-Etch Sealer and Hybrid Root SEAL) and 1 epoxy-amine resin sealer (AH Plus), according to American National Standards Institute/American Dental Association (ANSI/ADA) (2000) standards. In addition, the external and internal surface morphology of all sealers was analyzed by using scanning electron microscopy (SEM).

Materials and Methods

Setting time, ow, radiopacity, solubility, and dimensional change after setting for AH Plus (group I), Epiphany SE (group II), and Hybrid Root SEAL (group III) sealers were measured according to ANSI/ADA standards for dental root canal sealing materials (24) by one evaluator blinded to the identication of the materials. All tested materials were manipulated in accordance to the manufacturers instructions. AH Plus and Epiphany SE are displayed as 2 pastes, and for each tested sample, 15 mm of the sealer was dispensed onto a mixing pad and spatulated for 20 seconds to obtain a homogeneous mixture. Hybrid Root SEAL is a power/liquid sealer and was spatulated for 20 seconds onto a mixing pad in the proportion of 1:3. Because the sealers in groups II and III were dual-curable resins, they were mixed and handled in a darkroom under low wattage red safelight bulb (15 W). In addition, the specimens in groups II and III were light-cured (Ultralux; Dabi Atlante, Ribeirao Preto, SP, Brazil) for 40 seconds and 20 seconds, respectively. For physicochemical tests, the arithmetic mean of 5 replicates for each sealer was recorded and considered as the result of the test.

Radiopacity Test Five acrylic plates (2.2 cm 4.5 cm 1 mm), containing 4 wells measuring 1 mm in depth and 5 mm in diameter, were prepared and placed over a glass plate covered by cellophane sheet. In group I, the freshly mixed sealer was introduced into the wells by using a syringe to avoid the formation of bubbles. In groups II and III, the respective material applicators were used to ll the wells. Another glass plate covered with cellophane was placed on top until complete setting (chemically or light-cured), and any excess sealer was removed. Each plate was kept in an incubator (37  C, 95% relative humidity) for a period corresponding to 3 times the setting time. Each well was lled with one of the sealers, following a sequence according to the setting time of the material from the longest to the shortest, so that the samples would be ready for radiographic evaluation after the nal setting of all materials. Each one of the acrylic plates containing the root lling materials was positioned, at the time of the radiographic exposure, alongside another acrylic plate (1.3 cm 4.5 cm 1 mm) containing an aluminum step wedge made of 1100 alloy, with the thickness varying from 110 mm in uniform steps of 1 mm each (23). This set of acrylic plates was placed in front of this phosphor plate, and a digital radiograph was taken (Digora system; Soredex Orion Corporation, Helsinki, Finland). Care was taken to place the samples next to the aluminum step wedge and in the middle of the phosphor plate. Radiographic images were obtained by using the Spectro 70x x-ray machine (Dabi Atlante) at 70 kVp and 8 mA. The object-to-focus distance was 30 cm, and the exposure time was 0.2 seconds. Exposed imaging plates of the test samples were immediately scanned after exposure (Digora Scanner) and analyzed by using Digora for Windows 5.1 software. Dimensional Change after Setting Five Teon molds, prepared for the production of 3.58-mm-high cylindrical test bodies measuring 3 mm in diameter, were placed on a glass plate wrapped with a ne cellophane sheet. The molds were lled with a slight excess of freshly mixed sealers, and a microscope slide, also wrapped in cellophane, was pressed onto the upper surface of the mold. The assembled group was kept rmly joined with a C-shaped clamp and transferred to an incubator (37  C, 95% relative humidity) left to stand for a period corresponding to 3 times the setting time. After this period, the at ends of the molds, containing the samples, were grinded with 600-grit wet sandpaper. The samples were removed from the mold, measured with a digital caliper, stored in a 50-mL vessel containing 2.24 mL of distilled and deionized water, and kept in an incubator (37  C, 95% relative humidity) for 30 days. The sample was then removed from the container, blotted dry on absorbent paper, and measured again for length. The percentage of the dimensional alterations was calculated by using the following formula:
L30 L 100 L where L30 is the length of the sample after 30 days of storage, and L is the initial length of the sample.

Setting Time Five plasters of cast rings with an internal diameter of 10 mm and a thickness of 2 mm were prepared for each group. The external borders of the molds were xed with wax on a glass plate (75 25 1 mm). The molds were lled with the material and transferred to a chamber with 95% relative humidity and a temperature of 37  C. In addition, the specimens in groups II and III were light-cured for 40 and 20 seconds, respectively (Ultralux; Dabi Atlante). After 150 10 seconds from the onset of mixing, a Gilmore-type needle with a mass of 100 0.5 g and a at end of 2.0 0.1 mm in diameter was carefully lowered vertically onto the horizontal surface of the testing sample. The needle tip was cleaned, and the probing was repeated until indentations ceased to be visible. The time used from the start of mixing to this point was recorded. If the results differed by more than 5%, the test was repeated. Flow Test The amount of 0.5 mL of each sealer tested was placed on a glass plate (10 10 3 mm) by using a graduated disposable 3-mL syringe. At 180 5 seconds after the onset of mixing, another plate with a mass of 20 2 g and a load of 100 g were carefully applied on top of the material. Ten minutes after mixing the cement, the load was removed, and the major and minor diameters of the compressed discs were measured by using a digital calliper with a resolution of 0.01 mm (Mitutoyo MTI Corporation, Tokyo, Japan). If both measurements were consistent to within 1 mm, the results were recorded. If the major and minor diameter discs were not uniformly circular or did not match within 1 mm, the test was repeated.
1532
Marin-Bauza et al.

Solubility A 1.5-mm-thick cylindrical Teon (polytetrauroethylene; DuPont, HABIA, Knivsta, Sweden) mold measuring 7.75 mm in inner diameter was lled with freshly mixed sealers. The mold was supported by a larger glass plate and covered with a cellophane sheet. An impermeable nylon thread was placed inside the material, and another glass plate, also covered with cellophane lm, was positioned on the mold
JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
and pressed manually in such a way that the plates touched the entire mold in a uniform manner. The assembly was placed in an incubator (37  C, 95% relative humidity) and left to stand for a period corresponding to 3 times the setting time. As soon as the samples were removed from the mold, they were weighed 3 times each in an analytical balance (HM-200; A & D Engineering Inc, Bradford, MA), and the mean reading was recorded. The samples were suspended by nylon thread and placed 2-by-2 inside a plastic vessel with a wide opening containing 7.5 mL of distilled and deionized water, taking care to avoid any contact between them and the inner surface of the container. The containers were sealed and left for 7 days in an incubator (37  C, 95% relative humidity). After this period, the samples were removed from the containers, rinsed with distilled and deionized water, blotted dry with absorbent paper, and placed in a dehumidier for 24 hours. Afterwards, they were weighed again. The weight loss of each sample, expressed as percentage of the original mass (m% = mi mf), was taken as the solubility of the sealer. The liquid of the specimens immersion was analyzed by atomic absorption spectrometry (Perkin Elmer Instruments GmbH, Uberlingen, Germany) to quantify the levels of Ca2+, K+, Ni2+, and Zn2+ ions. Merck solutions with 1000 mg/L concentration (Merck, Darmstadt, Germany) were used to prepare the standard solution of the different metals. Metal analytical curves were obtained from appropriate dilutions of each respective stock solution. The concentration intervals of the solutions were Ca2+, 020 mg/L; K+, 0.21 mg/L; Zn2+, 01.5 mg/L; and Ni2+, 0.31.5 mg/L. During the chemical analyses of the solutions, the following wavelengths were used: Ca2+, 422.7 nm; K+, 766.5 nm; Zn2+, 213.9 nm; and Ni2+, 234.0 nm. This spectrophotometer is supplied with hollow cathode lamps with different light spectra exclusively for measuring metallic ions. The obtained solutions were sprayed into the atomic absorption spectrophotometer for ion quantication. The analyses were made in triplicate for each specimen; arithmetic mean was done and considered as the results of the concentration of Ca2+, K+, Zn2+, and Ni2+, expressed as mg/mL. time. After that, the samples were sectioned with a #15 disposable surgical scalpel blade, xed on a metallic stub (10 5 mm), and sputter-coated with gold-palladium (Bal-Tec AG, Balzers, Germany) at 20 mA. The internal surface morphology of the samples was qualitatively assessed by a eld emission SEM (JSM 5410; Jeol Technic Co, Tokyo, Japan) at an accelerating voltage of 15 kV at different magnications.

Results
Physicochemical Properties Table 1 shows the mean values and standard deviations of the physicochemical properties (setting time, ow, radiopacity, dimensional change after setting, and solubility) of the tested sealers. Setting Time The ANSI/ADA (2000) requires that the setting time of a sealer shall be within 10% of that stated by the manufacturers. According to them, the setting times of AH Plus, Epiphany SE, and Hybrid Root SEAL were 810 hours (480600 minutes), 25 minutes, and 16 hours (960 minutes), respectively. In this study, AH Plus and Epiphany SE were in agreement with the ANSI/ADA standards, and Hybrid Root SEAL did not conform to the standards. Statistical analysis demonstrated that the setting time of AH Plus (579.00 4.95 minutes) was signicantly longer than that of the other groups. Hybrid Root SEAL (63.40 2.70 minutes) showed intermediate results, and Epiphany SE (24.40 4.39 minutes) exhibited the inferior setting time (one-way ANOVA, post hoc Tukey-Kramer test, P < .05). In summary, AH Plus > Hybrid Root SEAL > Epiphany SE. Flow Test The ANSI/ADA (2000) requires that a sealer shall have a diameter of no less than 20 mm, and all groups of this study conformed to the standards. Statistical analysis showed that the mean values of Hybrid Root SEAL (45.94 0.61 mm) were signicantly higher than those of AH Plus (36.76 3.04 mm) and Epiphany SE (34.43 2.17 mm), which were statistically similar between themselves (one-way ANOVA, post hoc Tukey-Kramer test, P < .05). In summary, Hybrid Root SEAL > AH Plus = Epiphany SE. Radiopacity Test All materials showed radiopacity above the 3 mm of aluminum recommended by ANSI/ADA Specication 57 (2000). Statistical analysis demonstrated that AH Plus presented the superior results (5.97 0.24 mm Al), statistically different from the other groups. Epiphany SE (5.51 0.12 mm Al) showed intermediate values, followed by Hybrid Root SEAL (4.71 0.04 mm Al), which showed the inferior radiopacity (one-way ANOVA, post hoc Tukey-Kramer test, P < .05). In summary, AH Plus > Epiphany SE > Hybrid Root SEAL.

Statistical Analysis Five specimens from each group were tested, and the means were statistically compared. The Kolmogorov-Smirnov test showed that the results were consistent with a normal distribution curve. The parametric statistical analysis was performed (one-way analysis of variance [ANOVA] and post hoc Tukey-Kramer test), and the signicance level was set as 5% (GraphPad InStat; GraphPad Software Inc, San Diego, CA). SEM Examination For SEM examination, cylindrical Teon molds (3 4 mm) were lled with freshly mixed sealers. The molds were supported by a glass plate covered with a cellophane sheet and placed in a chamber (37  C, 95% relative humidity) for a period corresponding to 3 times the setting

TABLE 1. ANSI/ADA (2000) Standards, Mean Values, and Standard Deviations of Each Experimental Group Sealers Physicochemical properties
Setting time (min) Flow (mm) Radiopacity (mm Al) Dimensional change (%) Solubility (%)

AH Plus
579.00 4.95 36.76 3.04b 5.97 0.24a 1.69 0.31b 0.74 0.41a
a

Epiphany SE
24.40 4.39 34.43 2.17b 5.51 0.12b 2.22 0.41b 0.94 0.72a
b

Hybrid Root Seal

63.40 2.70c 45.94 0.61a 4.71 0.04c 3.45 0.34a 1.25 1.10b

The same superscripted letters indicate signicant statistical difference (comparison between lines, P < .05).

JOE Volume 36, Number 9, September 2010

Physicochemical Properties of Methacrylate Resinbased Root Canal Sealers

1533

Basic ResearchTechnology
TABLE 2. Levels of Metallic Ions, Expressed as mg/mL -1, Released in the Immersion Liquid of Samples Groups Metallic ions
Ca2+ K+ Ni2+ Zn2+

similar between themselves (one-way ANOVA, post hoc Tukey-Kramer test, P < .05). In summary, Hybrid Root SEAL > AH Plus = Epiphany SE.

AH Plus
43.22 11.39 0.58 0.36 <0.6 <0.2

Epiphany SE
7.80 3.43 0.32 0.26 <0.6 0.92 0.88

Hybrid Root SEAL

<0.5 0.74 0.10 <0.6 <0.2

Dimensional Change after Setting ANSI/ADA states that the maximum limit is 1% for linear shrinkage and 0.1% for expansion. The dimensional change of all sealers was greater than values considered acceptable by ANSI/ADA. Statistical analysis demonstrated higher dimensional change for Hybrid Root SEAL (3.45% 0.34%). AH Plus (1.69% 0.31%) and Epiphany SE (2.22% 0.41%) presented the lower results and were statistically

Solubility A root canal sealer should not exceed 3% by mass when the solubility of the set material is tested (ANSI/ADA 2000). Statistical analysis showed higher solubility for AH Plus (0.75% 0.41%) and Epiphany SE (0.94% 0.72%), which were statistically similar among themselves. The lower solubility was observed in Hybrid Root SEAL (1.25% 1.10%) group. The obtained results showed agreement with the ANSI/ADA standardization. Distilled and deionozed water used for the solubilty test was submitted to atomic absorption spectrometry analysis in order to check the amount of ions released from each sealer. Table 2 shows the ions released to the immersion liquid in each experimental group. It was veried that AH Plus sealer released signicant level of Ca2+ (43.22 11.39 mg mL1) when compared with the other groups. For K+ ions, the release was higher in Hybrid Root SEAL,

Figure 1. Internal portion of sealers after complete setting time. (A, B) AH Plus with original magnication of 350 and 3500, respectively. Note an irregular and compacted surface. (C, D) Epiphany SE with original magnication of 350 and 3500, respectively. Note the plate-shape monomers compacted in the sealer structure. (E, F) Hybrid Root SEAL with original magnication of 350 and 3500, respectively. Observe spherical monomers with inferior size when compared with AH Plus and Epiphany SE.

1534

Marin-Bauza et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
followed by AH Plus and Epiphany SE. Regarding Ni2+, all groups presented concentration inferior of 0.6 mg/L. For Zn2+, the largest value was observed for Epiphany SE, whereas AH Plus and Hybrid Root SEAL presented smaller concentrations than the smallest point of the analytical curve. ported ow values close to those obtained in this study for AH Plus and Epiphany sealers (7, 15). The highest ow obtained with Hybrid Root SEAL can be attributed to the presence of 4-META that has one hydrophilic and other hydrophobic radical (19), probably responsible for the increase of the sealer viscosity. SEM analysis showed that Hybrid Root SEAL is composed of small spherical particles that form a less compact and viscous structure than AH Plus and Epiphany SE, which might have contributed to the superior ow values achieved with Hybrid Root SEAL. A degree of radiopacity is indispensable for the control of root lling placement. Although the standards require only a lower limit to this property, it should be realized that extreme contrast in a material might lead to a false impression of a dense and homogenous ll (5). ANSI/ADA establishes that the radiopacity should be greater than or equal to 3 mm of aluminum. Although the radiopacity values of the sealers were statistically different among themselves, all groups were in agreement with ANSI/ADA specication. The difference in the results can be attributed to the radiopacifying agents of the studied sealers. AH Plus has calcium tungstate and zirconium oxide in its composition (18). Epiphany SE has bismuth oxychloride and barium borosilicate (7, 12), and Hybrid Root SEAL has only zirconium oxide in its formula (18). Bismuth oxychloride has higher radiopacifying effect, followed by zirconium oxide, calcium tungstate, and barium borosilicate (29). However, the resulting radiopacity depends on the chemical interaction among the radiopacifying agents (29). The results of this study can suggest that the zirconium oxide associated to the calcium tungstate in AH Plus produces superior radiopacity. Although Epiphany has bismuth oxychloride and barium borosilicate in its composition, the synergic effect did not guarantee the major radiopacity. Hybrid Root SEAL has only one radiopacifying agent and, therefore, was the least opaque material. Dimensional change demonstrates, in percentage, the shrinkage or expansion of the material after setting (5). ANSI/ADA states that the maximum limit is 1% for linear shrinkage and 0.1% for expansion. The dimensional change of all tested sealers was greater than the values considered acceptable by ANSI/ADA (AH Plus, 1.69%; Epiphany SE, 2.22%; and Hybrid Root SEAL, 3.46%). Previous investigations that evaluated the dimensional change of different sealers obtained values raging from 0.62%1.28% for AH Plus (7, 26) and 2.43%8.07% for Epiphany (7, 15). The results of this physicochemical property can be explained by water sorption after polymerization, a diffusioncontrolled process that occurs mainly in the resin matrix, because of its polar nature (30). It has been demonstrated that polymerized materials from mixtures of hydrophilic monomers had high water sorption (31). In this study, the presence of hydrophilic radicals in methacrylatebased sealers resulted in higher water sorption and, consequently, higher expansion (32). ANSI/ADA (2000) establishes that solubility of sealers should not exceed 3% by mass. Solubility results of all groups were within ANSI/ ADA recommendations (AH Plus, 0.75%; Epiphany SE, 0.94%; and Hybrid Root SEAL, 1.25%). Versiani et al (7) and Donnelly et al (33) found higher solubility values for Epiphany than those obtained in this research (3.41% and 4.02%, respectively). Resende et al (15) reported only 0.34% of solubility for Epiphany SE. Other authors (15, 26, 33) reported solubility values raging from 0.06%0.41% for AH Plus. The mass increase presented by Hybrid Root SEAL probably occurred as a result of the water absorption, because this sealer presented the larger dimensional alteration. Atomic absorption spectrometry analysis was conducted to determine the components released during the solubility test period. The analysis of the immersion liquid revealed highest calcium concentrations in the liquid of AH Plus. 1535

SEM
Fig. 1 illustrates the internal portion of sealers of each studied group. Photomicrographs of AH Plus revealed an irregular and compacted surface (Fig. 1A, B). Epiphany SE exhibited plate-shape monomers distributed in the sealer structure (Fig. 1C, D). Photomicrographs of Hybrid Root SEAL presented spherical monomers with inferior size when compared with the AH Plus and Epiphany SE sealers (Fig. 1E, F). At higher magnication (3500), polymeric chains of AH Plus and Epiphany SE seemed better compacted and organized than Hybrid Root SEAL (Fig. 1E).

Discussion
New dental materials intending to obtain a better clinical performance have been introduced on the market (25). On this concern, methacrylate resin-based root canal lling materials such as Epiphany and Hybrid Root SEAL were developed. Before the clinical use of these materials, it is necessary to perform several standardized tests to check their physiochemical and biological properties (5). In this study, the physiochemical tests were conducted on the basis of Specication 57 of ANSI/ADA (2000), following the modications proposed by Carvalho-Junior et al (26), which suggested a reduction of 80% in volume of the test samples dimensions, aiming to contribute to the rational use of endodontic materials without affecting the results. The setting time is primarily a control test on the stable behavior of a product and is dependent on the constituent components, their particle size, the ambient temperature, and relative humidity (3, 5). According to the ANSI/ADA standards, the setting time of a sealer can vary only 10% in relation to that established by the manufacturer. In the present study, only the setting time of the Hybrid Root SEAL did not conform to ANSI/ADA specications. AH Plus showed the highest mean values (579.00 minutes), followed by Hybrid Root SEAL (63.40 minutes) and Epiphany SE (24.40 minutes). These results agree with the literature reports that show a setting time between 494 and 817 minutes for AH Plus and between 23.10 and 25.03 minutes for Epiphany (7, 15). There is no report regarding the setting time of Hybrid Root SEAL, and the manufacturer informs that this sealer presents working time of 35 minutes and complete setting after 16 hours. AH Plus showed a setting time almost 9 times greater than Hybrid Root SEAL and 24 times greater than Epiphany SE as a result of the slow polymerization reaction of epoxy resin amines; hence the conversion of monomers into polymers occurs gradually (15, 27). Hybrid Root SEAL and Epiphany SE are dual-curable resin composites containing a catalyst component that accelerates the process (28). Another relevant aspect of the methodology is the fact that the methacrylate-based sealers were manipulated in a radiographic processing room to ensure that the sealer was not exposed to light during the experiment (10, 12, 15). The photo initiators present in Epiphany SE and Hybrid Root SEAL might start the polymerization process before their insertion into the molds. ANSI/ADA establishes as 20 mm the acceptable minimum value for the diameter of the disc formed by the sealer. In this study, the highest ow was presented in Hybrid Root SEAL group (45.94 mm), followed by AH Plus (36.76 mm) and Epiphany SE (34.43 mm); thus, all sealers were consistent with ANSI/ADA (2000). Previous investigations re-

JOE Volume 36, Number 9, September 2010

Physicochemical Properties of Methacrylate Resinbased Root Canal Sealers

Basic ResearchTechnology
Overall, this laboratory study disclosed that the physicochemical properties of the 3 tested sealers (Epiphany SE, Hybrid Root SEAL, and AH Plus) conformed to ANSI/ADA (2000) standardization, except for the setting time of Hybrid Root SEAL and the dimensional change of all sealers, which did not fulll the ANSI/ADA requirements. Nevertheless, it is important to highlight that these new resin-based lling materials have not had the same extensive evaluation that gutta-percha and conventional sealers have had. Further studies with these adhesive materials after longer periods of storage and under biologic conditions should be conducted.
14. Rached-Junior FJA, Souza-Gabriel AE, Alfredo E, Miranda CES, Silva-Sousa YTC, Sousa-Neto MD. Bond strength of Epiphany sealer prepared with resinous solvent. J Endod 2009;35:2515. 15. Resende LM, Rached-Junior FJA, Versiani MA, et al. A comparative study of physicochemical properties of AH Plus, Epiphany, and Epiphany SE root canal sealers. Int Endod J 2009;42:78593. 16. Pentron. Epiphany Soft Resin Endodontic Obturation System. Wallingford, CT: Pentron Clinical Technologies, LLC; 2007. 17. Lawson MS, Loushine B, Mai S, et al. Resistance of a 4-METAcontaining, methacrylate-based sealer to dislocation in root canals. J Endod 2008;34:8337. 18. Belli S, Ozcan E, Derinbay O, Eldeniz AU. A comparative evaluation of sealing ability of a new, self-etching, dual-curable sealer: Hybrid Root Seal (MetaSeal). Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:4552. 19. Chang JC, Hurst TL, Hart DA, Estey AW. 4-META use in dentistry: a literature review. J Prosthet Dent 2002;87:21624. 20. Van Landuyt K, Snauwaert J, De Munck JD, et al. Systematic review of the chemical composition of contemporary dental adhesives. Biomaterials 2007;28:375785. 21. Ames JM, Loushine RJ, Babb BR, et al. Contemporary methacrylate resin-based root canal sealers exhibit different degrees of ex vivo cytotoxicity when cured in their selfcured mode. J Endod 2009;35:2258. 22. Gillespie WT, Loushine RJ, Weller RN, et al. Improving the performance of EndoREZ root canal sealer with a dual-cured two-step self-etch adhesive II: apical and coronal seal. J Endod 2006;32:7715. 23. Perdigao J, Lopes MM, Gomes G. Interfacial adaptation of adhesive materials to root canal dentin. J Endod 2007;33:25963. 24. ANSI/ADA specication no. 57: endodontic sealing material, 2000. 25. Kim YK, Grandini S, Ames JM, et al. Critical review on methacrylate resin-based root canal sealers. J Endod 2010;36:38399. 26. Carvalho-Junior JR, Correr-Sobrinho L, Correr AB, Sinhoreti MA, Consani S, SousaNeto MD. Solubility and dimensional change after setting of root canal sealers: a proposal for smaller dimensions of test samples. J Endod 2007;33:11106. 27. Lin-Gibson S, Landis FA, Drzal PL. Combinatorial investigation of the structureproperties characterization of photopolymerized dimethacrylate networks. Biomaterials 2006;27:17117. 28. Pawinska M, Kierklo A, Marczuk-Kolada G. New technology in endodontics: the Resilon-Epiphany system for obturation of root canals. Adv Med Sci 2006;51:1547. 29. Bortoluzzi EA, Guerreiro-Tanomaru JM, Tanomaru-Filho M, Duarte MA. Radiographic effect of different radiopaciers on a potencial retrograde lling material. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108:62832. 30. Phillips RW, Skinner EW. Skinners science of dental materials. Philadelphia: Saunders; 1991:624. 31. Braden M, Clarke RL. Water absorption characteristics of dental microne composite lling materials: Iproprietary materials. Biomaterials 1984;5:36972. 32. Skrtic D, Antonucci JM. Dental composites based on amorphous calcium phosphate: resin composition/physicochemical properties study. J Biomaterials Applications 2007;21:37593. 33. Donnelly A, Sword J, Nishitani Y, et al. Water sorption and solubility of methacrylate resin-based root canal sealers. J Endod 2007;33:9904.

References
1. De-Deus G, Di Giorgi K, Fidel S, Fidel RA, Paciornik S. Push-out bond strength of Resilon/Epiphany and Resilon/Epiphany self-etch to root dentin. J Endod 2009; 35:104850. 2. Grossman LI. Solubility of root canal cements. J Dent Res 1978;57:927. 3. Gambarini G, Romeo U, Tucci E, et al. Cytotoxicity of Epiphany SE endodontic sealer: a comparative in vitro study. Med Sci Monit 2009;15:PI158. 4. Schroeder A. The impermeability of root canal lling material and rst demonstrations of new root lling materials. SSO Schweiz Monatsschr Zahnheilkd 1954;64:92131. 5. rstavik D. Materials used for root canal obturation: technical, biological and clinical testing. Endodontic Topics 2005;12:2538. 6. Sousa-Neto MD, Coelho FI, Marchesan MA, Alfredo E, Silva-Sousa YTC. Ex vivo study of the adhesion of an epoxy-based sealer to human dentine submitted to irradiation with Er: YAG and Nd: YAG. Int Endod J 2005;38:86670. 7. Versiani MA, Carvalho-Junior JR, Padilha MI, Lacey S, Pascon EA, Sousa-Neto MD. A comparative study of physicochemical properties of AH Plus and Epiphany root canal sealants. Int Endod J 2006;39:46471. 8. Bouillaguet S, Shaw L, Barthelemy J, Krejci I, Wataha JC. Long-term sealing ability of Pulp Canal Sealer, AH Plus, GuttaFlow and Epiphany. Int Endod J 2008;41:21926. 9. Shipper G, rstavik D, Teixeira FB, Trope M. An evaluation of microbial leakage in roots lled with a thermoplastic synthetic polymer-based root canal lling material (Resilon). J Endod 2004;30:3427. 10. Mathias-Junior O, Souza-Gabriel AE, Miranda CES, Pecora JD, Silva-Sousa YTC, Sousa-Neto MD. Solubility of Epiphany sealer prepared with resinous solvent. J Endod 2009;35:7158. 11. Tanomaru-Filho M, Jorge EG, Guerreiro-Tanomaru JM, Goncalves M. Radiopacity evaluation of new root canal lling materials by digitalization of images. J Endod 2007;33:24951. 12. Onay EO, Ungor M, Ozdemir BH. In vivo evaluation of the biocompatibility of a new resin-based obturation system. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;104:606. 13. Teixeira CS, Felippe WT. The effect of application time of EDTA and NaOCl on intracanal smear layer removal: a SEM analysis. J Endod 2004;38:28590.

1536

Marin-Bauza et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Effect of Calcium Phosphate Cements on Growth and Odontoblastic Differentiation in Human Dental Pulp Cells
Sun-Kyung Lee,* Sang-Kwang Lee, MSD,* Sang-Im Lee, MSD,* Jeong-Hui Park, MSD, Jun-Hyeog Jang, PhD, Hae-Won Kim, PhD, and Eun-Cheol Kim, DDS, PhD*
Abstract
Objective: Calcium phosphate cements (CPCs) are an interesting class of bone substitute materials. However, the biological effects of CPCs have not been well studied in human dental pulp cells (HDPCs). The purpose of this study was to investigate the effects of CPCs on the mechanical properties, growth, and odontoblastic differentiation in HDPCs compared with Portland cement (PC) and mineral trioxide aggregate (MTA). Methods: Experimental CPCs either containing chitosan (Ch-CPC) or without chitosan (CPC) were composed from the a-tricalcium phosphate powder. Setting time, compressive strength measurements, cell growth, alkaline phosphatase (ALP) activity, the levels of messenger RNA for differentiation-related genes, and mineralization of the HDPCs on various cements were assessed. Results: The setting time for CPC-Ch was 7.5 minutes, which was signicantly less than the 8.6 minutes for the CPC. On the seventh day of immersion, the compressive strength of CPC-CH reached 13.1 MPa, which was higher than 10.8 MPa of CPC. CPC and Ch-CPC-treated cells showed decreased cell proliferation but increased the levels of ALP activity, enhanced mineralized nodule formation, and upregulated odontoblastic markers messenger RNA including osteonectin, osteopontin, bone sialoprotein, dentin matrix protein-1, matrix extracellular phosphoglycoprotein, and dentin sialophosphoprotein (DSPP), compared with untreated control. The response of CPC and CP-CPC were similar to that of PC and MTA. However, the adhesion, growth, and differentiation in Ch-CPCtreated cells were similar to that in the CPC. Conclusion: CPC may be useful for pulp-capping applications based on its abilities to promote HDPC differentiation. (J Endod 2010;36:15371542)

Key Words
Calcium phosphate cement, chitosan, cytotoxicity, human dental pulp cell, odontoblastic differentiation

raditionally, different formulations of calcium hydroxide (CH) have been used for conservative pulp therapy (1). However, there are disadvantages associated with the use of CH, such as the presence of tunnels in the dentin bridge and the lack of adhesion and degradation after acid etching (2, 3). Recently, several materials have been proposed as candidates for direct pulp capping including mineral trioxide aggregate (MTA). MTA induces hard tissue repair of the exposed pulp in experimental animals and promotes dentin bridge formation (4, 5). Previously, we reported that the mean thickness of the dentin bridges observed in the MTA group was statistically greater than that of the CH group (6). In addition, significantly greater immunostaining for dentin sialoprotein (DSP) and heme oxygenase-1 (HO-1) was observed in the MTA group than in the CH group (6). Moreover, we showed that the combination of MTA and an enamel matrix derivative promotes more rapid differentiation of human dental pulp cells (HDPCs) than MTA alone (7). In addition, we reported that radiopaque Portland cement (PC) is biocompatible and facilitates odontoblastic differentiation of HDPCs (8). Although MTA is considered to be superior to CH, the hard tissueforming mechanism of MTA is basically similar to that of CH, which is known to cause inammatory and necrotic changes in the subjacent pulp (9). An alternative pulp-capping material to CH is self-setting calcium phosphate cement (CPC). CPC is a bioactive material that is available in powder and liquid forms, which, when mixed, primarily form hydroxyapatite (10). Self-setting CPC materials are widely used for orthopedic and dental applications and have the potential to stimulate osteogenesis (11). The self-setting, moderate compressive strength even in loading sites and highly biocompatibility properties (12) suggest that CPC is superior to pure CH, which means that this material may have applications in pulp capping to induce reparative dentin formation or as a lining material (13). Chitosan is a candidate material for biomedical applications because of its distinctive biological properties, which include good biodegradability, biocompatibility, and osteoconductivity (14). In the study of Xu et al (15), chitosan was incorporated into CPC, thereby increasing the exural strength, toughness, and strain to failure of the material (15). This CPC-chitosan scaffold was found to be biocompatible and to support the adhesion and proliferation of osteoblast cells (16). The addition of chitosan strengthened the CPC and increased alkaline phosphatase (ALP) activity in bone marrow stem cells (17). However, it has not been investigated whether HDPCs can differentiate into odontoblast-like cells when exposed to CPC or chitosansupplemented CPC (CPC-Ch). The aim of this study was to evaluate the mechanical and biological effects including setting time, compressive strength, surface morphology, cell adhesion,

From the *Department of Oral and Maxillofacial Pathology and Institute of Wonkwang Biomaterials Implant, School of Dentistry, Wonkwang University, Iksan, Korea; Department of Nanobiomedical Science and WCU Research Center, Dankook University, Cheonan, Korea; Department of Biochemistry, School of Medicine, School of Medicine, Inha University, Incheon, Korea; and Department of Biomaterials Science, School of Dentistry and Institute of Tissue Regeneration Engineering (ITREN), Dankook University, Cheonan, Korea. Hae-Won Kim and Eun-Cheol Kim contributed equally to this article. Supported by a grant from the Korea Healthcare technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A084458). Address requests for reprints to Dr Eun-Cheol Kim, Department of Oral and Maxillofacial Pathology, College of Dentistry, Wonkwang University, 344-2 Shinyong, Iksan, South Korea 570-749. E-mail address: eckwkop@wonkwang.ac.kr. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.04.027

JOE Volume 36, Number 9, September 2010

Effect of CPCs on HDPCs

1537

Basic ResearchTechnology
proliferation, and odontoblastic differentiation of CPC and CPC-Ch on HDPCs and to compare the results with MTA and PC.

Materials and Methods

Preparation of Experimental Cements The CPC powder a-tricalcium phosphate (TCP) was produced by the reaction of CaCO3 and dicalcium phosphate anhydrous at 1,400 C and followed by ball milling and sieving down to 45 mm. Commercial hydroxyapatite (Alfa Aesar, Ward Hill, MA) powder was added at 2 wt% to enhance the crystallization reaction. As the liquid phase, 5% wt Na2HPO4 with or without 2% chitosan solution (Sigma-Aldrich Chemical Co, St Louis, MO) was prepared. The setting reaction was made with the powder-to-liquid ratio (mixing ratio of powder with respect to liquid) of 3.0, and the mixture was stored at 37 C in 100% humidity chamber for different periods of time. The setting time of the cements was measured by means of Gillmore needle test according to American Society for Testing and Materials C266-99-A standard (18). Phase change of the cements during soaking in saline solution was characterized with x-ray diffraction (Shimadzu Co, Kyoto, Japan). Compressive strength of the samples was measured using a mechanical testing machine (Instron 3344; Instron Corp, Canton, MA). Five total specimens were tested under each condition (as hardened and stored for 7 days in humidity chamber) were tested. Morphology of the specimens was observed by scanning electron microscopy (Hitachi Co, Tokyo, Japan). Under aseptic conditions, white MTA (ProRoot; Dentsply, Tulsa, OK) and PC (Ssangyong, Seoul, Korea) were mixed according to the manufacturers instruction. Each sample (diameter, 6 mm; thickness, 2 mm) was allowed to set for 24 hours at 37 C in 100% humidity. Cell Culture We used the HDPCs lines immortalized by transfection with the telomerase catalytic subunit hTERT gene (19). Cells were cultured in Dulbeccos Modied Eagles Medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 mg/mL streptomycin in a humidied atmosphere of 5% CO2 at 37 C. For mineralization experiments, cells were cultured in 50 mg/mL ascorbic acid (AA) and 10 mmol/L b-glycerophosphate (bGP)-containing media as described previously (20). The cements were all placed at the bottom of 24well tissue culture plates, washed twice with phosphate buffer solution, dried under laminar ow for 24 hours at room temperature, and sterilized by gamma-radiation with 37.2 Gray before being used to culture cells. HDPCs were seeded at 1 105 cells per well on the prepared cements and cultured for up to 14 days.

Cell Adhesion and Viability At each culturing period (1, 7, and 14 days), the cells were harvested, and the cell viability was measured as the mitochondrial nicotinamide adenine dinucleotide and nicotinamide adenine dinucleotide phosphatedependent dehydrogenase activity using a cell proliferation assay kit (CellTiter 96 Aqueous One Solution; Promega, Madison, WI). A colorimetric measurement was performed using a microplate reader at an absorbance of 490 nm. The cell growth morphology was observed by scanning electron microscopy at an accelerating voltage of 15 kV after xing with glutaraldehyde (2.5%), dehydrating with a graded series of ethanol (75%, 90%, 95%, and 100%), and treating with hexamethyldisilazane and platinum coating. ALP Activity Incubated HDPCs were washed with PBS and then sonicated with a cell disruptor. ALP activity was measured using p-nitrophenyl phosphate (3 mmol/L nal concentration) as the substrate in 0.7 mol/L 2-amino-methyl-1-propanol, pH of 10.3, and 6.7 mmol/L MgCl2. Absorbance was measured at 410 nm using an enzyme-linked immunosorbent assay reader (Beckman Coulter, Fullerton, CA). Alizarin Red Staining Cells were cultured in DMEM containing 10% FBS, 50 mg/mL AA, and 10 mmol/L bGP in the absence or presence of cements for 14 days for mineralized nodule assay. The culture medium was replaced with fresh culture medium every 2 days. After 14 days of treatment, the calcium deposition of HDPCs was studied using 0.1% alizarin red S staining solution (Sigma-Aldrich, St. Louis, MO). The samples were xed with 70% ice-cold ethanol for 1 hour, rinsed twice with PBS, and stained with 40 mmol/L alizarin red solution for 10 minutes under conditions of gentle agitation. The pictures of alizarin red S staining were photographed under light microscopy. Reverse Transcription-Polymerase Chain Reaction The total RNA of pulp cells was extracted using the Trizol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturers instruction. Then, 1 mg RNA was reverse-transcribed for rst-strand complementary DNA (cDNA) synthesis (Gibco BRL, Rockville, MD). The cDNA was amplied in a nal volume of 20 mL containing 2.5 mmol/L magnesium dichloride, 1.25 U Ex Taq polymerase (Bioneer, Daejeon, Korea) and 1 mmol/L specic primers. Amplication was performed for 30 cycles in a DNA thermal cycler. Primer sequences for differentiation markers are detailed in Table 1. The PCR products were resolved on a 1.5% agarose gel and stained with ethidium

TABLE 1. Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) Primers Sequence Gene

Osteonectin (ON) Osteopontin (OPN) Bone sialoprotein (BSP) Dentin sialophosphoprotein (DSPP) Dentin-matrix protein-1 (DMP-1) MEPE GAPDH

Sequence (5 -3)
Forward: ACATGGGTGGACACGG Reverse: CCAACAGCCTAATGTGAA Forward: CCAAGTAAGTCCAACGAAAG Reverse: GGTGATGTCCTCGTCTGTA Forward: TGGAGATGACAGTTCAGAAG Reverse: GTACTGGTGCCGTTTATGC Forward: CAGTGATGAATCTAATGGC Reverse:CTGATTTGCTGCTGTCTGAC Forward:CAGGAGCACAGGAAAAGGAG Reverse:CTGGTGGTATCTTCCCCCAGGAG Forward: ATGCGAGTTTTCTGTGTGG Reverse: GTTTTCTTCCCCCAGGAG Forward:CGGAGTCAACGGATTTGGTCGTAT Reverse:AGCCTTCTCCAGGTGGTGAAGAC

Size (bp)
405 347 333 488 213 503 306

1538

Lee et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 2. Summary of Characteristics of the Experimental Calcium Phosphate Cements Used in This Study (CPC, CPC-Ch, PC and MTA) CPC
Starting powder Setting time (ave. min) Crystalline phase* Compressive strength (MPa) a-TCP 8.6 Hydroxyapatite 10.8 3.0

CPC-Ch
a-TCP + chitosan 7.5 Hydroxyapatite 13.1 2.1

PC
Calcium silicates 468 19.5 30.1

MTA
Calcium silicates, bismuth oxide 165 40.0

CPC, calcium phosphate cement; CPC-Ch, CPC with chitosan; PC, Portland cement; and MTA, mineral trioxide aggregate. *Newly formed after soaking for 7 days in saline. After 24 h of immersion in water.

bromide. The intensity of each band after normalization with glyceraldehyde 3-phosphate dehydrogenase messenger RNA was quantied on the photographed gels with a densitometer (Quantity One; Bio-Rad, Hercules, CA).

Statistical Analysis Data are presented as mean standard deviation. Continuous variables in the present study met the criteria for a normal distribution and were assumed to be parametric. The statistical analysis of the data was performed by one-way analysis of variance followed by a multiplecomparison Tukey test with the use of an SPSS program (SPSS 12.0; SPSS GmbH, Munich, Germany). Statistical signicance was determined at p < 0.05.

within 10 minutes of mixing with basic solvent, in a process driven by the acid-basic reaction, the PC and MTA have very long setting times, which have been concerned as the major drawback of the PC and MTA (21). The compressive strengths of CPCs were approximately 10 to 13 MPa, which were signicantly increased up to 20 to 25 MPa after incubation in water. The PC and MTA have compressive strengths in similar ranges of CPCs that were also increased when immersed in water. Figure 1 shows the SEM surface morphologies of the CPC and CPCCh specimens after soaking in PBS for 7 days. The surfaces of both cements showed similar morphologies with numerous ultrane crystallites in a rosette form, which is typical of the apatite crystals observed in CPC.

Results
Setting Time, Compressive Strength, and Surface Morphology of the Cements The characteristics of the experimental cements, CPC and CPC-Ch, are summarized in Table 2. Data on both PC and MTA are also presented for comparison (21, 22). Although both experimental CPCs hardened

Effects of Cements on Cell Viability and Morphology The levels of cell growth, as measured by the (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium, inner salt) assay, are represented in Figure 2A. It was observed that all the materials including MTA, PC, CPC, and CPC-Ch showed cytotoxicity against HDPCs when the cells were incubated at 24 hours. In contrast, the MTA and PC did not show signicant cytotoxicity against HDPCs at 14 days, whereas CPC and CPC-Ch showed

Figure 1. Scanning electron microscopic morphology of the (A) CPC and (B and C) chitosan CPC. Results were reproduced in two separate experiments.

JOE Volume 36, Number 9, September 2010

Effect of CPCs on HDPCs

1539

Basic ResearchTechnology
12 0 100

1 day

7 days

14 days

% of control

80 60 40 20 0 control

* *

* *

* *

* *

M TA

PC

CPC

CPC-ch

cements
MTA PC

CPC

CPC-ch

Figure 2. The effects of cements on (A) cell viability and (B) adhesion by the MTS method and SEM, respectively. )Statistically signicant difference as compared with control, p < 0.05. SEM micrographs of the HDPCs grown on the MTA, PC, CPC, and CPC cements for 7 days of culture. Results (n = 5) were reproduced in three separate experiments.

cytotoxicity at 1, 7, and 14 days. However, there was no signicant difference in cell viability between CPC and CPC-Ch for 1, 7, and 14 days. To investigate the inuence of the cements on cell spreading, SEM investigations of adhesion were performed (Fig. 2B). The HDPCs on all samples showed an elongated/spindle-shaped morphology. There were no differences in spreading among the MTA, PC, CPC, and CPC-Ch groups.

Effects of Cements on the Odontoblastic Differentiation of HDPCs To investigate the odontoblast-like differentiation of HDPCs treated with the cements, we assessed the levels of ALP activity, alizarin red staining, and messenger RNA expression of differentiation markers such as osteonectin (ON), osteopontin (OPN), bone sialoprotein (BSP), dentin matrix protein-1 (DMP-1), matrix extracellular

1540

Lee et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Figure 3. The effects of cements on (A) ALP activity; (B) the formation of calcication nodules; and (c) the expression of ON, OPN, BSP, MEPE, DMP-1 and DSPP messenger RNA in HDPCs. HDPCs were cultured in DMEM with 0.05 mmol/L AA and 10 mmol/L bGP in the presence of indicated kinds of cements for 14 days. The cells treated with bGP and AA were used as a negative control. Representative ALP data were the average of ve replicates. After 14 days, representative photographs of alizarin red S staining are shown (original magnication, 200). The RT-PCR data represent one of three independent experiments with similar results. (D) The relative density of cytokine mRNA to GAPDH messenger RNA. Levels of messenger RNA genes were measured by densitometry. The relative level of gene expression was normalized against the GAPDH messenger RNA signal, and the control was set as 1.0. Relative density values represent the mean standard deviation of three independent experiments. )Signicant difference from control (p < 0.05).
JOE Volume 36, Number 9, September 2010 Effect of CPCs on HDPCs

1541

Basic ResearchTechnology
phosphoglycoprotein (MEPE), and dentin sialophosphoprotein (DSPP) (23). Observation from alizarin red S staining at 14 days showed a strong positive calcium deposition in HDPCs exposed of all kind of cements, whereas the stain was not obvious in control cells. As shown in Figure 3A and B, the ALP activity and calcium deposition of the CPC-Chtreated group was similar to that of the CPC, PC, and MTA group. CPC and CPC-Ch increased the messenger RNA expression levels of the odontogenic genes DMP-1, DSPP, MEPE, ON, OPN, and BSP as compared with the control treatment at 1, 7, and 14 days of cultivation. In contrast, no signicant differences in the levels of mRNA expression for markers genes were observed between the CPC and CPC-Ch groups (Fig. 3C and D).
3. Cox CF, Hafez AA, Akimoto N, et al. Biocompatibility of primer, adhesive and resin composite systems on non-exposed and exposed pulps of non-human primate teeth. Am J Dent 1998;11:S5563. 4. Pitt Ford TR, Torabinejad M, Abedi HR, et al. Using mineral trioxide aggregate as a pulp-capping material. J Am Dent Assoc 1996;127:14914. 5. Faraco IM Jr, Holland R. Response of the pulp of dogs to capping with mineral trioxide aggregate or a calcium hydroxide cement. Dent Traumatol 2001;17:1636. 6. Min KS, Park HJ, Lee SK, et al. Effect of mineral trioxide aggregate on dentin bridge formation and expression of dentin sialoprotein and heme oxygenase-1 in human dental pulp. J Endod 2008;34:66670. 7. Min KS, Yang SH, Kim EC. The combined effect of mineral trioxide aggregate and enamel matrix derivative on odontoblastic differentiation in human dental pulp cells. J Endod 2009;35:84751. 8. Min KS, Lee SI, Lee Y, et al. Effect of radiopaque Portland cement on mineralization in human dental pulp cells. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108:e826. 9. Kuratate M, Yoshiba K, Shigetani Y, et al. Immunohistochemical analysis of nestin, osteopontin, and proliferating cells in the reparative process of exposed dental pulp capped with mineral trioxide aggregate. J Endod 2008;34:9704. 10. Dickens-Venz SH, Takagi S, Chow LC, et al. Physical and chemical properties of resin-reinforced calcium phosphate cements. Dent Mater 1994;10:1006. 11. Ruhe PQ, Hedberg-Dirk EL, Padron NT, et al. Porous poly (DL-lactic-co-glycolic acid)/calcium phosphate cement composite for reconstruction of bone defects. Tissue Eng 2006;12:789800. 12. Lye KW, Tideman H, Merkx MA, et al. Bone cements and their potential use in a mandibular endoprosthesis. Tissue Eng Part B Rev 2009;15:48596. 13. Chaung HM, Hong CH, Chiang CP, et al. Comparison of calcium phosphate cement mixture and pure calcium hydroxide as direct pulp-capping agents. J Formos Med Assoc 1996;95:54550. 14. Muzzarelli RAA, Biagini G, Bellardini M, et al. Osteoconduction exerted by methylpyrolidinone chitosan in dental surgery. Biomaterials 1993;14:3943. 15. Xu HHK, Quinn JB, Takagi S, et al. Processing and properties of strong and non-rigid calcium phosphate cement. J Dent Res 2002;81:21924. 16. Xu HHK, Simon CG Jr. Fast setting calcium phosphate-chitosan scaffold: mechanical properties and biocompatibility. Biomaterials 2005;26:133748. 17. Moreau JL, Xu HH. Mesenchymal stem cell proliferation and differentiation on an injectable calcium phosphate-chitosan composite scaffold. Biomaterials 2009;30: 267582. 18. Am Society for Testing and Materials. Standard test method for time of setting of hydraulic-cement paste by Gillmore needles. Annual book of ASTM standards. Conshohocken, PA: American Society for Testing and Materials; 2006;ASTM C26603. 19. Kitagawa M, Ueda H, Iizuka S, et al. Immortalization and characterization of human dental pulp cells with odontoblastic differentiation. Arch Oral Biol 2007;52:72731. 20. Yasuda Y, Ogawa M, Arakawa T, et al. The effect of mineral trioxide aggregate on the mineralization ability of rat dental pulp cells: an in vitro study. J Endod 2008;34: 105760. 21. Parirokh M, Torabinejad M. Mineral trioxide aggregate: a comprehensive literature review-part I: chemical, physical, and antibacterial properties. J Endod 2010;36: 40013. 22. Camilleri J. The physical properties of accelerated Portland cement for endodontic use. Int Endod J 2008;41:1517. 23. DSouza RN, Bachman T, Baumgardner KR, et al. Characterization of cellular responses involved in reparative dentinogenesis in rat molars. J Dent Res 1995; 74:7029. 24. Kouassi M, Michalesco P, Lacoste-Armynot A, et al. Antibacterial effect of a hydraulic calcium phosphate cement for dental applications. J Endod 2003;29:1003. 25. Yang SE, Baek SH, Lee W, et al. In vitro evaluation of the sealing ability of newly developed calcium phosphate-based root canal sealer. J Endod 2007;33:97881. 26. Lian Q, Li DC, He JK, et al. Mechanical properties and in-vivo performance of calcium phosphate cement-chitosan bre composite. Proc Inst Mech Eng H 2008;222:34753. 27. Pan Z, Jiang P, Fan Q, et al. Mechanical and biocompatible inuences of chitosan ber and gelatin on calcium phosphate cement. J Biomed Mater Res B Appl Biomater 2007;82:24652. 28. Khashaba RM, Chutkan NB, Borke JL. Comparative study of biocompatibility of newly developed calcium phosphate-based root canal sealers on broblasts derived from primary human gingiva and a mouse L929 cell line. Int Endod J 2009;42:7118. 29. Dagang G, Kewei X, Haoliang S, et al. Physicochemical properties of TTCP/DCPA system cement formed in physiological saline solution and its cytotoxicity. J Biomed Mater Res A 2006;77:31323. 30. Matsuya S, Takagi S, Chow LC. Effect of mixing ratio and pH on the reaction between Ca4(PO4)2O and CaHPO4. J Mater Sci Mater Med 2000;11:30511.

Discussion
Although CPCs are attractive materials for bone reconstruction (eg, alveolar ridge augmentation and craniofacial recovery) (14), their utility in the regeneration of dental pulp-dentin complex tissue has not been well documented. Bacterial growth inhibition and better mechanical properties than CH was observed for hydraulic CPC (24). In addition, newly developed CPC-based root canal sealers were well adapted to the canal wall and inltrated into the dentinal tubules (25). As the aTCPbased CPC used in the present study is cell compatible and stimulates osteoblastic differentiation and mineralization (10), it is expected to facilitate the growth and odontoblastic differentiation of pulp cells. Furthermore, the addition of chitosan to CPC, which is known to enhance mechanical properties (26, 27), is also required to address the efcacy in biological roles in dentin regeneration. Mechanical strength is a critical parameter for biomaterials used in tooth repair or pulp capping therapy. The compressive strength of the experimental CPCs was increased from 10.8 MPa to 13.1 MPa by the addition of chitosan, and this increase is attributed to the strengthening effect of the transformed apatite nanocrystallites, which interlock to form a more compact network than that formed by the original aTCPbased structure. We tested the experimental CPCs for their effects on HDPC proliferation and odontogenesis. We found that CPC and CPC-Ch were cytotoxic for HDPCs at 1, 7, and 14 days of cultivation. These results are supported by the data of Khashaba et al (28) who reported that CPC decreased the proliferation of human gingival broblasts and a mouse broblast cell line. In the CPC setting, TTCP (Ca4[PO4]2O) and dicalcium phosphate anhydrous (CaHPO4) dissolved in the liquid as Ca2+, PO43, and OH ions, which then reprecipitated to form hydroxyapatite (29). It seems likely that the ionic activities and pH changes that occurred during setting were responsible for the observed cytotoxicities (30). In our study, CPC and CPC-Ch promoted odontoblastic differentiation of HDPCs, as evidenced by the formation of mineralized nodules, the induction of ALP activity in the early stage, and the up-regulation of odontoblastic markers; these results are consistent with other studies in mesenchymal stem cells (17). But chitosan did not enhance adhesion, growth and differentiation of HDPCs on CPC. These contrary results may be due to different sources of chitosan. In summary, the study reports for the rst time that CPC facilitates the growth and differentiation of HDPCs, whereas these effects of CPC were not affected by chitosan.

References
1. Pereira JC, Segala AD, Costa CAS. Human pulpal response to direct pulp capping with an adhesive system. Am J Dent 2000;13:13947. 2. Cox CF, Subay RK, Ostro E, et al. Tunnel defects in dentin bridges: their formation following direct pulp capping. Oper Dent 1996;21:411.

1542

Lee et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Effect of Cement Type, Relining Procedure, and Length of Cementation on Pull-out Bond Strength of Fiber Posts
Vanessa Cruz Macedo, MD, DDS,* Andre Luis Faria e Silva, PhD, MD, DDS, and Luis Roberto Marcondes Martins, PhD, MD, DDS
Abstract
Introduction: As opposed to the cementation metal posts, the cementation of ber posts has several details that can signicantly inuence the success of post retention. This study evaluated the effect of the relining procedure, the cement type, and the luted length of the post on ber posts retention. Methods: One hundred eighty bovine incisors were selected to assess post retention; after endodontic treatment, the canals were ared with diamonds burs. Post holes were prepared in lengths of 5, 7.5, and 10 mm; the ber posts were relined with composite resin and luted with RelyX ARC, RelyX Unicem, or RelyX Luting 2. All cements are manufactured by 3M ESPE (St. Paul, MN). Samples were subjected to a pull-out bond strength test in a universal testing machine; the results (N) were submitted to a three-way analysis of variance and the Tukey post hoc test (a = 0.05). Results: The improvement of post retention occurred with the increase of the post length luted into the root canal; the relining procedure improved the pull-out bond strength. RelyX Unicem and RelyX ARC showed similar values of retention, both showing higher values than RelyX Luting 2. Conclusion: Post length, the relining procedure, and the cement type are all important factors for improving the retention of ber posts. (J Endod 2010;36:15431546)

Key Words
Bond strength, ber post, resin cement

ntraradicular posts are commonly used to restore endodontically treated teeth when their remaining coronal tissue can no longer provide adequate support and retention for the restoration (1, 2). For decades, endodontically treated teeth have been restored using cast metal posts; despite their high retention and thin resulting cement lm, these conventional posts have a high elastic modulus and can lead to root fractures (3, 4). On the other hand, the similar elastic modulus of ber posts, resin cements, and dentin is considered to be advantageous for improving the performance of restorations (5, 6). In contrast to rigid posts, ber posts do not need to be inserted to a length equal to or longer than the depth of the clinical crown in order to reduce the chance of root fracture (7). This is advantageous for short roots or for roots presenting a high degree of curvature. Once the ber posts are adhesively luted into the canal, a high in-depth insertion into the root canal is not necessary to improve retention (8). This is feasible because the bonding of the cement to dentin walls is more effective in the cervical region than in the apical region (9, 10). Despite the cited advantages, the mismatch between the diameters of the post space and the ber post remains a clinical problem (11, 12); prefabricated posts do not t well into elliptical-shaped canals (13) or ared canals that result from carious extension, trauma, pulpal pathosis, or iatrogenic misadventure (11). In such cases, if the post does not t well, the layer of resin cement might be excessively thick, favoring the formation of air bubbles and predisposing the post to debonding (14). One solution for this issue is to reline the ber post with resin composite (15); customizing the post increases its adaptation to the root walls and reduces the thickness of the resin cement (14). Improving the contact between the post and the canal walls may also reduce the dependence on the bonding for retention (16). It has been shown that cements with lower bonding potential but other favorable mechanical properties may be useful in luting relined ber posts (17). Therefore, the aim of this study was to evaluate the effect of ber post relining with resin composite, cement type used for luting, and post-hole length on the pull-out bond strength of ber posts luted to root canals.

Materials and Methods

From the *Department of Dental Materials and Prosthodon tics, Sao Jose dos Campos Dental School, Sao Paulo State University, Sao Jose dos Campos, SP, Brazil; Department of Dentistry, Federal University of Sergipe, Aracaju, SE, Brazil; and Department of Restorative Dentistry, Piracicaba Dental School, University of Campinas, Piracicaba, SP, Brazil. Address requests for reprints to Andre Luis Faria e Silva, Departamento de Odontologia, Campus da Saude Prof Joao Cardoso Nascimento Junior, Rua Claudio Batista S/N, Bairro Sanatorio, Aracaju/SE, Brazil 49060-100. E-mail address: andrelfsilva@hotmail.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.04.014

One hundred eighty recently extracted bovine incisors with similar root sizes and lengths were selected for this study. The crowns were removed above the cementumenamel junction with a low-speed diamond saw in order to obtain a remaining root height of 16 mm. For the endodontic treatment, a step-back preparation technique was used; to are the coronal and middle third of the canal, a #2 Gates Glidden drill was inserted several millimeters into the canal. This was repeated with #3 and #4 drills. Apical preparation was conducted with the nal master apical le of size 40; aring of the canal was completed with ling to size 70. All enlargement procedures were followed by irrigation with 2.5% sodium hypochlorite solution. The prepared root canals were obturated with gutta-percha cones using the lateral condensation technique and Sealer-26 resin sealer (Dentsply Ind Com Ltda, Petropolis, RJ, Brazil). The specimens were stored in 100% humidity for at least 72 hours to allow for the resin sealer to set. After the storage period, the coronal gutta-percha was removed with a hot Rhein instrument at 5-, 7.5-, and 10-mm depth. In order to obtain a standardized ared canal, the canals were enlarged using #4138 and 4137 high-speed diamond burs (KG Soren sen Ind e Com LTDA, Sao Paulo, SP, Brazil) under water irrigation. Three post space depths were prepared: 5, 7.5, and 10 mm. In order to facilitate handling, the roots

JOE Volume 36, Number 9, September 2010

Pull-out Strength of Fiber Posts

1543

Basic ResearchTechnology
TABLE 1. Description of Cements Used in This Study and the Application Protocols Cement
RelyX ARC

Classication
Resin cement

Application protocol
The canal walls were etched with 35% phosphoric acid for 15 seconds, rinsed for 15 seconds, and gently air dried. Excess water was removed from the post space with absorbent paper points. The Scotchbond Multipurpose Plus Activator (3M ESPE, St Paul, MN) was applied into the root canal with a microbrush of compatible size and air dried for 5 seconds. Afterwards, the Scotchbond Multipurpose Plus Primer (3M ESPE) and also air dried. The dual-cured resin cement RelyX ARC was mixed and placed over the posts, which was inserted into the root canal with light pressure. The excess of luting material was removed and light activation was performed for 40 seconds. The root canal walls were rinsed with water using a syringe and then gently dried with paper points. The cement capsule was activated for 2 seconds and mixed automatically in a high-speed triturator for 10 seconds. Afterwards, the resin cement was applied into the root canals by means of Elongation Tip (3M ESPE). The post was then seated in the root canal; the excess resin was subsequently removed. The light activation was performed for 40 seconds. The root canal walls are cleaned and dried as described for RelyX Unicem. The cement was mixed for 10 s inserted into the canal with 50 K-le. The post was inserted into the canal with light pressure and stabilized during 2 min. Then, the excess material was removed.

RelyX Unicem

Self-adhesive resin cement

RelyX Luting

Resin-modied glass ionomer

were embedded in polystyrene resin blocks. Parallelism between the post and resin block was obtained using a parallel meter. The prepared root canals received either relined or nonrelined ber posts; a 1.5-mm diameter glass ber-reinforced epoxy post system (Reforpost; Angelus, Londrina, Brazil) was used. The Adper Single Bond 2 adhesive system (3M ESPE, St Paul, MN) was applied and light cured over the previously silanized post before the cementation procedure. For the relining procedure, the ber post was treated as previously described. Afterwards, the canal walls were lubricated with a hydrosoluble gel; the ber post was covered with resin composite Filtek Z250 (3M ESPE) and inserted into the canal. The resin composite was light cured for 20 seconds, the relined ber post was removed, and the resin composite was light cured again for 20 seconds. Copious rinsing removed the lubricant gel from the root canal. The cements used for cementation and the details of the luting procedures are described in Table 1. The specimens were stored in 100% humidity at 37 C for 24 hours before testing. The pull-out test was performed parallel to the long axis of both the post and the tooth at a cross-head speed of 0.5 mm/min using a universal testing machine (Model 4411; Instron Inc, High Wycombe, UK). The force required to dislodge each post was then recorded in newtons. The independent variables tested in this study were the post relining, the cement type, and the depth of luting. Statistical analysis was performed by applying a three-way analysis of variance followed by the Tukey post hoc test at a 95% condence level.

Tukey test are shown in Tables 2 and 3. Because statistical signicances were detected only for the factors observed individually, comparisons were performed only for the pooled average. Relined posts presented signicantly higher pull-out bond strengths than nonrelined posts. Regarding the depth of post cementation, posts that were luted at a 10-mm depth presented the highest values followed by those luted at a 7.5-mm depth. RelyX ARC and RelyX Unicem showed similar results for retention, both showing signicantly higher bond strengths compared with RelyX Luting 2.

Discussion
When restoring root-lled teeth, clinicians are opting for materials that have an elastic modulus similar to that of dentin, that are capable of creating homogenous stress distribution, and that are able to decrease the incidence of catastrophic root fractures (18, 19). Fiber posts have elastic moduli similar to that of dentin and have the ability to bond to this substrate by the use of adhesive cements. Thus, it has been shown that it is not necessary to increase the depth of ber post cementation in order to improve fracture resistance (7). However, the results of the current study show that, irrespective of the type of cement used, increasing the depth of cementation is benecial to improve the post retention. Several studies have shown that the bond strength of resin cements to root canals is effective in the cervical third but weak in the apical third (11, 2022). Thus, it is reasonable to assume that the bond to the cervical third of the root canal could be enough to promote proper post retention. Regardless of the low bond strength in the apical regions, the frictional retention in these areas contributes to the dislocation resistance of the ber post (23, 24). Frictional retention is directly proportional to the contact area (the larger the contact surfaces, the better the retention). This explains the results of this

Results
Analysis of variance revealed statistically signicant differences for all factors (p < 0.01); however, there was no signicant effect for any interaction between the factors (p > 0.05). Comparisons using the

TABLE 2. Bond Strength Means (Standard Deviations) for All Cements and Post Types (N) Cements Post type
Non-relined post Relined post Pooled average

RelyX ARC
245.1 (74.5) 336.7 (91.6) 290.9 (94.8) A

RelyX Unicem
208.4 (83.8) 362.4 (87.8) 285.4 (115.2) A

RelyX Luting 2
101.7 (57.0) 147.1 (71.0) 124.4 (67.8) B

Pooled average
185.1 (94.3) B 282.1 (127.3) A

For the pooled averages, means followed by distinct letters are signicantly different (Tukey test, 95% condence level).

1544

Macedo et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 3. Bond Strength Means (Standard Deviations) for All Post Lengths and Post Types (N) Post length Post type
Non-relined post Relined post Pooled average

Cement
RelyX ARC RelyX Unicem RelyX Luting RelyX ARC RelyX Unicem RelyX Luting

5 mm
199.9 (72.3) 146.6 (46.5) 44.1 (10.8) 318.3 (67.6) 304.8 (57.9) 78.0 (23.9) 181.9 (116.1) C

7.5 mm
254.4 (65.1) 213.2 (91.0) 110.3 (34.3) 300.3 (36.4) 387.5 (84.1) 173.5 (48.0) 239.9 (108.6) B

10 mm
281.1 (67.8) 265.8 (65.0) 150.8 (52.2) 391.5 (126.4) 394.9 (94.1) 189.8 (72.4) 279.0 (122.2) A

Overall for cement by post type

245.1 (74.5) A 208.4 (83.8) A 101.7 (57.0) B 336.7 (91.6) A 362.4 (87.8) A 147.1 (71.0) B

Pooled average
185.1 (94.3) B 282.1 (127.3) A

For the pooled averages, means followed by distinct letters are signicantly different (Tukey test, 95% condence level).

study; ber posts luted deeper provided the highest pull-out bond strengths. In addition to the contact area, closer contact between cement type and dentin is also important in order to improve the frictional retention of the post. In this study, the relining procedure increased the post retention to the root canal. A higher post-to-root canal adaptation increases the sustained pressure during cementation (16). The application of sustained pressure results in better contact between the cement/post assembly and the dentin (25). This results in higher frictional retention compared with nonrelined posts and consequently in increased pull-out bond strengths. Furthermore, a higher sustained pressure during cementation reduces blister formation in the cement (25); blisters can act as aw-initiating sites during pull-out testing, interfering with the bond strength. Based on these results, it seems that the relining procedure increases the ber post retention by improving the contact between the cement and the adhesive rather than by reducing the defects observed in thin cement layers. Regarding the cement, the RelyX Luting 2 presented the lowest values of pull-out bond strength. This is a resin-modied glass ionomer, exhibiting lower bond strengths than other evaluated cements (26). The low bond strength is probably related to the application over the smear layer because no acid solution is applied before cementation. Thus, the retention provided by RelyX Luting 2 is more dependent on frictional retention than on its bonding to dentin. Because the mechanical properties of this cement are poorer compared with the resin cements, it is expected that this cement will have the lowest retention (27). Conversely, RelyX Unicem is a self-adhesive resin cement and interacts very supercially with the substrate, resulting in poor bond strength (28). However, it appears that RelyX Unicem has good chemical interaction with the calcium in hydroxyapatite, improving their mechanical properties (29). Furthermore, self-adhesive resin cements appear to have low shrinkage because of their viscoelastic properties, leading to better intimate contact of the resin cement with the root canal walls and higher frictional resistance (30). RelyX ARC is a conventional dual-cured resin cement that generates higher bond strength to dentin than the other cements evaluated (27). However, its high polymerization shrinkage and the resulting stress could impair the bonding to the root dentin (31). This increases the dependency on mechanical properties of cement for improving the post retention. Despite being a dual-cured material, deeper portions of cement are inaccessible to light, rendering the material dependent on the chemical curing. This can reduce the degree of conversion of the cement and consequently affect its mechanical properties. However, the use of a primer that is activated before insertion of the cement may improve the conversion. The primer contains sodium and sulnic acid salts that react with the primers acidic resin monomers to produce free radicals that would enhance the polymerization reaction of the cement (32, 33). Thus, the approach indicated by the manufacturer

of RelyX ARC improves both the mechanical properties of the cement and post retention. Although ber post retention seems to be derived predominantly from frictional retention, the luted length of the post is just as important to improve their retention. Relining the ber post reduces the thickness of the cement layer and also improves the frictional retention. The reduction of resin thickness may also reduce the polymerization stress because the stress development increases associated with increased volume of the resin cement (34). In conclusion, the use of cement with proper mechanical properties is essential for adequate post retention.

References
1. Fernandes AS, Shetty S, Coutinho I. Factors determining post selection: a literature review. J Prosthet Dent 2003;90:55662. 2. Soares CJ, Soares PV, de Freitas Santos-Filho PC, et al. The inuence of cavity design and glass ber posts on biomechanical behavior of endodontically treated premolars. J Endod 2008;34:10159. 3. Dietschi D, Duc O, Krejci I, et al. Biomechanical considerations for the restoration of endodontically treated teeth: a systematic review of the literature, Part II (Evaluation of fatigue behavior, interfaces, and in vivo studies). Quintessence Int 2008;39: 11729. 4. Nakamura T, Ohyama T, Waki T, et al. Stress analysis of endodontically treated anterior teeth restored with different types of post material. Dent Mater J 2006; 25:14550. 5. Schmitter M, Huy C, Ohlmann B, et al. Fracture resistance of upper and lower incisors restored with glass ber reinforced posts. J Endod 2006;32:32830. 6. DArcangelo C, De Angelis F, Vadini M, et al. In vitro fracture resistance and deection of pulpless teeth restored with ber posts and prepared for veneers. J Endod 2008;34:83841. 7. Santos-Filho PC, Castro CG, Silva GR, et al. Effects of post system and length on the strain and fracture resistance of root lled bovine teeth. Int Endod J 2008;41: 493501. 8. Braga NM, Paulino SM, Alfredo E, et al. Removal resistance of glass-ber and metallic cast posts with different lengths. J Oral Sci 2006;48:1520. 9. Faria e Silva AL, Arias VG, Soares LE, et al. Inuence of ber-post translucency on the degree of conversion of a dual-cured resin cement. J Endod 2007;33:3035. 10. Faria e Silva AL, Casselli DS, Ambrosano GM, et al. Effect of the adhesive application mode and ber post translucency on the push-out bond strength to dentin. J Endod 2007;33:107881. 11. DArcangelo C, Cinelli M, De Angelis F, et al. The effect of resin cement lm thickness on the pullout strength of a ber-reinforced post system. J Prosthet Dent 2007;98: 1938. 12. Teixeira CS, Silva-Sousa YT, Sousa-Neto MD. Bond strength of ber posts to weakened roots after resin restoration with different light-curing times. J Endod 2009;35:10349. 13. De-Deus G, Murad C, Paciornik S, et al. The effect of the canal-lled area on the bacterial leakage of oval-shaped canals. Int Endod J 2008;41:18390. 14. Grandini S, Goracci C, Monticelli F, et al. SEM evaluation of the cement layer thickness after luting two different posts. J Adhes Dent 2005;7:23540. 15. Grandini S, Sapio S, Simonetti M. Use of anatomic post and core for reconstructing an endodontically treated tooth: a case report. J Adhes Dent 2003;5:2437. 16. Faria-e-Silva AL, Pedrosa-Filho CF, Menezes MS, et al. Effect of relining on ber post retention to root canal. J Appl Oral Sci 2009;17:6004.

JOE Volume 36, Number 9, September 2010

Pull-out Strength of Fiber Posts

1545

Basic ResearchTechnology
17. Naumann M, Sterzenbach G, Rosentritt M, et al. Is adhesive cementation of endodontic posts necessary? J Endod 2008;34:100610. 18. Naumann M, Blankenstein F, Dietrich T. Survival of glass bre reinforced composite post restorations after 2 years-an observational clinical study. J Dent 2005;33:30512. 19. Bitter K, Noetzel J, Stamm O, et al. Randomized clinical trial comparing the effects of post placement on failure rate of postendodontic restorations: preliminary results of a mean period of 32 months. J Endod 2009;35:147782. 20. DArcangelo C, Zazzeroni S, DAmario M, et al. Bond strengths of three types of brereinforced post systems in various regions of root canals. Int Endod J 2008;41: 3228. 21. Giachetti L, Grandini S, Calamai P, et al. Translucent ber post cementation using light- and dual-curing adhesive techniques and a self-adhesive material: push-out test. J Dent 2009;37:63842. 22. Hayashi M, Okamura K, Wu H, et al. The root canal bonding of chemical-cured totaletch resin cements. J Endod 2008;34:5836. 23. Goracci C, Fabianelli A, Sadek FT, et al. The contribution of friction to the dislocation resistance of bonded ber posts. J Endod 2005;31:60812. 24. Faria-e-Silva AL, Reis AF, Martins LR. The effects of different luting procedures in the push-out bond strength of bre posts to the root canal. Braz J Oral Sci 2008;27: 16536. 25. Chief N, Chersoni S, Papacchini F, et al. The effect of application sustained seating pressure on adhesive luting procedure. Dent Mater 2007;23:1596. 26. Bonfante G, Kaizer OB, Pegoraro LF, et al. Tensile bond strength of glass ber posts luted with different cements. Braz Oral Res 2007;21:15964. 27. Amaral M, Santini MF, Wandscher V, et al. An in vitro comparison of different cementation strategies on the pull-out strength of a glass ber post. Oper Dent 2009;34:4435. 28. De Munck J, Vargas M, Van Landuyt K, et al. Bonding of an auto-adhesive luting material to enamel and dentin. Dent Mater 2004;20:96371. 29. Gerth HU, Dammaschke T, Zuchner H, et al. Chemical analysis and bonding reaction of RelyX Unicem and Bix compositesa comparative study. Dent Mater 2006;22: 93441. 30. Dauvillier BS, Feilzer AJ, De Gee AJ, et al. Visco-elastic parameters of dental restorative materials during setting. J Dent Res 2000;79:81823. 31. Tay FR, Loushine RJ, Lambrechts P, et al. Geometric factors affecting dentin bonding in root canals: a theoretical modeling approach. J Endod 2005;31:5849. 32. Arrais CA, Giannini M, Rueggeberg FA. Effect of sodium sulnate salts on the polymerization characteristics of dual-cured resin cement systems exposed to attenuated light-activation. J Dent 2009;37:21927. 33. Faria-e-Silva AL, Moraes RR, Ogliari FA, et al. The role of the primer. J Oral Sci 2009; 51:2559. 34. Braga RR, Boaro LC, Kuroe T, et al. Inuence of cavity dimensions and their derivates (volume and C factor) on shrinkage stress development and microleakage of composite restorations. Dent Mater 2006;22:81823.

1546

Macedo et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Comparative Evaluation of Modied Canal Staining and Clearing Technique, Cone-Beam Computed Tomography, Peripheral Quantitative Computed Tomography, Spiral Computed Tomography, and Plain and Contrast Mediumenhanced Digital Radiography in Studying Root Canal Morphology
Prasanna Neelakantan, MDS, Chandana Subbarao, BDS, and Chandragiri V. Subbarao, MDS
Abstract
Introduction: This study investigated the accuracy of cone-beam computed tomography (CBCT), peripheral quantitative computed tomography (pQCT), spiral computed tomography (SCT), plain (plain digi), and contrast mediumenhanced digital radiographs (contrast digi) in studying root canal morphology. Methods: The root canal anatomy was analyzed in 95 teeth using CBCT, pQCT, SCT, plain digi, and contrast digi. After ushing out the radiopaque dye, access cavities were sealed, and the teeth were subject to the modied canal staining and clearing technique. The number of root canals (Vertucci classication and Gulabivalas additional classes) was calculated by three calibrated endodontists and two maxillofacial radiologists. Erroneous or unsuccessful identications of root canals were statistically analyzed by one-way analysis of variance (p = 0.05). Results: The modied canal staining and clearing technique identied an average of 1.8 root canals per mandibular central incisor, 2.3 per maxillary rst premolar, 3.9 per maxillary rst molar, 3.8 per maxillary and mandibular second molar, and 4.3 per mandibular rst molar. CBCT and pQCT were erroneous in 0.29% and 2.05% cases, whereas SCT, contrast digi, and plain digi were unsuccessful in 15.58%, 14.7%, and 23.8%, respectively. There was a signicant difference between all the methods (p < 0.05) in the unsuccessful identication of root canals except between CBCT and modied canal staining and clearing technique where there was no signicant difference (p > 0.05). Conclusions: CBCT and pQCT were as accurate as the modied canal staining and tooth clearing technique in identifying root canal systems. (J Endod 2010;36:15471551)

Key Words
Cone-beam computed tomography, contrast media, digital radiograph, modied staining and clearing, morphology, peripheral quantitative computed tomography, root canal, spiral computed tomography

uccessful endodontic therapy stems from thorough canal debridement and effective lling of the root canal system, for which knowledge of morphology of the root canals is a critical prerequisite (1). Internal complexities of the root canal are genetically determined and have denitive importance in anthropology (2). These factors necessitate the identication of a method that accurately determines the root canal morphology. There are numerous reports on the root canal morphologies of different populations, which is extremely important for an endodontist as well as general dental practitioners. Also of interest to us are the methods that have been used in these studies (3). The methods most commonly used in analyzing the root canal morphology are canal staining and tooth clearing (46), conventional radiographs (79), digital and contrast mediumenhanced radiographic techniques (10, 11), radiographic assessment enhanced with contrast media (12, 13), and more recently computed tomographic techniques (1416). Canal staining and tooth clearing (CS) is generally considered the gold standard in these studies (4, 5, 17). A modication of this technique proposed by Weng et al (18) is accurate, allows the appreciation of intricate details, and is nondestructive. An ideal technique would be one that is accurate, simple, non-destructive, and, most importantly, feasible in the in vivo scenario. The gold standard methods CS and the modied method cannot be used in vivo. The application of computed tomography (CT) scans in endodontics was rst reported by Tachibana and Matsumoto in 1990 (19). A CT scan uses a fan-shaped beam and multiple exposures around an object to reveal the internal architecture of this object, thereby helping the clinician to view morphologic features as well as pathology from different three-dimensional (3D) perspectives. The distinct advantage of a CT scan is that it allows for 3D reconstruction of root canal systems. CT scanning has been suggested as the preferential imaging modality in difcult situations demanding localization and description of root canal systems because of its ability to render 3D information (2022)

From the Department of Conservative Dentistry and Endodontics, Saveetha Dental College and Hospitals, Saveetha University, Chennai, India. Address requests for reprints to Dr Neelakantan Prasanna, Plot 1500, 16th Main Road, Anna Nagar West, Chennai, Tamil Nadu, India. E-mail address: prasu_endo@ yahoo.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.008

JOE Volume 36, Number 9, September 2010

Comparative Evaluation of Various Imaging Techniques in Studying Root Canal Morphology

1547

Basic ResearchTechnology

Figure 1. Classication of canal congurations according to Vertucci.

Cone-beam computed tomography (CBCT) scanning or digital volume tomography (DVT) uses an extraoral imaging scanner to produce 3D scans of the maxillofacial skeleton at a considerably lower radiation dose than conventional CT scanning. CBCT scanning has been shown to be more accurate than digital radiographs in determining root canal systems. CBCT scanning has also been used in vivo in diagnosis and preoperative assessment (20, 2326). Another computed tomography technique, peripheral quantitative computed tomography (pQCT), was originally introduced for bone mineral analysis. The only report on pQCT in studying root canal anatomy showed that this method offers accurate 3D reconstruction of the root canal systems (RCS) and analysis of endodontic procedures (16). The planar resolution of pQCT is approximately 70 70 mm, which is lower than mCT. Nevertheless, it might prove to be a nondestructive method of investigation at low cost and shorter scanning times. Spiral CT (SCT) has been used in several cases of diagnosis of aberrant root canal systems (2732) as well as in the identication of the root canal morphology of Indian molars (33). Despite a plethora of studies and clinical reports showing the canal morphology of different populations, the most accurate and ideal method is yet to be ascertained with scientic evidence. The hypothesis tested in this study was that there is a difference in the identication of the number of root canal systems between the modied canal staining and clearing technique (gold standard) versus CBCT, pQCT, SCT, contrast mediumenhanced digital radiographs, and plain digital radiographs. The aim of this in vitro investigation was to compare the efcacy of the six aforementioned methods in identifying root canal systems.

Board of the university. The teeth were washed under tap water immediately after extraction and stored in distilled water with thymol iodide crystals until the collection was complete. After this, the samples were washed thoroughly under tap water and immersed in 2.5% sodium hypochlorite for 30 minutes to remove adherent soft tissue.

CT Scans The teeth were randomly inserted into foam arches in close contact to each other to simulate their natural alignment in a dental arch. An acrylic facing was placed on the facial side to mimic soft tissue on the radiographs. All the teeth were scanned by a CBCT scanner (3DAccuitomo; J Morita Corporation, Osaka, Japan), a pQCT scanner (Research SA+; Stratec Medizin-technik GmbH, Pforzheim, Germany), and a SCT scanner (GE Electricals, Milwaukee, WI) with constant thicknesses of 125 m/slice, 250m/slice, and 650m/slice respectively. The teeth were viewed both cross-sectionally and longitudinally. Volume-rendering and multiplanar volume reconstruction were performed using the Advantage Windows workstation (GE Systems, Milwaukee, WI). Digital Radiography with Contrast Medium Following the scanning procedure, all the teeth were exposed to a radiograph using a digital radiography unit (DSX 730; Owandy Dental Imaging, France; and Kodak 2100 X ray unit; Kodak Dental Systems, Atlanta, GA). Access cavities were prepared in all the teeth. After gaining entry into the pulp chamber, the pulp tissue was extirpated with a ne barbed broach (Denstply Maillefer, Ballaigues, Switzerland). The teeth were placed in 5% sodium hypochlorite for 30 minutes to dissolve the pulp tissue (10), washed in water, dried, and a water-soluble lowviscosity radiopaque medium (diatrizoate sodium, Hypaque; Amersham Health Inc, Princeton, NJ) was delivered into the root canals with a monoject syringe with a needle gauge 23 to 27 depending on the teeth and root canal size. The teeth were subjected to vacuum for 2 minutes (24 mm Hg), and a reapplication of vacuum was performed after 3 minutes. The penetration of the dye into the niches of the root canal system was enhanced by ultrasonication for 2 minutes. The teeth were placed in a model simulating the maxillary or mandibular arches, and radiographs were taken in the buccolingual direction using a digital radiography unit (DSX 730 and Kodak 2100). All radiographs (before

Materials and Methods

A total of 95 recently extracted teeth (20 mandibular rst molars, 20 maxillary rst molars, 20 mandibular second molars, 20 maxillary second molars, 7 maxillary rst premolars, and 8 mandibular central incisors) with intact roots and mature apices were collected from the Oral Surgery Department of the University and stored at 100% humidity. It was ensured that for every tooth type, the number of roots was standardized (mandibular incisors, 1; maxillary rst premolars, 2; maxillary rst and second molars, 3; and mandibular rst and second molars, 2). The methodology was approved by the Institutional Review

Figure 2. Additional classes of canal congurations according to Gulabivala.

1548

Neelakantan et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 1. Average Number of Root Canals Identied with the Six Methods Tooth
Mandibular central incisor (n = 8) Maxillary rst premolar (n = 7) Maxillary rst molar (n = 20) Maxillary second molar (n = 20) Mandibular rst molar (n = 20) Mandibular second molar (n = 20)

Conventional radiographs
1 2 3.1 3.1 2.8 2.8

CBCT
1.8 2.3 3.8 3.8 4.3 3.8

pQCT
1.8 2.1 3.8 3.7 4.2 3.3

SCT
1.8 2.1 3.3 2.7 3.4 2.9

Digital radiograph with contrast medium

1.8 2 3.2 3.5 3.1 2.7

Modied canal clearing and staining

1.8 2.3 3.9 3.8 4.3 3.8

CBCT, cone-beam computed tomography; pQCT, peripheral quantitative computed tomography; SCT, spiral computed tomography.

and after introducing contrast medium) were taken in two horizontal angulations: 0 and 30 .

Modied Canal Staining and Clearing Technique The contrast medium was ushed out with water followed by 5.25% sodium hypochlorite and then rinsed again with water. The access cavities were sealed with light cure composite resin and light cured for 40 seconds. The modied method of canal staining has been described in detail elsewhere (18). Briey, after immersion of the teeth in India ink, the teeth were placed in a hyperbaric oxygen chamber set to a pressure of 0.6 MPa for 2 hours. The method of decalcication and rendering the teeth transparent was adopted from Robertson et al (34). After 12 hours of drying, the teeth were decalcied in 10% nitric acid for 28 to 30 hours. The acid was changed after 24 hours and was stirred every 8 hours. The endpoint of decalcication was determined by taking periodic radiographs of ve sample teeth. After thorough washing of the decalcied teeth in running tap water for 4 hours, the samples were dehydrated in ascending concentrations of ethanol (70%, 80%, 95%, and 100%) for 1 day, and the samples were rendered transparent by immersing in methyl salicylate for 2 days. The samples were analyzed using a magnifying lens (3). For all the methods, the evaluation was performed by three precalibrated and standardized endodontists and two oral and maxillofacial radiologists. To eliminate any bias, the evaluators were asked to analyze every slice (both cross-sectional and longitudinal) as well as the images obtained by multiplanar volume reconstruction to quantify the number of root canals. The number of root canals per tooth identied with modied canal staining and clearing technique was used as the gold standard to compare the number of root canals identied with CBCT, pQCT, SCT, contrast mediumenhanced digital radiographs, and digital radiography. The identication of root canal systems was done according to Vertuccis classication (4) and Gulabivalas additional classes (17) (Figs. 1 and 2). The total number of root canals was calculated for each tooth studied. In any canal system with divisions of the canal, the maximum number of divisions was considered as the number of canals. For example, type II canal system was considered as two canals, type 1-3-1 as three canals, and type VI as two canals. The average number of root canals identied per tooth type was calculated for all the methods. Descriptive statistical analysis was per-

formed by calculating the percentage of root canals found from the modied canal staining and clearing method with those identied with CBCT, pQCT, SCT, contrast mediumenhanced digital radiographs, and digital radiographic images. Interrater agreement was measured between all three endodontist evaluators and two radiologists as well as between each pair of endodontist evaluators and the radiologists. Intrarater agreement was measured by having all three endodontists and the two radiologists evaluate one half of the CBCT, pQCT, SCT, contrast mediumenhanced digital radiographs, and digital radiographic images at each of two separate sessions. The percentage of unsuccessful identications as compared with the modied canal staining and clearing method (0% unsuccessful) was statistically analyzed by oneway analysis of variance with a p value of 0.05 considered signicant.

Results
The average number of root canals identied by all the methods is depicted in Table 1. The percentage of root canals found from the modied canal staining and clearing analysis with those identied with CBCT, pQCT, SCT, contrast mediumenhanced digital radiographs, and digital radiographic images is given in Table 2. Table 3 presents the percentage of unsuccessful identication of root canals by all the methods analyzed in this study.

Interrater and Intrarater Agreement For the modied canal staining and clearing method, each pair of endodontist evaluators agreed with each other 100% of the time. For CBCT and pQCT, each pair of endodontist evaluators and radiologists agreed with each other between 98% and 100% of the time, whereas for SCT they agreed with each other 95% to 98% of the time. For digital radiography and contrast mediumenhanced digital radiographs, each pair of endodontists agreed with each other 80% to 83% and 86% to 89% of the time. Among all three endodontist evaluators and two radiologist evaluators, total agreement was found 100% of the time for the modied canal staining and clearing method, 99% for CBCT and pQCT, 98% for SCT, and 82% for digital radiography and 84% for contrast mediumenhanced digital radiographs. Intrarater agreement results showed that the endodontist evaluators agreed with themselves 100% of the time for the modied canal

TABLE 2. Correct Identication of the Total Number of Root Canals by All the Methods Evaluator
Endodontist 1 Endodontist 2 Endodontist 3 Radiologist 1 Radiologist 2 All

Digital radiograph (% correct)

80 83 82 80 80 82

CMDR (% correct)
88 86 89 88 87 84

CBCT (% correct)
99 100 99 98 100 99

pQCT (% correct)
100 98 98 99 99 99

SCT (% correct)
98 96 95 98 96 98

MCS (% correct)
100 100 100 100 100 100

CMDR, contrast mediumenhanced digital radiography; CBCT, cone-beam computed tomography; pQCT, peripheral quantitative computed tomography; SCT, spiral computed tomography; MCS, modied canal staining.

JOE Volume 36, Number 9, September 2010

Comparative Evaluation of Various Imaging Techniques in Studying Root Canal Morphology

1549

Basic ResearchTechnology
TABLE 3. Unsuccessful Identication of Root Canals in Individual Teeth in All the Techniques Compared with the Modied Canal Staining and Clearing Technique Missed one RCS in a tooth (%)
0 0 0.58 0.29 0.88

Method
Digital radiograph CMDR SCT CBCT pQCT

Missed two or more RCS in a tooth (%)

23.8 14.7 15 0 1.17

Total (%)
23.8 14.7 15.58 0.29 2.05

CMDR, contrast mediumenhanced digital radiography; CBCT, cone-beam computed tomography; pQCT, peripheral quantitative computed tomography; SCT, spiral computed tomography; RCS, root canal staining.

staining and clearing technique interpretation; 98% for CBCT, pQCT, and SCT; between 78% and 84% for digital radiographs; and 85% to 93% for contrast mediumenhanced digital radiograph interpretation. The evaluators failed to identify one or more root canals with digital radiographs in 23.8% of teeth, contrast mediumenhanced digital radiographs in 14.8%, SCT in 15.58%, CBCT in 0.29%, and pQCT in 2.05% of the teeth. There was a signicant difference between all the methods (p < 0.05) in unsuccessful identication of root canals.

Discussion
Failure to identify extra canals is implicated as one of the most common reasons for the failure of endodontic treatment (1). The analysis of the root canal morphologies of teeth of different populations, ethnicities, and preoperative assessment of root canal systems is of importance in this regard. This study compared the efcacy of the six methods (modied canal staining and clearing, CBCT, pQCT, SCT, digital radiography, and contrast mediumenhanced digital radiography) in identifying root canal systems. In the absence of overlying bone, tissue, and anatomic features, it may not be possible to directly extrapolate the results of this in vitro investigation to the clinical scenario. The canal staining and clearing technique is considered the gold standard method of studying root canal anatomy. We used the modied canal staining and clearing technique in this study. Although nondestructive and more accurate than the conventional staining and clearing method, the main disadvantage of this method is that it cannot be used in vivo (18). A method that has the accuracy of the canal staining and clearing technique and yet is clinically feasible is essential in endodontic practice. Radiographs taken after the introduction of radiopaque contrast media are more useful than plain radiographs in the assessment of root canal anatomy (12). Hypaque is a tensioactive, water-soluble contrast medium with a specic gravity similar to sodium hypochlorite. It is an aqueous solution of two iodine salts: sodium iodine and diatrizoate meglumine. The low surface tension enables its penetration into the niches of the root canal system (12). This alteration of subject contrast induced by variations in transmission of the radiographic beam between the tooth and the contrast medium denitely improves the visibility of canal systems in comparison with conventional radiographs. Digital radiographs were used in this study because they are more accurate than conventional radiographs in analyzing root canal anatomy (10). Exposing radiographs in two different horizontal angulations (30 shift and orthoradial position) provide additional information on root canal systems but show some amount of inherent distortion, which makes radiographic determination of some characteristics difcult. Our study showed that endodontists and radiologists only identied 1550

86% to 89% of root canals as compared with the modied canal staining and clearing method. Passive injection of the contrast medium results in the entrapment of air bubbles, hampering proper visualization of apical anatomy (10, 12, 35). Despite the injection of contrast media under pressure in our study, it is possible that complete perfusion may not occur (35). This is probably the reason why this method failed to identify the intricacies of root canal anatomy. The clearing and staining technique has been shown to be more accurate than the radiographic technique with contrast medium (35). Our results are in agreement. CT scanning is currently widely used in implantology, maxillofacial reconstruction, and endodontic diagnosis as well as for the assessment of canal preparation, obturation, and the removal of root llings (20, 23, 27, 28). The main advantages of CT scanning are that it is nondestructive and allows 3D reconstruction and visualization of the external and internal anatomy of the teeth. The results of our study showed that CBCT scanning was as accurate as the modied canal staining and clearing technique in identifying canal anatomy. CBCT scanning was erroneous only in 0.29% of the teeth examined. The differences between the results obtained by the evaluators were not signicantly different when endodontists were compared with radiologists. However, in some instances of analysis of the CT images, radiologists were able to interpret the number of root canals better than endodontist evaluators, but this was not true for all endodontist evaluators. Nevertheless, the experience of the radiologists in the analysis of CT images may be an attributable factor. The endodontic evaluators and radiologists identied 98% to 100% of root canals as compared with modied canal staining and clearing method when CBCT and pQCT were used. The slice thickness for CBCT ranges from 80 to 200 m. The slice thickness used in our study was 125 m. The primary advantages of CBCT are signicantly lower effective radiation dose, short exposure time (2-5 seconds), less expensive than conventional CT scanning, and highly accurate. Also, CBCT measurements are geometrically accurate because of the fact that the CBCT voxels (3D pixels containing data) are isotropic (20, 26). pQCT, with a slice thickness of 250 m failed to identify 2.05% of RCS as compared with the gold standard. Also, the large pixel size as mentioned earlier is responsible for a low resolution, which may cause errors in the identication of intricate details of the root canal system. The advantages of pQCT are that it is economic, scanning times are shorter, and it allows mapping of multiple teeth at the same time (16). Our study showed that SCT did not identify 15.58% of root canals. The frequent errors were in differentiating type II canal systems from type III and 1-3-1 systems and in differentiating type IV canal systems from type 3-2. We attribute this to the slice thickness. The usual thickness of each slice in SCT ranges from 0.65 mm to 1.0 mm (33). In our study, it was 0.65 mm/slice. It is possible that this slice thickness does not offer sufcient reproducibility along the length of the canal (14). For example, type VI canal system may be misinterpreted as a type IV system if the fusion and subsequent division of the canals is covered in the same slice. Another main disadvantage of SCT is the increased radiation dose as compared with CBCT scanning (20).

Conclusions
CBCT and pQCT are as accurate as the gold standard (modied canal staining and clearing technique) in identifying root canal anatomy. The application of 3D reconstruction based on CBCT and pQCT require further evaluation to validate its clinical application.

References
1. Grossman IL, Oliet S, Del Rio E. Endodontic Practice. 11th ed. Philadelphia, PA: Lea and Fabringer; 1988:14551.

Neelakantan et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
2. Sperber GH. The phylogeny and odontogeny of dental morphology. In: Sperber GH, ed. From Apes to Angels. New York, NY: Wiley-Liss; 1990:2159. 3. Cleghorn BM, Christie WH, Dong CC. Root and root canal morphology of the human permanent maxillary rst molar: a literature review. J Endod 2006;32:81321. 4. Vertucci FJ. Root canal morphology of mandibular premolars. J Am Dent Assoc 1978;97:4750. 5. Alavi AM, Opasanon A, Ng YL, et al. Root and canal morphology of Thai maxillary molars. Int Endod J 2002;35:47885. 6. Awawdeh L, Abdullah H, Al-Qudah A. Root form and canal morphology of Jordanian maxillary rst premolars. J Endod 2008;34:95661. 7. Pineda F, Kuttler Y. Mesiodistal and buccolingual roentgenographic investigation of 7275 root canals. Oral Surg Oral Med Oral Pathol 1972;33:10110. 8. Weine FS, Hayami S, Hata G, et al. Canal conguration of the mesiobuccal root of the maxillary rst molar of a Japanese sub-population. Int Endod J 1999;32:7987. 9. Pattanshetti N, Gaidhane M, Al Kandari AM. Root and canal morphology of the mesiobuccal and distal roots of permanent rst molars in a Kuwait populationa clinical study. Int Endod J 2008;41:75562. 10. Fan B, Gao Y, FanW, et al. Identication of a C-shaped canal system in mandibular second molars. Part II. The effect of bone image superimposition and intraradicular contrast medium on radiograph interpretation. J Endod 2008;34:1605. 11. Patel S, Dawood A, Whaites E, et al. New dimensions in endodontic imaging: part 1. Conventional and alternative radiographic systems. Int Endod J 2009;42:44762. 12. Naoum HJ, Love RM, Chandler NP, et al. Effect of X-ray beam angulation and intraradicular contrast medium on radiographic interpretation of lower rst molar root canal anatomy. Int Endod J 2003;36:129. 13. Thomas RP, Moule AJ, Bryant R. Root canal morphology of maxillary permanent rst molar teeth ate various ages. Int Endod J 1993;26:25767. 14. Plotino G, Grande NM, Pecci R, et al. Three dimensional imaging using micrcomputed tomography for studying tooth macromorphology. J Am Dent Assoc 2006; 137:155561. 15. Fan B, Yang J, Gutmann JL, et al. Root canal systems in mandibular rst premolars with C-shaped root congurations. Part I: Microcomputer tomography mapping of the radicular groove and associated root canal cross-sections. J Endod 2008;34:133741. 16. Sberna MT, Rizzo G, Zacchi E, et al. A preliminary study of the use of peripheral quantitative computed tomography for investigating root canal anatomy. Int Endod J 2009;42:6675. 17. Gulabivala K, Aung TH, Alavi A, et al. Root and canal morphology of Burmese mandibular molars. Int Endod J 2001;34:35970. 18. Weng XL, Yu SB, Zhao SL, et al. Root canal morphology of permanent maxillary teeth in the Han nationality in Chinese Guanzhong area: a new modied root canal staining technique. J Endod 2009;35:6516. 19. Tachibana H, Matsumoto K. Applicability of x-ray computerized tomography in endodontics. Endod Dent Traumatol 1990;6:1620. 20. Patel S, Horner K. The use of cone beam computed tomography in endodontics. Int Endod J 2009;42:7556. 21. La SH, Jung DH, Kim EC, et al. Identication of independent middle mesial canal in mandibular rst molar using cone-beam computed tomography imaging. J Endod 2010;36:5425. 22. Bornstein MM, Wolner-Hanssen AB, Sendi P, et al. Comparison of intraoral radiography and limited cone beam computed tomography for the assessment of rootfractured permanent teeth. Dent Traumatol 2009;25:5717. 23. Nair MK, Nair UP. Digital and advanced imaging in endodontics: a review. J Endod 2007;33:16. 24. Peters OA, Laib A, Ruegsegger P, et al. Three-dimensional analysis of root canal geometry using high-resolution computed tomography. J Dent Res 2000;79: 14059. 25. Mozzo P, Procacci C, Tacconi A, et al. A new volumetric CT machine for dental imaging based on the cone-beam technique: preliminary results. Eur Radiol 1998;8:155864. 26. Matherne RP, Angelopoulos C, Kulilid JC, et al. Use of cone-beam computed tomography to identify root canal systems in vitro. J Endod 2008;34:879. 27. Gopikrishna V, Reuben J, Kandaswamy D. Endodontic management of a maxillary rst molar with two palatal roots and a single fused buccal root diagnosed with spiral computed tomographya case report. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;105:e748. 28. Sachdeva GS, Ballal S, Gopikrishna V, et al. Endodontic management of a mandibular second premolar with four roots and four root canals with the aid of spiral computed tomography: a case report. J Endod 2008;34:1047. 29. Aggarwal V, Singla M, Logani A, et al. Endodontic management of a maxillary rst molar with two palatal canals with the aid of spiral computed tomography: a case report. J Endod 2009;35:1379. 30. Chandra SS, Rajasekaran M, Shankar P, et al. Endodontic management of a mandibular rst molar with three distal canals conrmed with the aid of spiral computerized tomography: a case report. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108:e7781. 31. Ma L, Chen J, Wang H. Root canal treatment in an unusual maxillary rst molar diagnosed with the aid of spiral computerized tomography and in vitro sectioning: a case report. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;107: e6873. 32. Rani AK, Metgud S, Yakub SS, et al. Endodontic and esthetic management of maxillary lateral incisor fused to a supernumerary tooth associated with a talon cusp by using spiral computed tomography as a diagnostic aid: a case report. J Endod 2010; 36:3459. 33. Reuben J, Velmurugan N, Kandaswamy D. The evaluation of root canal morphology of the mandibular rst molar in an Indian population using spiral computed tomography scan: an in vitro study. J Endod 2008;34:2125. 34. Robertson D, Leeb IJ, McKee M, et al. A clearing technique for the study of root canal systems. J Endod 1980;6:4214. 35. Bedford JM, Martin DM, Youngson CC. Assessment of a contrast medium as an adjunct to endodontic radiography. Int Endod J 2004;37:80613.

JOE Volume 36, Number 9, September 2010

Comparative Evaluation of Various Imaging Techniques in Studying Root Canal Morphology

1551

Basic ResearchTechnology

Design Improvement and Failure Reduction of Endodontic Files through Finite Element Analysis: Application to V-Taper File Designs
Rui He, PhD,* and Jun Ni, PhD
Abstract
Introduction: Torsional stiffness and bending exibility are essential characteristics as far as the performance and safety of the endodontic les are concerned. Inadequacy in addressing these requirements in le design leads to increased risk of le failure. The stiffness and exibility of the endodontic le are greatly dependent on its geometric design. The aim of this study was to evaluate the inuence of geometric features on the mechanical performance of endodontic les through numerical simulations. Methods: Finite element models of V-Taper le were developed, and the mechanical behavior of the le under bending and torsional loads was simulated. The inuence of helix angle, taper, and ute length was evaluated through parametric studies. Results: In the helix angle range between 5 and 40 degrees, the bending exibility and torsional stiffness both improve with increasing helix angle. The torsional stiffness increases with increasing taper or decreasing ute length, accompanied by a decrease in bending exibility. Changing the ute length alone does not result in a change in the stress prole in the tip section. The elastic limit of V-Taper le tip section was estimated in the form of transverse deection and angular deformation in bending and in torsion, respectively. Conclusions: The inuence of helix angle, taper, and ute length on the bending exibility and torsional stiffness of V-Taper les was quantitatively assessed through parametric studies with nite element method. The elastic limit of the V-Taper le tip section was estimated. A design methodology for achieving improved mechanical performances was proposed. (J Endod 2010;36:15521557)

Key Words
Bending exibility, elastic limit, endodontic le design, failure reduction, nite element method, helix angle, taper, torsional stiffness

From *Research and Development, Boston Scientic Corporation, St Paul, Minnesota; and Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan. Address requests for reprints to Dr Rui He, Boston Scientic Corporation, Research and Development, 4100 Hamline Ave N, St Paul, MN 55112-5798. E-mail address: ruihe@umich.edu. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. All rights reserved. doi:10.1016/j.joen.2010.06.002

n recent decades nickel-titanium (NiTi) rotary les have gained increasing popularity over stainless steel les in root canal preparation. Mainly due to the superelastic behavior of its base material of nitinol, NiTi les are more exible than stainless steel les (1, 2) and better preserve the canal anatomy with less canal aberrations such as zips, ledges, apical foramen transportation, or perforations (35). Despite these advantages, however, fracture of NiTi les remains a concern in clinical practices (6, 7). A recent questionnaire survey revealed that a signicant population of general dental practitioners and endodontists have encountered endodontic le fractures (8). An endodontic le faces 2 obvious challenges in canal instrumentation. It rst needs to be strong to facilitate effective cutting of the dentin material for canal shaping purposes. Adequate torsional strength is necessary to maintain the cutter geometry and edge strength of the le, preventing torsional damage. A le also needs to be exible in bending to follow the canal anatomy during instrumentation with good centering ability (9), and to minimize the aberrations to the canal. In addition, adequate exibility helps lower the bending stresses and reduce the risk of exural fatigue failure (10). Both torsional stiffness and bending exibility are essential characteristics for le performance and safety concerns. Inadequacy in addressing these performance requirements in le design leads to increased risk of le failure. Studies suggested that exural failure and torsional failure are the 2 major failure modes for rotary NiTi les (6, 7, 1113). Numerous studies also showed that the stiffness and exibility of endodontic les are greatly dependent on their geometric design, including taper, helix angle, cross-section shape, tip size, and length, etc. (1420). The mechanical performance of endodontic les can be evaluated through lab testing or numerical simulation. Numerical simulation is a low cost alternative to lab testing. The numerical simulation method is often more convenient than lab testing when it comes to studying what-if scenarios. Numerical modeling has been increasingly used to study the mechanical behavior of the endodontic le in the past decade. With a boundary integral method, Turpin et al (16) calculated the stresses on les in bending and in torsion and found that the cross-section shape has a marked inuence on the stress prole. With a nite element model, Berutti et al (17) evaluated the bending and torsional behavior of NiTi les and found that a le with concave cross section is more exible in bending but less stiff in torsion compared to one with a convex cross section. Xu et al (14) studied the inuence of cross-sectional prole on torsional and exural behavior of NiTi les and found that the peak torsional stress decreases with increasing inner core diameter. Kim et al (21) created full-length models of 3 commercial NiTi les and calculated stress distribution and force responses in simulated canal shaping and residual stresses after unloading. Necchi et al (22) simulated the insertion and withdrawal of a commercial le against a rigid root canal and obtained a strain map of the le. They also investigated the inuence of le material and root canal geometry on the strain prole. Their work was extended in another study (23) to include the effect of several NiTi material variants on the strain prole. Kim et al (24) and Kim et al (25) studied the effect of cross-section shape on the stress prole during simulated bending, torsion, and canal shaping by using full-length numerical model of commercial les. It should be pointed out that numerical simulation has its own limitations. For example, a model that contains only the gross geometry of an endodontic le might

1552

He and Ni

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
not allow for the study of the effects of material imperfection or manufacturing defects present in physical instruments (26). and on the 20-degree helix model (61 degrees). This result indicates that the torsional stiffness of the le increases with increasing helix angle. This correlation between torsional stiffness and the helix angle was also reported in the nite element studies on twist drills (27, 28). The stress proles on the les under bending and torsion were also simulated. It was found that the helix angle greatly affects the bending stress distribution. Although a minimal stress zone is always located on the neutral plane, the location of peak stress varies with the helix angle. To illustrate this point, Fig. 1c, d present the bending stress proles on the cross-sectional views of a 15-degree helix model and a 35-degree helix model, respectively. The peak stress occurs right on the cutting edge on the 15-degree model but shifts to the base of the ute on the 35-degree model. Regardless of helix angle, however, the peak stress is always located in the plane of bending. These observations agree with the results reported in other studies (16, 17, 25). Due to the fact that the base is closer to the neutral plane than the edge is, the peak stress on the 35-degree helix is lower than that on the 15-degree helix. No correlation was observed between the peak bending stress and the helix angle. In contrast, a correlation was observed between peak stress and helix angle under torsional deformation. The peak stress on the model was found to decrease continuously as the helix angle increases from 5 degrees to 25 degrees across models when loaded with a 0.4 Nmm torque. The peak torsional stress is consistently located at the base of the ute regardless of helix angle, an observation also reported in other studies (16, 17, 25). Fig. 1e, f show the torsional stress distribution on the 15-degree and 35-degree helix models, respectively. Because of the consistency in peak stress location across models, the peak stress value correlates inversely with the torsional stiffness of the le. It is noteworthy that after the helix angle reaches 25 degrees, further increase in helix angle results in only minimal change in peak torsional stress. This is because at or above 25-degree helix angle, the stresses on the model decrease to the nitinol superelasticity plateau, where a large change in strain results in only minimal change in stress.

Material and Methods

In this work numerical models were developed to study the inuence of geometric features on the mechanical behavior of endodontic les. The numerical models discussed throughout the scope of this work were constructed using the cross-sectional prole of a V-Taper le (Guidance Endo, Albuquerque, NM). A non-linear stress-strain curve of nitinol presented in the literature (17) was used in this study to approximate the mechanical behavior of the base material. This material model describes the behavior of nitinol in 3 distinct phases, including an initial linear elastic phase, a superelastic plateau phase, and a nal strain-hardening phase. While studying the effect of the helix angle (angle between the tangent of the helical cutting edge and the longitudinal axis of the le), a set of 8 le models was created. The helix angle was xed on each model but increased evenly from 5 to 40 degrees across models by increment of 5 degrees. All models have the same length of 1.8 mm and the same taper of zero degree. A short length was chosen to lower the computational cost. All models share a common taper to avoid possible variation in results caused by a change in taper. The diameter of the cross section is 0.40 mm, coincident to that of an ISO size 40 tip. To verify the generality of results from xed-helix models, a le with varying helix angle and zero taper was also included in this study. The helix angle on this le increases from 9 degrees at the tip to 20 degrees at the shank end, matching the measured helix prole of a V-Taper le. While studying the effect of taper, 2 le models were created, with zero taper on one model and a 0.06 taper on the other. The helix angle on both models increases from 9 degrees at the tip end to 20 degrees at the shank end. Both models have the same length of 1.8 mm. While evaluating the impact of ute length, 8 models were created, with le length increasing evenly from 6 to 20 mm across models by increments of 2 mm. All models have a size 40 tip, a 0.02 taper, and an identical helix prole that increases from the tip to the shank. The helix angle at the tip is 10 degrees. The helix angle at the shank varies with the model length, from 16 degrees on the 6-mm length model to 30 degrees on the 20-mm length model. A cantilever beam boundary condition was applied to all le models. The models were constrained on all translational and rotational degrees of freedom at the shank end and were loaded at the tip end.

Results
Effect of the Helix Angle A bending moment of 0.7 Nmm was applied to the tip end of the le on 8 models of xed helix angles. As the helix angle increases from 5 degrees to 40 degrees across the models, the simulated tip deection increases continuously from 2.52 to 3.55 mm (Fig. 1a). Similar results were observed when a transverse load of 0.4 N was applied to the tip. Interestingly, the tip deection on the model with helix varying from 920 degrees (2.66 mm) falls in between those observed on the 5degree helix model (2.52 mm) and on the 20-degree helix model (2.79 mm). This result suggested that the bending exibility of the le increases with increasing helix angle (or decreasing pitch). Torsional behavior was also simulated on the same set of models. When subjected to a 0.4 Nmm torque, the angular deformation at the tip end decreases continuously with increasing helix angle, from 97 degrees on the 5-degree helix model to 36 degrees on the 40-degree helix model (Fig. 1b). When the same torque was applied to the model with helix varying from 920 degrees, its angular deformation (94 degrees) falls in between those observed on the 5-degree helix model
JOE Volume 36, Number 9, September 2010

Effect of the Taper The taper feature obviously adds more material to the le and makes a stiffer le body, which affects both bending and torsional behavior of the le. When subjected to a 0.7 Nmm bending moment, the zero-taper model reported a tip deection of 2.66 mm, compared to a 1.04-mm tip deection on the 0.06-taper model. Similarly, when subjected to a 0.7 Nmm torque, the tip angular deformation drops from 352 degrees on the zero-taper model to 122 degrees on the 0.06-taper model. These observations indicate that an increase in taper results in an increase in torsional rigidity but a decrease in bending exibility. This result is in agreement with the reported testing results of a commercial endodontic le (19). The taper also has a great impact on the stress prole. The taper feature makes the shank stiffer than the tip and directs the higher stresses to the tip, a result also reported by Kim et al (24). In addition, an increase in taper adds more material to the overall body, leading to an increase in stiffness and decrease in stress. Under 0.7 Nmm loading, the peak stress in bending and in torsion drops by 19% and 10% respectively as the taper increases from zero to 0.06. Effect of the Flute Length An increase in the ute length adds slenderness to the le body, resulting in decreased resistance to deformation. When a 3.0 Nmm bending moment was applied to the tip of a le with a size 40 tip and a 0.02 taper, the tip deection increases continuously with increasing ute length, from 4.5 mm on the 6-mm-length model to 7.7 mm on
1553

Design Improvement and Failure Reduction of Endodontic Files through Finite Element Analysis

Basic ResearchTechnology

Figure 1. Simulated mechanical behavior of V-Taper endodontic les in bending and torsion. (This gure is available in color online at www.aae.org/joe/.)

the 20-mm-length model. This indicates that a le of longer ute length has better bending exibility. The transverse deection of the cutting edge along the le length was plotted in Fig. 1g for 3 models of different length. This plot shows that the deection peaks at the tip and drops quickly within a few millimeters of the tip, regardless of the le length. This quick drop of deection is mainly due to the effect of the taper. Moving toward the xed shank end, the deection approaches asymptotically to zero. Interestingly, despite the marked difference in their deection proles, models of different length have an identical stress prole in their common 1554

section as a result of the identical boundary condition on them (Fig. 1h). Fig. 1h also shows that the stress reaches its peak at a very short distance from the tip, where the helical cutting edge crosses the bending plane. The stress drops rapidly to the superelasticity stress plateau of nitinol within less than 3 mm from the tip. The angular deformation and torsional stresses were also examined for models of different length. Similar results were obtained as in the case of bending. Angular deformation increases with increasing ute length, indicating that the torsional stiffness of a le decreases with increasing ute length.

He and Ni

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
Discussion
Impact of Geometric Parameters The taper feature introduces base material to the le body and increases the stiffness of the le (15, 19, 20). A larger taper offers stronger resistance to torsional breakage and helps maintain the cutting edge strength and cutting efcacy. Apparently a large taper makes the rotary le a bigger cutter and helps remove more dentin material from the canal wall, but it also encounters a higher torque during preparation (29, 30). On the other hand, the taper makes a stiffer le, leading to higher risk of root canal straightening, abbreviation, or transportation during instrumentation (9, 15). It also results in higher bending stresses and lower resistance to exural fatigue failure (30). Therefore, a large taper le is appropriate when strong stiffness or high cutting efcacy is needed, such as during initial orice enlargement (31) or while preparing the rst half of the working length using a crown-down technique (4). It can also be used to address calcied but relatively straight canals. In contrast, a le with small taper is suitable for severely curved or S-shaped canals where great exibility is needed (32, 33). The disadvantage of using a small taper le is that it has a relatively poor capability of producing a taper-shaped canal (33). The stiffness of a le decreases with increasing ute length. Although a longer le has superior bending exibility, it provides less resistance to twist deformation. Files of different ute length share an identical stress prole in their common section, suggesting an equal propensity for le damage. In a canal treatment, the selection of ute length is obviously dependent on the canal anatomy and the instrumentation protocol. The simulation results showed that severe deformation and plastic stresses are concentrated within 3 mm of the tip regardless of ute length, suggesting that le damage is most likely to occur in the tip section. This observation agrees with results from other studies (6, 19, 34). To date, studies on the helix angle and its inuence on mechanical strength of endodontic les are very limited. Low et al (35) acknowledged that the spiral ute design adds complexity to the exural behavior and developed a curve-tting based model to include the effect of ute helix. Our work is probably one of the rst studies that examined the helix angle in relation to the structural strength of endodontic les in details. Our study suggested that the helix angle inuences the mechanical strength of the le in a different way than other geometric parameters do. As discussed earlier, an increase in bending exibility is always accompanied by a decrease in torsional stiffness as the taper or the ute length of a le changes, or vice versa. Other studies suggested that the geometric feature of cross-sectional prole inuences the mechanical strength of the le in a similar way as the taper or length factor does (16, 17). Our study showed that as the helix angle increases, the
bending stiffness decreases, but the torsional stiffness increases. In other words, the 2 seemingly competing characteristics of bending exibility and torsional stiffness can be simultaneously improved with increasing helix angle. Files with large helix angle reportedly have longer ex fatigue life than les with small helix angle (18), which can be largely credited to the high exibility and low bending stresses associated with large helix angle. The favorable correlation between helix angle and mechanical performances could imply new alternatives for le selection in clinical practices. In situations where more exibility is needed to negotiate severely curved canals or more stiffness is needed to address highly calcied canals, or a combination of both cases, endodontists can switch to les of larger helix angle potentially to solve the problem. If the endodontists have access to a variety of helix angles within the same le family, it will make this solution even more attractive. This is because endodontists thus are able to keep their choice on other features of the le, such as cross-sectional shape, tip size, taper, length, which could be driven by personal experience or clinical protocols.

Safe Use of Endodontic Files Empowered by the superelasticity behavior of nitinol, NiTi endodontic les can take fairly large deformation and stay in the elastic deformation domain. However, severe loading will still deform the material plastically and expose the le to permanent damage. The load at which plastic deformation develops is referred to as the elastic limit of the le. If the load on the le is kept below the elastic limit of the le during instrumentation, the risk of le failure can be greatly reduced (36). With numerical simulations the elastic limit of a le can be readily estimated. The numerical model allows for calculation of the peak stress on a le under specic loading conditions, thus a curve of peak stress versus load can be obtained. The load at the transition point from elastic stresses to plastic stresses on this curve is the elastic limit of the le. As a simple example, the stress-load curves of a size 25, 0.04 taper, 1.8-mm length V-taper tip section under pure bending and pure torsion as a cantilever beam were established (Fig. 2). The peak stress reaches the yield point of nitinol ($530 MPa) when the le is bent under a 0.6 Nmm bending moment, or when the tip deection reaches 0.8 mm. In this case, the elastic limit of the le segment was depicted by either the bending moment or the tip deection. Similarly, as the le is loaded in torsion, the elastic limit of the le was identied as 0.5 Nmm torque or 80-degree twist of the tip. Following this approach, the elastic limit of a full-length endodontic le under more complex loading conditions can be determined. The bending load applied to a le during the instrumentation is largely determined by the morphology of a root canal system (canal

Figure 2. Estimated elastic limit of V-Taper le tip section. (This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Design Improvement and Failure Reduction of Endodontic Files through Finite Element Analysis

1555

Basic ResearchTechnology
design is obtained, it is evaluated against the design requirements on torsional stiffness and bending exibility. The design for mechanical performances is completed if the requirements on le stiffness and exibility are met. The core of this design process is to use nite element models to assist evaluating a design against the performance requirements and making design changes accordingly. Performance requirements dene the upper limit of load or deformation that is safe to a le, which is essentially the elastic limit of the le. Whether a design survives a specic elastic limit can be readily evaluated with nite element analysis. If detrimental stresses are reported from the simulation, the design iteration starts where 1 or more geometric parameters will be optimized. The work of design optimization could benet from the knowledge of mechanical behavior in its relation to the design parameters.

Conclusions
Figure 3. Finite element analysisassisted endodontic le design for mechanical performance. (This gure is available in color online at www.aae.org/joe/.)

angle, curvature, length, etc), which can be clinically determined, such as through a preoperative radiographic study (37). The torsional load during instrumentation, which is dependent on the canal anatomy and the cutter geometry of the rotary le, can be mathematically computed with analytical models such as the one proposed by He (38). The clinical load that a le receives during instrumentation is then fully determined. If the clinical load exceeds the elastic limit of the le, the le is at risk of being damaged. Once such risk is identied, the same process can be followed to select a le with an elastic limit greater than the clinical load. The risk of le damage in a canal treatment is then minimized. Alternatively, the risk of le damage can be lowered by using the torque-limiting feature on the handpiece. This feature allows the motor to automatically stop and then reverse the rotation once the preset torque is reached, thus protecting the endodontic le against excessive torsional deformation (39). Setting the torque limit in a way such that the torsional load in combination with the expected bending load is below the elastic limit of the le, the risk of developing le failure is reduced. It is worth noting, though, that the actual torque output on certain models of torque-control motor was higher than the preset torque and the reported torque at fracture for several rotary les (40), which limits the effectiveness of using those motors for preventing le damage. It is also noteworthy that some studies suggested that as soon as the NiTi material enters the superelastic plateau where the stress-induced phase transformation from austenite to martensite begins, the material is subjected to an accelerated rate of crack propagation in low cycle fatigue (41, 42). To maximize the fatigue life of the endodontic le, the elastic limit should be redened as the load at which the phase transformation of NiTi material starts.

In this study, the impact of helix angle, taper, and ute length on mechanical performances of V-Taper endodontic le designs was quantitatively assessed through parametric studies with numerical simulations. Improvements on torsional stiffness and bending exibility were attained through modication of geometric parameters will be optimized. The elastic limit of the V-Taper le tip section was also estimated in both bending and torsion. A design methodology for achieving improved mechanical performances was proposed.

Acknowledgments
The authors deny any conicts of interest.

References
1. Walia HM, Brantley WA, Gerstein H. An initial investigation of the bending and torsional properties of Nitinol root canal les. J Endod 1988;14:34651. 2. Canalda-Sahli C, Brau-Aguade E, Berastegui-Jimeno E. A comparison of bending and torsional properties of K-les manufactured with different metallic alloys. Int Endod J 1996;29:1859. 3. Esposito PT, Cunningham CJ. A comparison of canal preparation with nickeltitanium and stainless steel instruments. J Endod 1995;21:1736. 4. Schafer E, Lohmann D. Efciency of rotary nickel-titanium FlexMaster instruments compared with stainless steel hand K-Flexole: part 1shaping ability in simulated curved canals. Int Endod J 2002;35:50513. 5. Schafer E, Schulz-Bongert U, Tulus G. Comparison of hand stainless steel and nickel titanium rotary instrumentation: a clinical study. J Endod 2004;30:4325. 6. Sattapan B, Nervo GJ, Palamara JE, et al. Defects in rotary nickel-titanium les after clinical use. J Endod 2000;26:1615. 7. Parashos P, Gordon I, Messer HH. Factors inuencing defects of rotary nickeltitanium endodontic instruments after clinical use. J Endod 2004;30:7225. 8. Parashos P, Messer HH. Questionnaire survey on the use of rotary nickel-titanium endodontic instruments by Australian dentists. Int Endod J 2004;37:24959. 9. Vaudt J, Bitter K, Neumann K, et al. Ex vivo study on root canal instrumentation of two rotary nickel-titanium systems in comparison to stainless steel hand instruments. Int Endod J 2009;42:2233. 10. Pruett JP, Clement DJ, Carnes DL Jr. Cyclic fatigue testing of nickel-titanium endodontic instruments. J Endod 1997;23:7785. 11. Barbosa FO, Gomes JA, de Araujo MC. Fractographic analysis of K3 nickel-titanium rotary instruments submitted to different modes of mechanical loading. J Endod 2008;34:9948. 12. Ugur I, Nihan G. Deformation and fracture of Mtwo rotary nickel-titanium instruments after clinical use. J Endod 2009;35:13969. 13. Gianluca P, Nicola MG, Massimo C, et al. A review of cyclic fatigue testing of NickelTitanium rotary instruments. J Endod 2009;35:146976. 14. Xu X, Eng M, Zheng Y, et al. Comparative study of torsional and bending properties for six models of nickel-titanium root canal instruments with different crosssections. J Endod 2006;32:3725. 15. Schafer E, Dzepina A, Danesh G. Bending properties of rotary nickel-titanium instruments. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:75763. 16. Turpin YL, Chagneau F, Vulcain JM. Impact of two theoretical cross-sections on torsional and bending stresses of nickel-titanium root canal instrument models. J Endod 2000;26:4147.

Finite Element Analysisassisted Endodontic File Design In previous sections we evaluated the impact of geometric parameters on the mechanical behavior of endodontic les with nite element models. We also estimated the elastic limit of the les under different modes of deformation. Integrating these 2 pieces of work together, we developed a methodology that provides a means to guide the le design process to achieve improved mechanical performances. As illustrated in Fig. 3, the design process starts by specifying the radial and longitudinal dimensions of the le. It is followed by parallel efforts to design the taper, helix, and cross-sectional prole. Once a prototype
1556
He and Ni

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
17. Berutti E, Chiandussi G, Gaviglio I, et al. Comparative analysis of torsional and bending stresses in two mathematical models of nickel-titanium rotary instruments: ProTaper versus ProFile. J Endod 2003;29:159. 18. Tripi TR, Bonaccorso A, Condorelli GG. Cyclic fatigue of different nickel-titanium endodontic rotary instruments. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:e10614. 19. Melo MC, Pereira ES, Viana AC, et al. Dimensional characterization and mechanical behaviour of K3 rotary instruments. Int Endod J 2008;41:32938. 20. Camara AS, de Castro Martins R, Viana ACD, et al. Flexibility and torsional strength of ProTaper and ProTaper Universal rotary instruments assessed by mechanical tests. J Endod 2009;35:1136. 21. Kim HC, Cheung GS, Lee CJ, et al. Comparison of forces generated during root canal shaping and residual stresses of three nickel-titanium rotary les by using a threedimensional nite-element analysis. J Endod 2008;34:7437. 22. Necchi S, Taschieri S, Petrini L, et al. Mechanical behaviour of nickel-titanium rotary endodontic instruments in simulated clinical conditions: a computational study. Int Endod J 2008;41:93949. 23. Petrini L, Necchi S, Taschieri S, et al. Numerical study on the inuence of material characteristics on Ni-Ti endodontic instrument performance. Journal of Materials Engineering and Performance 2009;18:6317. 24. Kim HC, Kim HJ, Lee CJ, et al. Mechanical response of nickel-titanium instruments with different cross-sectional designs during shaping of simulated curved canals. Int Endod J 2009;42:593602. 25. Kim TO, Cheung GS, Lee JM, et al. Stress distribution of three NiTi rotary les under bending and torsional conditions using a mathematic analysis. Int Endod J 2009;42: 1421. 26. Kim HC, Yum J, Hur B, et al. Cyclic fatigue and fracture characteristics of ground and twisted Nickel-Titanium rotary les. J Endod 2010;36:14752. 27. Chen W- C. Applying the nite element method to drill design based on drill deformations. Finite Elements in Analysis and Design 1997;26:5781. 28. Selvam SVM, Sujatha C. Twist drill deformation and optimum drill geometry. Computers & Structures 1995;57:90314. 29. da Silva FM. Kobayashi C, Suda H. Analysis of forces developed during mechanical preparation of extracted teeth using RaCe rotary instruments and ProFiles. Int Endod J 2005;38:1721. 30. Grande NM, Plotino G, Pecci R, et al. Cyclic fatigue resistance and three-dimensional analysis of instruments from two nickel-titanium rotary systems. Int Endod J 2006; 39:75563. 31. Bryant ST, Dummer PM, Pitoni C, et al. Shaping ability of .04 and .06 taper ProFile rotary nickel-titanium instruments in simulated root canals. Int Endod J 1999;32: 15564. 32. Schafer E, Vlassis M. Comparative investigation of two rotary nickel-titanium instruments: ProTaper versus RaCepart 2: cleaning effectiveness and shaping ability in severely curved root canals of extracted teeth. Int Endod J 2004;37: 23948. 33. Yang GB, Zhou XD, Zhang H, et al. Shaping ability of progressive versus constant taper instruments in simulated root canals. Int Endod J 2006;39:7919. 34. Shen Y, Cheung GS, Peng B, et al. Defects in nickel-titanium instruments after clinical use: part 2fractographic analysis of fractured surface in a cohort study. J Endod 2009;35:1336. 35. Low D, Ho AW, Cheung GS, et al. Mathematical modeling of exural behavior of rotary nickel-titanium endodontic instruments. J Endod 2006;32:5458. 36. Gambarini G. Rationale for the use of low-torque endodontic motors in root canal instrumentation. Endod Dent Traumatol 2000;16:95100. 37. Schafer E, Diez C, Hoppe W, et al. Roentgenographic investigation of frequency and degree of canal curvatures in human permanent teeth. J Endod 2002;28: 2116. 38. He R. Modeling, analysis, and experimental investigation of root canal instrumentation (PhD dissertation). Ann Arbor, MI: University of Michigan; 2007. 39. Yared GM, Kulkarni GK. Failure of ProFile Ni-Ti instruments used by an inexperienced operator under access limitations. Int Endod J 2002;35:53641. 40. Yared G, Kulkarni GK. Accuracy of the DTC torque control motor for nickel-titanium rotary instruments. Int Endod J 2004;37:399402. 41. Dauskardt RH, Duerig TW, Ritchie RO. Effects of in situ phase transformation on fatigue-crack propagation in Titanium-Nickel shape-memory alloys. In: Proceedings of the MRS International Meeting on Advanced Materials, May 31-June 3, 1988, Tokyo, Japan:243249. 42. McKelvey AL, Ritchie RO. Fatigue-crack propagation in Nitinol, a shape-memory and superelastic endovascular stent material. Journal of Biomedical Materials Research 1999;47:3018.

JOE Volume 36, Number 9, September 2010

Design Improvement and Failure Reduction of Endodontic Files through Finite Element Analysis

1557

Basic ResearchTechnology

Tissue Dissolution by Sodium Hypochlorite: Effect of Concentration, Temperature, Agitation, and Surfactant
Sonja Stojicic, DDS, MSc,* Slavoljub Zivkovic, DDS, PhD, Wei Qian, DDS, PhD,* Hui Zhang, DDS, PhD,* and Markus Haapasalo, DDS, PhD*
Abstract
Aim: Sodium hypochlorite is the most commonly used endodontic irrigant because of its antimicrobial and tissue-dissolving activity. The aim of this study was to evaluate and compare the effects of concentration, temperature, and agitation on the tissue-dissolving ability of sodium hypochlorite. In addition, a hypochlorite product with added surface active agent was compared with conventional hypochlorite solutions. Methods: Three sodium hypochlorite solutions from two different manufacturers in concentrations of 1%, 2%, 4%, and 5.8% were tested at room temperature, 37 C, and 45 C with and without agitation by ultrasonic and sonic energy and pipetting. Distilled and sterilized tap water was used as controls. Pieces of bovine muscle tissue (68 3 mg) were placed in 10 mL of each solution for ve minutes. In selected samples, agitation was performed for one, two, or four 15-second periods per each minute. The tissue specimens were weighed before and after treatment, and the percentage of weight loss was calculated. The contact angle on dentin of the three solutions at concentrations of 1% and 5.8% was measured. Results: Weight loss (dissolution) of the tissue increased almost linearly with the concentration of sodium hypochlorite. Higher temperatures and agitation considerably enhanced the efcacy of sodium hypochlorite. The effect of agitation on tissue dissolution was greater than that of temperature; continuous agitation resulted in the fastest tissue dissolution. Hypochlorite with added surface active agent had the lowest contact angle on dentin and was most effective in tissue dissolution in all experimental situations. Conclusions: Optimizing the concentration, temperature, ow, and surface tension can improve the tissue-dissolving effectiveness of hypochlorite even 50-fold. (J Endod 2010;36:15581562)

Key Words
Agitation, Chlor-Xtra, sodium hypochlorite, surfactant, temperature, tissue dissolution

uccess in endodontic treatment depends to a great extent on chemomechanical debridement of the canals. Although instruments remove most of the canal contents in the main root canal area, irrigation plays an indispensable role in all areas of the root canal system, in particular those parts that are inaccessible for instrumentation (1). The most favorable features of irrigants are their ushing action, tissue-dissolving ability, antimicrobial effect, and low toxicity (2, 3). Sodium hypochlorite is the most commonly used endodontic irrigant because of its well-known antimicrobial and tissue-dissolving activity (46). The dissolving capability of sodium hypochlorite relies on its concentration, volume, and contact time of the solution but also on the surface area of the exposed tissue (7). However, high concentrations are potentially toxic for periapical tissue (810). Also, changes in mechanical properties such as decreased microhardness and increased roughness of radicular dentin have been reported after exposure to sodium hypochlorite in concentrations of 2.5% and 5.25% (11). Possible ways to improve the efcacy of hypochlorite preparations in tissue dissolution are increasing the pH (12) and the temperature of the solutions, ultrasonic activation, and prolonged working time (13). Although there is a general consensus that increased temperature enhances the effectiveness of hypochlorite solutions, there are only a few published articles about this (1416). It has been suggested that preheating low-concentration solutions improves their tissue-dissolving capacity with no effect on their short-term stability. Also, systemic toxicity is lower compared with the higher-concentration solutions (at a lower temperature) with the same efcacy (15). The impact of mechanical agitation of the hypochlorite solutions on tissue dissolution was found to be very important by Moorer and Wesselink (7) who emphasized the great impact of violent uid ow and shearing forces caused by ultrasound on the ability of hypochlorite to dissolve tissue. However, the mechanisms involved are not completely understood (13). Despite several separate reports of the various ways to improve the effectiveness of tissue dissolution by sodium hypochlorite, the relative importance of temperature, concentration, and agitation remains unclear. In the present study, all these factors were examined under controlled conditions to allow comparison of their role. Finally, a hypochlorite product with an added surface active agent was compared with conventional products in the different experimental settings.

From the *Division of Endodontics, Department of Oral Biological & Medical Sciences, Faculty of Dentistry, University of British Columbia, Vancouver, Canada; and Department for Restorative Dentistry and Endodontics, Faculty of Dentistry, University of Belgrade, Belgrade, Serbia. Address requests for reprints to Dr Markus Haapasalo, Division of Endodontics, Oral Biological & Medical Sciences, UBC Faculty of Dentistry, 2199 Wesbrook Mall, Vancouver, BC, Canada V6T 1Z3. E-mail address: markush@interchange.ubc.ca. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.021

Materials and Methods

Solutions Sodium hypochlorite solutions in concentrations of 1%, 2%, 4%, and 5.8% were tested. Stock solution of 6 % sodium hypochlorite (Regular 1; EMD Chemicals Inc, Gibbstown, NJ) and 5.8% sodium hypochlorite (Regular 2; Inter-Med, Inc/Vista Dental Products, Racine, WI) were obtained from the manufacturers. Two different hypochlorite products with 5.8% sodium hypochlorite were included in the experiments: one conventional solution (Regular 2, Vista-Dental) and one with a surface active agent added (Chlor-Xtra, Vista-Dental). The amount of available chlorine was obtained by the manufacturers. The solutions were kept at 4 C following the recommendations of the manufacturer and brought to room temperature (RT) before use. One percent,

1558

Stojicic et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
2%, 4%, and 5.8% solutions of sodium hypochlorite were prepared by diluting the stock solution in distilled water. Distilled and sterilized tap water was used as controls. The specimens were weighed on an electronic balance (FX-300; A&D Company, Ltd, Tokyo, Japan) before the hypochlorite treatments and placed each in 10 mL of preheated hypochlorite solution in separate beakers in the temperature-controlled water bath. Five parallel samples per group were included in all experiments. After 5 minutes in the solution, the samples were blotted dry and weighed again. The percentage of weight loss was calculated.

Dissolution of Tissue Bovine meat was used as a tissue sample in the experiment. It was kept frozen at 15 C in 100% humidity. Frozen tissue was cut into pieces of 4 4 2 mm using stainless steel blade. Because the surface area has a great impact on the tissue dissolution, each sample had a similar size and shape. The samples had an original weight of 68 3 mg with no signicant difference between groups. The experiments were done at RT and at 37 C and 45 C. A water bath (Water Bath Digital 10L; Fisher Scientic, Ottawa, Ontario, Canada) was used for the experiments at 37 C and 45 C. The temperature of the solutions was conrmed using a thermometer (Fisher Scientic). Three different means of agitation were tested: ultrasonic, sonic, and pipetting. An ultrasonic system (Varios Lux 350; NSK, Kanuma, Japan) with a slender steel tip (E 7) at power setting 4 was used to deliver ultrasonic energy. Sonic vibration at 10,000 cpm was applied by EndoActivator (Dentsply, Tulsa, OK) using a exible polymer tip size 25/04 (medium). The tips (ultrasonic and sonic) were immersed in the hypochlorite solutions into a depth of 10 mm, 5 mm away from the tissue specimens, without touching them. A transfer pipette (grad, 5.8 mL, Fisher Scientic) was used for mechanical agitation of the hypochlorite solutions; the tip of the pipette was at 5 mm distance from the tissue specimen. Agitation was performed 15 seconds per each minute during the 5-minute incubation period. The three hypochlorite solutions in selected concentrations were tested at RT and at 45 C with and without agitation. In addition, the effect of duration of agitation by pipetting was studied using the 2% and 5.8% solutions of the Regular 2 hypochlorite solution (Vista-Dental) at RT as follows: the tissue specimens were treated for 5 minutes in the hypochlorite solutions, pipetting was done as above either for one or two 15-second periods per each minute, or continuously throughout the 5-minute experiment.

Contact Angle Measurement Extracted human maxillary canine and mandibular premolars were used to prepare the dentin surfaces for contact angle measurements. After cutting off the crown and apical third of the root, each tooth was split in half labiopalatally using a low-speed diamond saw. Each cut surface was polished using a series of abrasive papers (CarbiMet; Buehler, Lake Bluff, IL) in the following sequence (120/P120, 180/P180, 240/P280, 320/P400, 400/P800, and 600/P1200). A 1.5-mL droplet of 1% and 5.8% hypochlorite solutions or distilled water (control) was placed on coronal root dentin using a 2-mL pipette. The contact angle was measured within 30 seconds using a NRL Contact Angle Goni ometer (Rame-hart, Netcong, NJ). Six parallel measurements were performed with each solution on dentin surfaces of both teeth. Data Analysis Weight loss was expressed as mean value standard deviation of the percentage of the tissue weight loss. Data were analyzed using oneway analysis of variance followed by the Tukey post hoc test for multiple comparisons (SPSS Inc, Chicago, IL). Statistical signicance was considered at p < 0.05.

Results
Tissue weight loss after 5 minutes in different concentrations of sodium hypochlorite at three different temperatures is shown in Table 1. Weight loss increased with increasing concentrations of sodium hypochlorite. A signicant difference in weight loss was observed after exposure to 2% Chlor-Xtra and 4% and 5.8% (p <

TABLE 1. Tissue Weight Loss (% Standard Deviation) After 5 Minutes of Exposure to Three Sodium Hypochlorite Products at Different Concentrations and Temperatures % of weight loss
Distilled water (control) Sterilized water (control) 1% Regular 1 Regular 2 Chlor-Xtra 2% Regular 1 Regular 2 Chlor-Xtra 4% Regular 1 Regular 2 Chlor-Xtra 5.8% Regular 1 Regular 2 Chlor-Xtra

RT
0.87 1.59 2.26 1.35 3.21 1.85 5.52 1.38 6.51 3.05 4.41 1.79 3.95 2.04 7.75 2.08 20.52 3.00 18.13 2.22 28.19 3.27, 29.93 2.24 27.28 4.47 41.55 5.22,

37 C
1.40 2.33 2.36 1.14 0.27 2.23 0.29 1.24 1.19 1.26 10.72 2.30 8.25 2.30* 13.07 0.80 26.63 3.13 24.93 1.32 37.90 5.77,k 40.68 2.64 38.31 3.29 59.05 6.20,k

45 C
10.91 1.71 8.26 1.86 1.16 1.92 1.72 1.87 4.74 2.12 14.61 5.27 14.18 3.45 21.10 2.16*, 36.26 2.95 33.61 3.58 49.17 5.20,k 49.10 6.52 48.18 4.45 67.28 8.80,k

RT, room temperature; regular 1, sodium hypochlorite (EMD Chemicals Inc); regular 2, sodium hypochlorite (Inter-Med, Inc./Vista Dental Products); Chlor-Xtra, sodium hypochlorite with added surface active agent (Inter-Med, Inc./Vista Dental Products). *p < 0.05. p < 0.001 versus distilled water. p < 0.05. p < 0.01. k p < 0.001 versus regular 1.

JOE Volume 36, Number 9, September 2010

Tissue Dissolution by NaClO

1559

Basic ResearchTechnology
0.001) of Regular 1 and Regular 2 sodium hypochlorite compared with controls at RT. Tissue specimens immersed in 1% sodium hypochlorite at RT increased in weight after 5 minutes. Heating the hypochlorite solutions greatly increased tissue dissolution; the magnitude of the increase compared with RT varied from 30% to 300% depending on the concentration, temperature, and type of hypochlorite (Table 1). Of the three solutions, the one with added surface active agent dissolved signicantly more tissue than the two other solutions in all temperature/concentration groups (Table 1). The effect of the three different methods of agitation on the tissuedissolving ability of 2% and 5.8% sodium hypochlorite is shown in Table 2. The tissue weight loss was signicantly higher at both tested temperatures when sodium hypochlorite solutions were agitated than without agitation (p < 0.001 for 2% and p < 0.01 for 5.8% sodium hypochlorite, Table 2). Under agitation at RT, tissue weight loss was signicantly higher with 5.8% Chlor-Xtra than with both 5.8% regular hypochlorite products (Table 2). Agitation experiments with an increased time of active agitation from 25% to 100% showed continuous increase in tissue dissolution as the relative time of agitation increased (Table 3). The increase of tissue dissolution from no agitation to 100% agitation (pipette) in Regular 2 hypochlorite at RT was 12.7-fold with 2% and 2.1-fold with 5.8% solution. Tissue weight loss was signicantly higher after simultaneous action of temperature and agitation than by either one alone (Tables 1-3). Chlor-Xtra (5.8%) had the lowest contact angle of the three hypochlorite solutions. There was no signicant difference in contact angle between the 1% solutions (p > 0.05) (Table 4).
73.004.12 71.685.90 87.937.13, 35.94 3.02* 39.16 3.60 48.91 7.62 49.10 6.52 48.18 4.45 67.28 8.80 74.99 7.17 75.51 7.98 81.38 2.88 69.964.78 68.827.09 78.262.98* 37.312.25 33.952.37 40.901.80 20.56 1.67 17.80 0.85 20.65 1.68 14.61 5.27 14.18 3.45 21.10 2.16 29.85 2.89 29.15 3.66 38.37 3.72 29.011.05 26.071.06 35.175.47 11.912.39 11.041.74 2.18 1.44 1.48 1.06 10.91 1.71 8.26 1.86 10.50 4.38 10.03 3.55 10.903.71 8.582.06

TABLE 2. Effect of Three Different Methods of Agitation on Tissue Dissolution (% Tissue Weight Loss Standard Deviation) by the Three Hypochlorite Products

45 C P No agitation US

EA

Discussion
A great number of studies have focused on the tissue-dissolving ability of sodium hypochlorite. It has been found that the solvent capability of sodium hypochlorite depends on its concentration; time; volume; pH; temperature; agitation; and the type, amount, and surface area of the tissue (2, 5, 7, 12). However, great variations among these factors contribute to the difculty of making comparisons between different studies and the relative importance of each factor (17). The present study evaluated the effect of concentration, temperature, and agitation on sodium hypochlorite ability to dissolve organic material in a standardized setting. Tissues from a number of different sources have been used in studies about sodium hypochlorite tissue-dissolving ability (6). Porcine muscle tissue (1, 12), rabbit liver (7), rat connective tissue (5), pig palatal mucosa (18), bovine muscle tissue (2), and bovine pulp (3) have been used to determine dissolution ability of different irrigants. The reasons for using different tissue instead of dental pulp have been availability and easier standardization of the surface area of each specimen (2). Tissue specimens used in the present study were prepared from bovine muscle tissue with a standardized weight of 68 3 mg. The meat specimens were cubical in shape (4 4 2 mm) giving an equal surface area. Pilot experiments had shown that it was difcult to determine the endpoint of complete dissolution of the tissue because of a great number of bubbles (result of saponication reaction); therefore, xed time was used instead, and the samples were weighed before and after exposure. Other methods have used different approaches (eg, measuring the changes in the solutions, such as the amount of available chlorine in the solution after completed dissolution [7] or the amount of hydroxyproline in the residual tissue after incubation with the solution [19]). Previous studies have shown that the tissue-dissolving ability of sodium hypochlorite solution decreases if it is diluted (2, 5, 16). Results from the present study showed that 5.8% sodium

12.76 1.63 12.83 2.47 16.25 0.88 4.41 1.79 3.95 2.04 7.75 2.08 11.76 1.95 11.393.98 16.96 3.28

RT

0.87 1.59 2.26 1.35

No agitation

29.93 2.24 27.28 4.47 41.55 5.22

36.28 3.89* 34.00 3.41 48.03 6.48

US, ultrasound; EA, endoactivator; P, pipetting. *p < 0.05. p < 0.01. p < 0.001 versus no agitation. p < 0.001 versus regular 1.

1560

Distilled water Sterilized tap water 2% Reg 1 Reg 2 Chlor-Xtra 5.8% Reg 1 Reg 2 Chlor-Xtra

2.35 1.26 1.78 1.19

US

34.19 3.52 32.76 3.38 47.76 6.14

1.86 1.33 2.67 2.08

EA

Stojicic et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 3. Effect of Agitation by Pipetting for 0%, 25%, 50%, and 100% of the 5-Minute Exposure Time on Tissue Dissolution (% Standard Deviation) by 2% and 5.8% Regular (Reg 2) Hypochlorite at Room Temperature Agitation/no. of agitation per each minute Solutions
2% sodium hypochlorite 5.8% sodium hypochlorite
*p < 0.05. p < 0.001 versus no agitation.

No agitation
2.71 1.68 29.94 2.31

15/45
14.51 2.07 39.11 4.44*

15/15/15/15
24.56 2.37 50.36 2.86

Continuous
34.68 3.84 63.14 6.90

hypochlorite was the most effective in agreement with the previous studies. Interestingly, tissue weight loss after 5 minutes in 2% of sodium hypochlorite solution was equal to controls with distilled and sterilized tap water. It should be emphasized, however, that the weight loss in water did not change after 5 minutes, whereas it continued to increase in 2% sodium hypochlorite after the rst 5minute period (results not shown). Tissue specimens immersed in 1% solutions increased in weight during the 5-minute exposure. This can probably be explained by hydration of the tissue after the initialdissolving action of sodium hypochlorite on the meat. Other studies have also reported a similar effect on tissue by low-concentration hypochlorite solutions (5). However, longer exposure times will cause weight loss also with mild hypochlorite solutions. The present study showed that heating the sodium hypochlorite solution enhanced its ability to dissolve organic material. These results are in accordance with other studies (1416). Kamburis et al (20) found that heated sodium hypochlorite solutions were more effective in removing organic debris from dentin shavings than unheated solutions at the same concentration. It has also been shown that sodium hypochlorite solutions were stable for a period of 4 hours when heated to 37 C (21). Sirtes et al (15) found that 1%, 2.62%, and 5.25% solutions had an unchanged quantity of available chlorine for 1 hour at 45 C and 60 C, respectively. Therefore, in the present study, all experiments were nished in less than 1 hour to exclude the possibility of the loss of activity of sodium hypochlorite during experimentation at higher temperatures. The effect of the surface active agent to hypochlorite was rst shown by Cameron (22) who showed that the addition of the surface modiers enhanced the ability of sodium hypochlorite to dissolve organic material. Clarkson et al (6) tested the dissolution ability of three different brands of sodium hypochlorite available in Australia and reported that the products with surfactants dissolved porcine pulp in a shorter time than regular sodium hypochlorite at the same concentration. The present study not only conrmed these ndings but also showed that the advantage over regular products remained also when the products were diluted, heated, or agitated. Although not known in detail, it is possible that the better performance of Chlor-Xtra over regular hypochlorite is based partly on better contact on the surface of the tissue as well as faster exchange with fresh solution, both facili-

tated by the surface active agent. The lower contact angle of 5.8% ChlorXtra as compared with the two regular hypochlorites supports the assumption of better wetting of the surface of the substrate. When diluted to 1%, there was no signicant difference in contact angle between the products, obviously because of the higher proportion of the water. The importance of agitation on the tissue-dissolving ability of sodium hypochlorite has been reported, but there are very few studies on the effect of agitation on this (7). Aqueous solution of sodium hypochlorite is a dynamic balance of sodium hydroxide and hypochlorous acid. When sodium hypochlorite is in contact with organic material, sodium hydroxide reacts with fatty acids creating soap and glycerol, which is known as saponication reaction. It also reacts with amino acids creating salt and water (neutralization). In addition, hypochlorous acid acts with amino acids creating chloramine and water. These reactions, which happen mostly at the surface, lead to liquefaction of the organic tissue (23). At the same time, molecules of sodium hypochlorite involved in the reactions are consumed, resulting in the decline of local activity. It is therefore important to supply active hypochlorite to the area and also remove the remnants of dissolved tissue. In this study, the effect of three different agitation methods on the tissue-dissolving ability of sodium hypochlorite was investigated. It was found that the agitation of the solution improved their dissolving ability; however, there were no signicant differences between the three methods of agitation. The use of ultrasound energy in root canal preparation and irrigation has for a long time been a matter for debate. Although some researchers have reported great success of ultrasonics in root canal cleaning (2428), others did not nd ultrasound superior compared with root canal irrigation using the syringe (2931). The mechanism of passive ultrasonic action has been attributed to acoustic streaming (microstreaming) and cavitation (13, 32). A sonic action has a similar mechanism of action to ultrasonic although the pattern of the oscillating le is different. However, according to Lumley et al (33), cavitation is limited to the distance of less than 100 mm. All experiments in the present study were performed in the beaker with the ultrasonic/sonic tip immersed in the solutions to a depth of 10 mm, 5 mm away from the tissue specimen without touching it. Therefore, the possibility to evaluate the direct impact of ultrasonic or sonic energy by cavitation was excluded. However, acoustic streaming developed by this

TABLE 4. Contact Angle (Mean Standard Deviation) on Dentin of Three Sodium Hypochlorite Solutions (1% and 5.8%) 1% Distilled water
39.33 12.35

5.8% Chlor-Xtra
41.42 9.59

Reg 1
52.17 13.58

Reg 2
44.83 12.33

Reg 1
71.92 12.76

Reg 2
54.58 13.91*,

Chlor-Xtra
36.25 8.25

*p < 0.05. p < 0.001 versus distilled water. p < 0.01. p < 0.001 versus 5.8% Reg 1.

JOE Volume 36, Number 9, September 2010

Tissue Dissolution by NaClO

1561

Basic ResearchTechnology
mechanism creates stirring action and rapid movements of the liquid further away from the energy source. This is likely to be the mechanism by which ultrasonic affects cleaning of the peripheral parts in root canal in vivo. Within the limitations of the present study, all three methods of agitation improved tissue dissolution. In clinical endodontics penetration of irrigants to the most apical canal and to other peripheral area such as ns and webs remains a great challenge, and both ultrasonic and sonic agitation may in that environment have advantages which could not be demonstrated in the present study design. In conclusion, an increase in concentration and temperature of sodium hypochlorite greatly increased its efcacy in tissue dissolution. Refreshing the hypochlorite solution at the site of dissolution by agitation, preferably continuous, also resulted in a marked increase of hypochlorite effect. High temperature and agitation had an additive effect on the tissue dissolution. The sodium hypochlorite product with added surface active agent was the most effective in tissue dissolution at all concentrations and temperatures.
13. Zehnder M. Root canal irrigants. J Endod 2006;32:38998. 14. Rossi-Fedele G, De Figueiredo JA. Use of a bottle warmer to increase 4% sodium hypochlorite tissue dissolution ability on bovine pulp. Aust Endod J 2008;34: 3942. 15. Sirtes G, Waltimo T, Schaetzle M, et al. The effects of temperature on sodium hypochlorite short-term stability, pulp dissolution capacity, and antimicrobial efcacy. J Endod 2005;31:66971. 16. Abou-Rass M, Oglesby SW. The effects of temperature, concentration, and tissue type on the solvent ability of sodium hypochlorite. J Endod 1981;7:3767. 17. Beltz RE, Torabinejad M, Pouresmail M. Quantitative analysis of the solubilizing action of MTAD, sodium hypochlorite, and EDTA on bovine pulp and dentin. J Endod 2003;29:3347. 18. Naenni N, Thoma K, Zehnder M. Soft tissue dissolution capacity of currently used and potential endodontic irrigants. J Endod 2004;30:7857. 19. Koskinen KP, Stenvall H, Uitto VJ. Dissolution of bovine pulp tissue by endodontic solutions. Scand J Dent Res 1980;88:40611. 20. Kamburis JJ, Barker TH, Bareld RD, et al. Removal of organic debris from bovine dentin shavings. J Endod 2003;29:55961. 21. Frais S, Ng YL, Gulabivala K. Some factors affecting the concentration of available chlorine in commercial sources of sodium hypochlorite. Int Endod J 2001;34: 20615. 22. Cameron JA. The effect of a uorocarbon surfactant on the surface tension of the endodontic irrigant, sodium hypochlorite. A preliminary report. Aust Dent J 1986;31:3648. 23. Estrela C, Estrela CR, Barbin EL, et al. Mechanism of action of sodium hypochlorite. Braz Dent J 2002;13:1137. 24. Sabins RA, Johnson JD, Hellstein JW. A comparison of the cleaning efcacy of shortterm sonic and ultrasonic passive irrigation after hand instrumentation in molar root canals. J Endod 2003;29:6748. 25. Cameron JA. The synergistic relationship between ultrasound and sodium hypochlorite: a scanning electron microscope evaluation. J Endod 1987;13:5415. 26. Ahmad M, Pitt Ford TJ, Crum LA. Ultrasonic debridement of root canals: acoustic streaming and its possible role. J Endod 1987;13:4909. 27. Ahmad M, Pitt Ford TR, Crum LA. Ultrasonic debridement of root canals: an insight into the mechanisms involved. J Endod 1987;13:93101. 28. Weller RN, Brady JM, Bernier WE. Efcacy of ultrasonic cleaning. J Endod 1980;6: 7403. 29. Mayer BE, Peters OA, Barbakow F. Effects of rotary instruments and ultrasonic irrigation on debris and smear layer scores: a scanning electron microscopic study. Int Endod J 2002;35:5829. 30. Cymerman JJ, Jerome LA, Moodnik RM. A scanning electron microscope study comparing the efcacy of hand instrumentation with ultrasonic instrumentation of the root canal. J Endod 1983;9:32731. 31. Tauber R, Morse DR, Sinai IA, et al. A magnifying lens comparative evaluation of conventional and ultrasonically energized ling. J Endod 1983;9:26974. 32. van der Sluis LW, Versluis M, Wu MK, et al. Passive ultrasonic irrigation of the root canal: a review of the literature. Int Endod J 2007;40:41526. 33. Lumley PJ, Walmsley AD, Laird WR. Streaming patterns produced around endosonic les. Int Endod J 1991;24:2907.

References
1. Hasselgren G, Olsson B, Cvek M. Effects of calcium hydroxide and sodium hypochlorite on the dissolution of necrotic porcine muscle tissue. J Endod 1988;14:1257. 2. Turkun M, Cengiz T. The effects of sodium hypochlorite and calcium hydroxide on tissue dissolution and root canal cleanliness. Int Endod J 1997;30:33542. 3. Okino LA, Siqueira EL, Santos M, et al. Dissolution of pulp tissue by aqueous solution of chlorhexidine digluconate and chlorhexidine digluconate gel. Int Endod J 2004;37:3841. 4. Rosenfeld EF, James GA, Burch BS. Vital pulp tissue response to sodium hypochlorite. J Endod 1978;4:1406. 5. Hand RE, Smith ML, Harrison JW. Analysis of the effect of dilution on the necrotic tissue dissolution property of sodium hypochlorite. J Endod 1978;4:604. 6. Clarkson RM, Moule AJ, Podlich H, et al. Dissolution of porcine incisor pulps in sodium hypochlorite solutions of varying compositions and concentrations. Aust Dent J 2006;51:24551. 7. Moorer WR, Wesselink PR. Factors promoting the tissue dissolving capability of sodium hypochlorite. Int Endod J 1982;15:18796. 8. Mehra P, Clancy C, Wu J. Formation of a facial hematoma during endodontic therapy. J Am Dent Assoc 2000;131:6771. 9. Gernhardt CR, Eppendorf K, Kozlowski A, et al. Toxicity of concentrated sodium hypochlorite used as an endodontic irrigant. Int Endod J 2004;37:27280. 10. Barnhart BD, Chuang A, Lucca JJ, et al. An in vitro evaluation of the cytotoxicity of various endodontic irrigants on human gingival broblasts. J Endod 2005;31:6135. 11. Ari H, Erdemir A, Belli S. Evaluation of the effect of endodontic irrigation solutions on the microhardness and the roughness of root canal dentin. J Endod 2004;30:7925. 12. Christensen CE, McNeal SF, Eleazer P. Effect of lowering the pH of sodium hypochlorite on dissolving tissue in vitro. J Endod 2008;34:44952.

1562

Stojicic et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Detection of Dentinal Cracks after Root-end Resection: An Ex Vivo Study Comparing Microscopy and Endoscopy with Scanning Electron Microscopy
Thomas von Arx, Prof.Dr.med.dent.,* Renato Kunz, Dr.med.dent., Adrienne Christina Schneider, Dr.med.dent.,* Walter Burgin, med.ing.,* and Adrian Lussi, Prof.Dr.med.dent.
Abstract
Introduction: Dentinal cracks are occasionally observed at the cut root face after root-end resection in apical surgery. The objective of this ex vivo study was to evaluate and compare the efciency of visual aids to identify root-end dentinal cracks. Methods: Twenty-six extracted human molars were decoronated, and the root canals were instrumented and lled. The apical 3 mm of the roots were resected, and the cut root faces were assessed with microscopy at 16 and 24 magnication and with endoscopy at 8 and 64 magnication (four visual aids). Roots were then duplicated for inspection with scanning electron microscopy. The presence, type, and location of cracks were registered by a blinded observer, with the scanning electron microcopy serving as the reference. The percentages of correct identication of dentinal cracks were then statistically compared among the four test congurations. Results: Endoscopy 64 showed the highest sensitivity for crack identication, irrespective of the applied methodology (ie, per root and per crack). However, higher scores of false-positive cracks (lower specicity) were found with endoscopy 64 than with the other tested visual aids. The correct detection and location of complete canal cracks (55.3%, 52.6%, 68.4%, and 78.9%) were higher than the detection of incomplete canal cracks (42.2%, 42.2%, 52.0%, and 64.7%) using the four tested visual aids (microscopy at 16 and 24 magnication and endoscopy at 8 and 64 magnication, respectively). Only one of ve intradentin cracks was identied with endoscopy 64. Conclusions: Overall, endoscopy 64 proved the most accurate visual aid for the identication of dentinal cracks after root-end resection in extracted human teeth; however, it also provided the most false identications. (J Endod 2010;36:1563 1568)

Key Words
Apical surgery, dentinal crack, endoscopy, microscopy, root end

he ideal goal of apical surgery is to create optimum conditions for healing by sealing any path from the root canal to the periradicular tissues (1). Therefore, the identication and treatment of possible pathways, such as isthmuses, accessory canals, and cracks or microfractures, is of utmost importance to avoid the recurrence of leakage with subsequent infection. The introduction of microsurgical principles almost 15 years ago, including microinstruments, well-focused illumination, and magnication, led to a realization of the importance of detecting microscopic ndings to improve the healing outcome of apical surgery (24). Working with loupes or with a surgical microscope has become widely accepted and recommended in conventional and surgical endodontics. Several authors have suggested that the use of aids to enhance vision might be decisive in achieving high success rates in apical surgery (3, 5, 6). The saying you can only treat what you can see, however, is greatly inuenced by the clinicians ability to (1) identify microscopic ndings and (2) correctly interpret these microelements at the cut root face. Hence, the accuracy of identication of microscopic ndings will play a decisive role in apical surgery. Recent experimental studies have evaluated and compared the effectiveness of different visual aids in identifying microstructures after root-end resection. An experimental study showed enhanced accuracy of crack identication in extracted human teeth with increasing magnication, but the overall accuracy ranging between 39% and 58% was lower than expected (7). Using transillumination of the root end, whether alone or in combination with a dye, was reported as a more accurate way of diagnosing root-end dentinal cracks (8). Another experimental study compared endoscopy with scanning electron microscopy for the detection of microelements at the cut root face in extracted human teeth after root-end resection (9). That research showed the correct identication of isthmuses in all teeth, but cracks were less accurately identied. The objective of the present study was to evaluate and compare the accuracy of the microscope and the endoscope for the identication of dentinal cracks after root-end resection in extracted human teeth.

Materials and Methods

The material comprised 26 extracted human molars with closed apices. None of the teeth presented with a root canal lling (teeth were radiographed mesiobuccally and

From the Departments of *Oral Surgery and Stomatology and Preventive, Restorative and Pediatric Dentistry, School of Dental Medicine, University of Bern, Bern, Switzerland. Address requests for reprints to Dr Thomas von Arx, Department of Oral Surgery and Stomatology, School of Dental Medicine, University of Bern, Freiburgstrasse 7, CH-3010 Bern, Switzerland. E-mail address: thomas.vonarx@zmk.unibe.ch. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.016

JOE Volume 36, Number 9, September 2010

Detection of Dentinal Cracks After Root-end Resection

1563

Basic ResearchTechnology
orofacially). After the removal of all periodontal tissues from the root surfaces, the teeth were stored in a saturated mineral solution (10) at 37 C. For root canal treatment, the crowns were partially removed (3 mm from the occlusal plane), and the pulps were shaped according to the balanced force technique (11). Canals were irrigated with NaOCl 3%. After drying with paper points, canal obturation was accomplished with lateral condensation using gutta-percha points and sealer (AH Plus; Dentsply DeTrey, Konstanz, Germany). Access cavities were closed with glass ionomer cement (Ketac-Fil; 3M ESPE, Seefeld, Germany). Subsequently, all teeth were stored at 37 C and 100% humidity for a minimum of 4 weeks to allow complete setting of the canal sealer. Root-end resection was performed perpendicular to the long axis of the root at a level of 3 mm from the apex using a cylindrical ssure bur under copious irrigation. In addition, a notch was prepared at the junction of the cut root face and the buccal root surface to standardize the orientation of the resection plane for the subsequent analysis. The roots were air dried and stained with 2% methylene blue dye (12). After the methylene blue was rinsed off with water, root surfaces were dried again. Subsequently, the root-end faces were examined with a surgical microscope (Moller Denta 300; Haag-Streit International, Koeniz, Switzerland) at 16 and 24 magnication and with a rigid endoscope (Hopkins Tele-Otoscope 70o; Karl Storz GmbH, Tuttlingen, Germany) at 8 and 64 magnication, yielding four different test congurations. Video sequences limited to 2 minutes per root were recorded for subsequent image analysis. The primary study parameter was the presence (or absence) of dentinal cracks at the resection plane. A crack was dened as any dark line within the resected dentinal surface that appeared to disrupt the integrity of the dentin, not including artifacts produced by the resection of the root end (8). Three different types of dentinal cracks were distinguished (13): incomplete canal cracks, complete canal cracks, and intradentin cracks. After the microscopic and endoscopic visual evaluation, root ends were duplicated for scanning electron microscopy (SEM) analysis by taking impressions with silicone (Impregum Penta Soft, 3M ESPE). Impressions were poured with resin (Epox, Struers, Birmensdorf, Switzerland) and sputter coated with gold and palladium (Sputter SCD 050; Balzers Union AG, Balzers, Liechtenstein). The SEM was determined as the reference baseline. The SEM examination was performed by a blinded observer, with the time limited to 2 minutes. SEM evaluation was done between 50 and 200 magnication (Stereoscan 200; Cambridge Instruments Ltd, Cambridge, United Kingdom). The following analyses were performed to compare the four test visual aids with SEM: (i) analysis 1: presence (or absence) and type of crack, with the root as the unit of evaluation, and (ii) analysis 2: concurrence between test visual aid and SEM, with regard to type and location of crack, using a 12-sector grid (Figs. 1-7).

Figure 1. A schematic illustration of a cut root face with various cracks. A 12sector grid was used for location of cracks. The notch was always made at the midbuccal aspect of the cut root face.

Results
During the evaluation process, three teeth had to be discarded because root-end inspection was impossible because of artifacts, root fracture, or irregular root-face cutting. This resulted in 23 teeth with a total of 52 roots to be assessed (52 roots 12 sectors/root = 624 sectors). The evaluation per sectors showed good concurrence between all test congurations and SEM, with kappa values of 0.90 (microscopy 16), 0.90 (microscopy 24), 0.87 (endoscopy 8), and 0.84 (endoscopy 64).

Statistics For comparison of the four methods (microscopy 16, microscopy 24, endoscopy 8, and endoscopy 64) with SEM, Cohen kappa values were calculated based on the sector data (14). For comparisons among the different methods, the sensitivity and specicity (for denitions, see http://en.wikipedia.org/wiki/Sensitivity_and_ specicity) of crack identication were assessed using the root data. Furthermore, the cracks percentage values were transformed with an Arcsine transformation, and their effect size was compared with Cohens critical hc (15).
1564
von Arx et al.

Figure 2. An SEM reference image (28.9 magnication) of cut root face presenting with one incomplete canal crack (upper right quadrant) and two complete canal cracks (lower quadrants). The notch on the midbuccal aspect is clearly visible. The black holes on the cut root face are caused by entrapped tiny air bubbles during impression taking for making the root replicas.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Figure 3. An image of the same cut root face obtained with a microscope (16 magnication); the two complete canal cracks in the lower quadrants can be identied, but the incomplete canal crack in the upper right quadrant is barely visible. (This gure is available in color online at www.aae.org/joe/.)

Figure 5. An image of the same cut root face obtained with an endoscope (8 magnication); all three cracks are clearly visible. (This gure is available in color online at www.aae.org/joe/.)

Analysis 1 SEM showed that 49 of 52 roots presented with a crack at the resection plane (Table 1). Twenty-eight roots had a complete canal crack, 44 roots had an incomplete canal crack, and only 5 roots had an intradentin crack. Endoscopy 64 showed the highest sensitivity of all tested visual aids with regard to crack detection for categories all cracks (95.9%), complete canal cracks (92.9%), and incomplete canal cracks (90.9%). The statistical evaluation for sensitivity was signicantly higher for endoscopy 64 than for both microscope magnications (p < 0.05). In contrast, the specicity of endoscopy 64 was lower (all cracks, 33.3%; complete canal cracks, 87.5%; incomplete canal cracks, 75.0%) than the specicity of endoscopy 8, microscopy 16, and microscopy 24 for the same categories, with statistically signicant differences between endoscopy 64 and

both microscope magnications only for all cracks (p < 0.01). For intradentin cracks, no signicant differences were found for specicity and sensitivity when the four test congurations were compared.

Analysis 2 SEM showed a total of 145 cracks (38 complete canal cracks, 102 incomplete canal cracks, and 5 intradentin cracks) that were seen at the cut root faces of the 52 roots (Table 2). Endoscopy 64 exhibited the highest percentages of concurrence between tested visual aid and SEM with regard to the type and location of cracks for all categories (ie, all cracks, 66.9%; complete canal cracks, 78.9%; incomplete canal cracks, 64.7%; and intradentin cracks, 20%). For the category all cracks, endoscopy 64 was signicantly better than all other test congurations (p < 0.05). For the categories complete canal cracks and incomplete canal cracks, endoscopy 64 was signicantly better than microscopy 16 and 24 (p < 0.05), but no signicant

Figure 4. An image of the same cut root face obtained with a microscope (24 magnication); all three cracks are visible, but the incomplete canal crack in the upper right is still difcult to identify. (This gure is available in color online at www.aae.org/joe/.)

Figure 6. An image of the same cut root face obtained with an endoscope (64 magnication): a view of the incomplete canal crack in the upper right quadrant. (This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Detection of Dentinal Cracks After Root-end Resection

1565

Basic ResearchTechnology
Correctly absent
33.3cd 21 87.5 6 75.0 45 62.5 46 75.0 45 75.0 46 5 47 0 0
a,b,c,d,e,g,h

Endoscope 64

Correctly present

40 90.9gh

47 95.9ab

26 92.9ef

Correctly absent

66.7 23

95.8 5

1 20.0 1 20.0 0 0 97.8

f

Endoscope 8

Correctly present

44 89.8

24 85.7

Figure 7. An image of the same cut root face obtained with an endoscope (64 magnication): a view of the two complete canal cracks in the lower quadrants. (This gure is available in color online at www.aae.org/joe/.)

Correctly absent

91.7 6

36 81.8

Correctly absent

Discussion
The present study evaluated the identication of dentinal cracks after root-end resection in extracted human molars. Microscopy (16 and 24 magnication) and endoscopy (8 and 64 magnication) were the visual aids tested, and SEM served as the reference. A total of 145 cracks were identied with SEM in 52 roots. The majority (70.3%) of detected cracks were incomplete canal cracks. However, because crack identication was only performed at the resection plane (ie, 3 mm from the apex), it can be speculated that incomplete canal cracks might be complete canal cracks and vice versa in a different plane (16). It has been shown that instrumentation for orthograde root canal preparation can lead to the development of cracks, in particular in the apical area (17). The clinical signicance of dentinal cracks identied during apical surgery has not yet been claried. A complete canal crack might contribute to leakage resulting in the recurrence of periapical infection and might explain surgical (and nonsurgical) failures in an unknown number of cases. Concerns that root-end cavity preparation using sonic- or ultrasonic-powered microtips might augment crack formation have been refuted in three cadaver studies (18-20) and in one clinical study (21). Reasons for crack development in the present material included the extraction of the tooth and ex vivo treatment (crown resection, endodontic preparation and obturation, and root-end resection) (16, 20, 22, 23). None of the teeth in this study was endodontically treated before extraction. 1566

Microscope 16

Correctly present

100c 22

Crack absent

24

TABLE 1. Identication of Cracks per Root (n = 52)

SEM (reference)

Crack present

49

28

44

30 68.2g

37 75.5a

19 67.9e

91.7 6

difference was found between endoscopy 64 and endoscopy 8. For intradentin cracks, no signicant differences were found among the four test congurations. False-positive cracks were more frequently seen with endoscopy 64 than with the other tested visual aids, with signicant differences for the category all cracks when comparing endoscopy 64 with microscopy 16 and 24 (p < 0.01), for the category incomplete canal cracks when comparing endoscopy 64 with microscopy 16 and 24 (p < 0.05), and for the category intradentin cracks when comparing endoscopy 64 with microscopy 16 (p < 0.05) (Table 3).

Microscope 24

Correctly present

37 75.5b

100d 22

29 65.9h

20 71.4f

95.7
Signicant differences (p < 0.01). Signicant differences (p < 0.05).

von Arx et al.

All cracks Sensitivity (%) Specicity (%) Complete canal cracks Sensitivity (%) Specicity (%) Incomplete canal cracks Sensitivity (%) Specicity (%) Intradentin cracks Sensitivity (%) Specicity (%)

JOE Volume 36, Number 9, September 2010

97.8

95.7

Basic ResearchTechnology
The main nding of the present study was that endoscopy at 64 magnication yielded the highest percentage of correct identication (66.9%) of root-end dentinal cracks compared with the other tested visual aids (44.1%-55.2%). Similar results were recently presented in a study that assessed the identication of articially created cracks in extracted human permanent incisors after root-end resection (7). That study reported the following enhanced accuracies of crack identication with increasing magnication: 39% accuracy for unaided vision, 45% accuracy for loupes (3.3 magnication), 53% accuracy for the microscope (10 magnication), and 58% accuracy for the endoscope (35 magnication). The difference in accuracy among the four visualization techniques was signicant (p = 0.0007). However, in contrast to the present study, Slaton et al (7) assessed if a specimen (root-end surface) was correctly identied as having or not having a crack, without specic determination of the type and location of the crack. A detailed analysis of the data of the present study with regard to the identication of a specic type of crack showed that sensitivity was always better for complete canal cracks (52.6%-78.9%) than for incomplete canal cracks (42.2%-64.7%) and intradentin cracks (0%-20%). None of the intradentin cracks was detected with microscopy, and only one intradentin crack (20%) was identied with endoscopy. Similarly, lower sensitivity of crack detection of intradentin cracks (36%) compared with incomplete canal cracks (73%) and complete canal cracks (97%) was reported in a previous experimental study comparing endoscopy and SEM (9). In the present study, the misinterpretation of incomplete canal cracks as complete canal cracks was found to range from 5.9% to 15.7% and of complete canal cracks as incomplete canal cracks from 7.9% to 15.8%. In contrast, a complete canal crack or an incomplete canal crack was never misinterpreted as an intradentin crack or vice versa. Interestingly, crack misinterpretation differed when comparing microscopy with endoscopy. Complete canal cracks were more frequently misinterpreted as incomplete canal cracks with microscopy than with endoscopy, whereas the misinterpretation of incomplete canal cracks as complete canal cracks was higher with endoscopy than with microscopy. In addition, endoscopy 8 and 64 showed more false-positive crack identication (6.9% and 12.4%) compared with microscopy 16 and 24 (each 4.8%). In fact, endoscopy 64 yielded relatively high false-positive results for incomplete canal cracks (13.7%) and intradentin cracks (40%). When the endoscope is used at this high magnication, the optical lens and light source are very close to the root-end surface (0.5-1 mm), and grooves or surface irregularities caused by the resection bur may produce a change in light reection, creating the illusion of a crack (8). Slaton et al (7) also described an interesting nding when load was applied for articially creating cracks. They reported that the dentin around the area of strain became opaque or frosted in appearance before a crack developed. The authors speculated that these changes were caused by the formation of many microscopic cracks that had not yet formed a macrocrack. This nding should heighten the clinicians suspicion that a crack may be present. Studies about crack propagation have shown microcracking at the crack tip and crack opening (macrocracking) in the so-called crack wake (24, 25). Staining (eg, with methylene blue) may not necessarily enhance the detection of cracks because the dye cannot ow into craze lines unless there is a break in the surface. Once a crack has propagated into a ssure or fracture, these can be stained with dyes. A recent study evaluated and compared crack identication using transillumination and various dyes (8). The technique that provided the best discrimination between cracked and noncracked resected roots was methylene blue plus transillumination. In a clinical situation, however, 1567
29 20.0d,e,f 5 13.2i 20 19.6m,n,o 4 15.7l 0 64.7j,k 1 36.3o 4 11.8 0 52.0 1 48.0n 5 9.8 0 42.2k 0 52.0m 5 5.9l 0 42.2j 0
Crack not identied

Endoscope 64

Location correct, but wrong type

19 13.1 3

7.9 16

97 66.9a,b,c 30

Type and location correct

78.9g,h 66

Crack not identied

49 33.8f 8

Endoscope 8

Location correct but wrong type

16 11.0 4

10.5 12

21.1 37

Type and location correct

80 55.2c 26

68.4 53

Crack not identied

66 45.5e 12

31.6i 49

Microscope 24

Location correct but wrong type

16 11.0 6

15.8 10

Type and location correct

63 43.5b 20

52.6h 43

Crack not identied

69 47.6d 11

28.9 53

Microscope 16

TABLE 2. Identication of Cracks by Type and Location (N = 145)

Location correct but wrong type

12 8.3 6

15.8 6

Type and location correct

55.3g 43

64 44.1a 21

SEM (reference)

0
a,b,d,e,f,h,j,k,m,n,o c,g,I,l

0
Signicant differences (p < 0.01). Signicant differences (p < 0.05).

Crack present

145

38

102

JOE Volume 36, Number 9, September 2010

All cracks Percentage Complete canal cracks Percentage Incomplete canal cracks Percentage Intradentin cracks Percentage

100

100

20.0

80.0

20.0

80.0

Detection of Dentinal Cracks After Root-end Resection

Basic ResearchTechnology
TABLE 3. False-Positive Cracks Identied with Visual Aid but not Present in SEM Microscope 16
No. of additional cracks % false-positives (= n/145) No. of additional complete canal cracks % false-positives (= n/38) No. of additional incomplete canal cracks % false positive (= n/102) No. of additional intradentin cracks % false positive (= n/5)
a,b c,d,e

Microscope 24
7 4.8b 1 2.6 5 4.9d 1 20.0

Endoscope 8
10 6.9 1 2.6 8 7.8 1 20.0

Endoscope 64
18 12.4ab 2 5.3 14 13.7cd 2 40.0e

7 4.8a 2 5.3 5 4.9c 0 0e

Signicant differences (p < 0.01). Signicant differences (p < 0.05).

transillumination may be difcult to apply if the root end is level with the bone after resection. Wright et al (8) suggested removing a small amount of buccal bone overlying the root surface or transilluminating through bone. Whether radicular dentinal cracks can enlarge during functioning and in this way result in future leakage pathways or root fractures remains to be investigated (20). Whether there is a need for extending root-end cavity preparation to include and to obturate canal cracks identied at the resection level is a topic for future clinical research.

Conclusions
Endoscopy at 64 magnication yielded the highest percentages of correct identication of root-end dentinal cracks (categories: all cracks, complete canal cracks, and incomplete canal cracks). With respect to the correct detection of a specic type of crack, complete canal cracks showed the highest sensitivity of identication with endoscopy at 64 magnication. The correct identication of intradentin cracks was low with endoscopy and nonexistent with microscopy. A relatively high number of false-positive incomplete canal cracks and intradentin cracks was registered with endoscopy at 64 magnication compared with the other visual aids.

Acknowledgment
The authors thank Brigitte Megert, Department of Preventive, Restorative and Pediatric Dentistry, University of Bern, for her help with the scanning electron microscopy evaluation.

References
1. European Society of Endodontology. Quality guidelines for endodontic treatment: consensus report of the European Society of Endodontology. Int Endod J 2006; 39:92130. 2. Kim S. Principles of endodontic microsurgery. Dent Clin N Am 1997;41:48197. 3. Rubinstein RA, Kim S. Short-term observation of the results of endodontic surgery with the use of a surgical operation microscope and Super-EBA as root-end lling material. J Endod 1999;25:438. 4. Kim S, Kratchman S. Modern endodontic surgery concepts and practice: a review. J Endod 2006;32:60123. 5. Rubinstein RA, Kim S. Long-term follow-up of cases considered healed one year after microsurgery. J Endod 2002;28:37883.

6. Taschieri S, del Fabbro M, Testori T, et al. Endodontic surgery using 2 different magnication devices: preliminary results of a randomized controlled study. J Oral Maxillofac Surg 2006;64:23542. 7. Slaton CC, Loushine RJ, Weller RN, et al. Identication of resected root-end dentinal cracks: a comparative study of visual magnication. J Endod 2003;29:51922. 8. Wright HM, Loushine RJ, Weller RN, et al. Identication of resected root-end dentinal cracks: a comparative study of transillumination and dyes. J Endod 2004;30:7125. 9. von Arx T, Montagne D, Zwinggi C, et al. Diagnostic accuracy of endoscopy in periradicular surgerya comparison with scanning electron microscopy. Int Endod J 2003;36:6919. 10. Zero DT, Rahbek I, Fu J, et al. Comparison of the iodide permeability test, the surface microhardness test, and mineral dissolution of bovine enamel following acid challenge. Caries Res 1990;24:1818. 11. Roane J, Sabala C, Ducanson M. The balanced force concept for instrumentation of curved canals. J Endod 1985;11:20311. 12. Cambruzzi JV, Marshall FJ, Pappin JB. Methylene blue dye. An aid to endodontic surgery. Endo Report 1985;11:3114. 13. Layton CA, Marshall JG, Morgan LA, et al. Evaluation of cracks associated with ultrasonic root-end preparation. J Endod 1996;22:15760. 14. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York, NY: John Wiley & Sons; 1981. 15. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. New York, NY: Psychology Press; 1988. 16. Altshul JH, Marshall G, Morgan LA, et al. Comparison of dentinal crack incidence and of post removal time resulting from post removal by ultrasonic or mechanical force. J Endod 1997;23:6836. 17. Adorno CG, Yoshioka T, Suda H. The effect of root preparation technique and instrumentation length on the development of apical root cracks. J Endod 2009;35:38992. 18. Calzonetti KJ, Iwanowski T, Komorowski R, et al. Ultrasonic root-end cavity preparation assessed by an in situ impression technique. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:2105. 19. Gray GJ, Hatton JF, Holtzmann DJ, et al. Quality of root-end preparation using ultrasonic and rotary instrumentation in cadavers. J Endod 2000;26:2813. 20. De Bruyne MAA, de Moor RJG. SEM analysis of the integrity of resected root apices of cadaver and extracted teeth after ultrasonic root-end preparation at different intensities. Int Endod J 2005;38:3109. 21. Morgan LA, Marshall JG. A scanning electron microscopic study of in vivo ultrasonic root-end preparations. J Endod 1999;25:56770. 22. Dang DA, Walton RE. Vertical root fracture and root distortion: effect of spreader design. J Endod 1989;15:294301. 23. Onnink PA, Davis RD, Wayman BE. An in-vitro comparison of incomplete rootfractures associated with three obturation techniques. J Endod 1994;20:327. 24. Nalla RK, Kinney JH, Ritchie RO. On the fracture of human dentin: is it stress- or strain-controlled? J Biomed Mater Res 2003;67A:48495. 25. Kruzic JJ, Nalla RK, Kinney JH, et al. Mechanistic aspects of in-vitro fatigue-crack growth in dentin. Biomaterials 2005;26:1195204.

1568

von Arx et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

The Quality of Root Canal Preparation and Root Canal Obturation in Canals Treated with Rotary versus Self-adjusting Files: A Three-dimensional Micro-computed Tomographic Study
Zvi Metzger, DMD,* Raviv Zary, DMD, Raphaela Cohen, DMD, Ehud Teperovich, DMD, and Frank Paque, DMD
Abstract
Aim: The study was designed to quantitatively evaluate the quality of root canal preparation and root canal obturation in canals treated with either rotary or self adjusting les, using three-dimensional microcomputed tomographic (CT) analysis. Methodology: Pair-matched root canals were instrumented with either rotary nickel-titanium les or self-adjusting les following the manufacturers instructions. The area of the canal wall unaffected by the preparation procedure was analyzed using before and after micro-CT images. Root canal obturation was done using lateral compaction with gutta-percha and AH26 (Dentsply-DeTrey, Konstanz, Germany). Teeth were scanned a third time, and the adaptation of the lling material to the canal walls was evaluated three-dimensionally by micro-CT analysis and the area of canal wall untouched by the lling was determined. The correlation between these two parameters within each of the groups was studied using the Pearson correlation test. Results: A high percentage of unaffected root canal walls (60% 14%) and areas untouched by the root canal lling (45% 15%) were found in canals treated with rotary les. Both parameters were signicantly smaller in canals treated with self-adjusting les (17% 9% and 17% 11%, respectively) (p < 0.01). No correlation was found between these parameters within each of the groups. Conclusion: Within the limitations of the present study, the self-adjusting les allowed better cleaning and shaping and better adaptation of the root canal lling than those allowed by rotary les. (J Endod 2010;36:15691573)

Key Words
MicroCT, obturation, root canal lling, SAF, self adjusting le

oot canal obturation is an essential stage of root canal treatment aimed to seal the root canal in order to prevent future bacterial contamination/recontamination of the canal space (1). Many obturation methods have been introduced over the years, each attempting to provide a better seal of the root canal (2). All have in common the assumption that the root canal is properly cleaned and shaped before the obturation stage. It is assumed by all that if the root canal is not adequately prepared and if tissue remnants and debris are present along the walls, proper sealing may be jeopardized, even with the best root canal lling method (3, 4). When simple, narrow, straight root canals with round cross-sections are considered, most current rotary nickel-titanium le systems will adequately clean and shape the canal with favorable results. The case is different in oval, at, or curved root canals. In at root canals, rotary le systems often fail to adequately clean and shape the canal, leaving ns that may have not been prepared (24). In such a case, even warm gutta-percha obturation methods will fail to adequately seal the root canal (4). Clinical buccolingual radiographs will fail to detect such discrepancy. Microcomputed tomographic (CT) studies by Peters et al (6) have further extended the understanding of the limitations of rotary le systems. They clearly showed that inadequate preparation also often occurs in curved root canals, even if they do not have a at cross-section. In upper molars treated with the ProTaper system (DentsplyMaillefer, Ballaigues, Switzerland), 49% (29%) of the canal wall was left untouched, even in the larger palatal canals (6). Again, two-dimensional, clinical periapical radiographs cannot disclose the discrepancies. This led to the recent introduction of a new self-adjusting le (SAF), which not only adapts itself longitudinally to a curved canal, as most rotary nickel-titanium les do, but also adapts itself to the cross-section of the canal (5). Rather than machining each canal into one with a circular cross-section, it removes an even dentin layer from all around the root canal, thus respecting the shape of a given root canal rather than imposing a circular cross-section on every canal no matter what its shape (5). Recent micro-CT analysis of root canals prepared with this new le indicated that this new technology allows for a higher percent of the root canal surface to be affected by the procedure (5). The resulting apical size is usually at least equivalent to a #40 le.

From the *Department of Endodontology, The Goldschleger School of Dental Medicine, Tel Aviv University, Tel Aviv; Redent-Nova Inc, Raanana, Israel; and Department of Preventive Dentistry, Peridontology and Cariology, University of Zurich, Zurich, Switzerland. Dr Ehud Teperovich, Dr Raphaela Cohen, and Dr Raviv Zary are employed by ReDent-Nova, manufacturer of the SAF le. Dr Zvi Metzger serves as a scientic consultant to the same company. Address requests for reprints to Dr Zvi Metzger, School of Dental Medicine, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel. E-mail address: metzger@post.tau. ac.il. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.003

JOE Volume 36, Number 9, September 2010

Root Canal Preparation and Obturation in Canals Treated with Rotary vs Self-adjusting Files

1569

Basic ResearchTechnology
Following the concept of the previously mentioned micro-CT analysis methods, a new method for quantitative three-dimensional analysis of the adaptation of root canal lling material to the canal walls was designed. This new method was applied in the present study to quantitatively evaluate and compare the results of instrumentation with an SAF to those obtained with nickel-titanium rotary les. irrigation with 17% EDTA. A nal ush with 5 mL 3% NaOCl was used to remove the EDTA, and the canal was dried using paper points.

Materials and Methods

Experimental Design The study was designed to compare two instrumentation methods using two three-dimensional parameters: (1) the quality of root canal cleaning and shaping, as expressed by the root canal wall area affected/unaffected by the procedure, and (2) the quality of obturation, as expressed by the percent of the root canal wall area (after preparation) touched/untouched by the root canal lling material. Teeth Teeth were selected from a large random collection of extracted human teeth that were recently extracted for reasons unrelated to the present study and kept in 10% buffered formalin. Both threedimensional images and two-dimensional cross-sections obtained using micro-CT scanning were available for each tooth in this collection. Ten pairs of roots were selected based on matching root canal morphology. These 20 roots included two mesial roots of lower molars, two distal roots of lower molars, six premolars, eight incisors, and two canines. Pair selection was based on visual similarity in shape, size, atness, and curvature of the root canals, as seen in a set of threedimensional micro-CT images of each of the roots. The two roots of each pair were randomly assigned to one of the two treatment groups (rotary les or SAFs). Root Canal Cleaning and Shaping Two le systems were used: rotary nickel-titanium les (ProTaper, Dentsply-Maillefer, Ballaigues, Switzerland) and SAFs (ReDent-Nova Ltd, Raanana, Israel). Each instrument was applied following its manufacturers instructions. Rotary Files The rotary les (ProTaper) were operated with a handpiece attached to a speed- and torque-controlled motor (X-Smart, Dentsply-Maillefer) at 250 rpm. The sequence used was ProTaper S1, S2, F1, F2, and F3 with RC-Prep (Premier, Plymouth Meeting, PA) used as a chelator/lubricant with each le. The canal was irrigated with 5 mL 3% NaOCl between the instruments. A nal ush with 5 mL 17% EDTA was applied, followed by an additional ush with 5 mL 3% NaOCl to remove the EDTA, and the canal was dried using paper points. SAF
The SAF (ReDent) was operated for 4 minutes using a GentlePower Lux 20LP KaVo handpiece, (KaVo, Biberach, Germany) adapted with a RDT3 head (ReDent-Nova, Raanana, Israel) (5). The micromotor rotation speed was set at 5,000 rpm, which resulted in an in-and-out operation of 5,000 vibrations per minute with an amplitude of 0.4 mm. The le was used with a manual in-and-out motion to the working length. Continuous irrigation was applied throughout the procedure (5) at 5 mL/min using a special irrigation apparatus (VATEA irrigation device, part of the SAF-System, ReDent); 3% NaOCl was used for irrigation during the rst 3 minutes of the operation, followed by 1 minute of 1570

Micro-CT Evaluation of the Root Canal Preparation Before preparation and scanning, each experimental tooth was mounted on scanning electron microscopy carriers (014001-T; BalTec AG, Balzers, Liechtenstein) to allow exact repositioning in the scanning system. Specimens were scanned before and after preparation by using a commercially available micro-CT system (mCT 40; Scanco Medical, Bruttisellen, Switzerland). Teeth were scanned at 70 kV and 114 mA with an isotropic resolution of 18 mm. Although the mounting device ensured almost exact repositioning of the specimens, superimposition was further calculated with newly developed software (IPL Register 1.01, Scanco Medical), as previously reported by Paque et al. (7). The nal exact superimposition of the teeth before and after preparation was with a precision better than one voxel. Individual root canal models were reconstructed up to the level of the cementoenamel junction using specially developed software (IPL V5.06B, Scanco Medical). Superimposition of the root canals before and after preparation enabled visualization and quantitatively three-dimensional evaluation of the amount of un-instrumented areas (Fig. 1). This parameter was expressed as a percentage of the number of static surface voxels of the total number of surface voxels. Root Canal Obturation Root canal obturation was performed using the lateral compaction technique with gutta-percha and AH26 sealer (Dentsply-DeTrey, Konstanz, Germany). A gutta-percha master cone (DiaDent, Almere, The Netherlands) was tted with tug-back in each root canal. Sealer was placed into the canal using a lentulo spiral followed by insertion of the master cone to the predetermined working length. Nickeltitanium nger spreaders (Dentsply-Maillefer) were used to conduct the lateral compaction using XXF/ XF accessory cones (Sure-Endo, Paris, France). When no additional cones could be inserted, the gutta-percha mass was cut off 1 mm apical to the canal orice using a heated plugger (Dentsply-Maillefer). The freshly cut surface was vertically condensed using a cold plugger (Dentsply-Maillefer). The sealer was then allowed to set for 4 days at 37 C and 100% humidity. Evaluation of Obturation Quality by Three-Dimensional Micro-CT Analysis Each tooth was subjected to a third micro-CT scan as detailed earlier. Differences in the radiopacity between the root canal lling and the root dentin allowed differentiation between the two. For each root, the lling material was three-dimensionally reconstructed and superimposed with the image of the root canal after cleaning and shaping. Superimposition of root llings and the prepared root canals (third and second scanning) allowed a three-dimensional analysis of the areas of the root canal surface, which were touched/untouched by the root canal lling. Statistical Methods The area unaffected by instrumentation and the area untouched by the root canal lling of the two instrumentation groups were compared with each other using the Student t test. The one-tailed Pearson correlation test was used to study the correlation within each group between the unaffected surface in a given root canal and the area untouched by the root canal lling in the same canal.
JOE Volume 36, Number 9, September 2010

Metzger et al.

Basic ResearchTechnology

Figure 1. Three-dimensionally reconstructed micro-CT images of root canal preparation and obturation. (A) A at root canal prepared with the SAF le. (B) A at root canal prepared with a rotary le. (C) A good root canal lling adaptation with 98.1% of the canal wall in contact with the root canal lling material. (D) A poor root canal lling adaptation with only 68.9% of the root canal wall in contact with the root canal lling material. Note the uninstumented lingual n in B that was most probably full of debris, which prevented the sealer from owing into it in D. Red: root canal surface before treatment. Blue: root canal surface post-treatment. Yellow: area touched by the root canal lling. Right, in each panel: buccal view, left: distal view. (resolution = 18 mm). (This gure is available in color online at www.aae.org/joe/.)

Results
Percent Root Canal Surface Unaffected by Root Canal Preparation The root canal surface unaffected by root canal preparation was calculated as a percent of the root canal surface area before preparation (Table 1 and Fig. 1A and B). A wide range of this three-dimensional parameter was recorded, between 5.3% and 76.6%. The mean unaffected area was 16.7% (8.9%) and 60.2% (13.6%) in the self SAF and rotary le groups, respectively (Fig. 2).
JOE Volume 36, Number 9, September 2010

Percent Root Canal Surface After Preparation Untouched by the Root Canal Filling Material The root canal surface untouched by the root canal lling material was calculated as a percent of the root canal surface area after root canal preparation (Table 1 and Fig. 1C and D). A wide range of this parameter was recorded, between 1.9% and 75.1%. The mean area untouched by the root canal lling was 17.0% (11.0%) and 44.6% (14.5%) in the SAF and rotary le groups, respectively (Fig. 2).
1571

Root Canal Preparation and Obturation in Canals Treated with Rotary vs Self-adjusting Files

Basic ResearchTechnology
TABLE 1. A Three-Dimensional Micro-CT Analysis of the Quality of Cleaning and Shaping and Root Canal Filling Adaptation to the Canal Walls Pair #
1 2 3 4 5 6 7 8 9 10

Type of canal
R-S R-S R-S R-C R-C F-S F-S F-S F-S F-C

Method
RF SAF RF SAF RF SAF RF SAF RF SAF RF SAF RF SAF RF SAF RF SAF RF SAF

Area unaffected by root canal preparation (%)

66.7 14.9 64.4 7.1 49.4 20.8 28.0 33.8 76.6 5.3 63.0 21.6 73.0 5.8 60.9 20.2 59.3 21.1 61.1 15.9

Area untouched by root canal lling (%)

28.9 1.9 64.3 15.8 42.4 8.5 38.8 23.5 37.6 5.8 44.1 30.5 75.1 15.1 47.0 8.0 37.1 29.4 31.1 31.2

R-S, round cross-section, straight; R-C, round cross-section, curved; F-S, at cross-section, straight; F-C, at cross-section, curved; RF, Rotary les.

Correlation Between the Area Unchanged by the Root Canal Preparation and the Area Untouched by the Root Canal Filling No correlation was found within each of the groups between the percent of the area unaffected by the root canal preparation and the area untouched by the root canal lling.

Discussion
Many obturation methods are used today, ranging from traditional lateral compaction to a variety of heat-softened gutta-percha techniques. All are aimed at providing a good adaptation of the root canal lling material to the canal walls, thus ensuring an adequate seal that
100

Rotary File
80

Self Adjusting File

Percent

60

40

20

Area Unaffected by C&S

Area Untouched by RCF

Figure 2. Comparison between the quality of the root canal preparation and adaptation of the root canal lling in root canals treated with rotary les or SAFs. The quality of root canal preparation is expressed as a percent of the root canal surface that was unaffected by the le. Adaptation of the root canal lling is expressed as a percent of the root canal surface after preparation that was untouched by the root canal lling. RCF, root canal lling; C&S, cleaning and shaping.

will prevent bacterial contamination/re-contamination of the root canal system. When applied in adequately prepared canals with no tissue remnants and with a clean, prepared dentin surface, this goal may be rather easily achieved. The case may be different in root canals that were inadequately cleaned and shaped. Tissue and debris remaining in parts of the canal that were unaffected by the procedure may present a barrier that does not allow for the root canal lling to intimately touch the root canal wall, thus forming the weakest link in the chain of steps aimed to properly seal the canal. This may happen in curved canals in which the les failed to touch some of the walls (6) but constitutes an even greater problem in the case of at or ribbon-shaped canals. In these canals, a rotary nickel-titanium le alone may be unable to adequately prepare the canal (Fig. 1B and D). Its action may result in a canal prepared to accommodate a certain thickness of master cone or root canal lling but may allow for buccal and/or lingual ns full of tissue remnants and debris to remain untouched (3, 4). These buccal and/or lingual defects may go unobserved in regular periapical radiographs; the root canal lling that is present in the central part of the canal will most likely mask them. Bacterial retention in or penetration into and through these defects may result in endodontic failure in an apparently radiographically acceptable case. Even though the relation between the quality of cleaning and shaping and the potential of the root canal lling to intimately touch the walls of the prepared root canal is readily understood, it has never, to the best of our knowledge, been established quantitatively for the whole canal. A study aimed to investigate the correlation between these parameters calls for a high variety of root canal cleaning and shaping scores that could later be analyzed against root canal lling adaptation to the walls of the same canals. The roots selected for the present study intentionally included a random variety of root canal morphologic shapes. This was done so that a wide spectrum of cleaning and shaping results would be available for analysis. These roots ranged from simple straight root canals with a round cross-section, which were likely to score high in effective cleaning and shaping using any le system, to curved roots or those with at root canals that were likely to result in a higher

1572

Metzger et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
percent of root canal wall unaffected by rotary nickel-titanium les (5, 6). The present study clearly showed that the two instrumentations groups differed from each other. This difference was found in both the quality of cleaning and shaping, as expressed by the percent of the root canal surface affected/unaffected by the procedure, and in the quality of root canal obturation, as expressed by the percent of the canal wall that is/is not in intimate contact with the root canal lling material. Treatment with the SAF allowed for better results with both parameters (Fig. 2). No correlation could be found between the two parameters within each group. This could result from the relatively small number of specimens in each of the groups. Further studies with larger groups may be needed to establish such correlation. Lateral compaction was used in the present study because it is the most commonly used obturation method. It may be of interest to test the same concept with heat-softened gutta-percha obturation methods, which are commonly expected to provide a better adaptation to the canal walls; nevertheless, this was beyond the scope of the present study. Micro-CT scanning has been used previously to evaluate the quality of root canal llings. Jung et al (7) have shown that the root canal lling may be differentiated from the canal wall in a micro-CT scan using digital root slices. Former studies of root canal obturation quality commonly used two-dimensional analysis of either root slices (710) or digital cross-sections generated from micro-CT scans (7). These could at best serve as a semiquantitative representation of what happens in the canal at large. A three-dimensional analysis of micro-CT images, similar to the one used in the present study, was rst applied by Zakizadeb et al (11) to evaluate intraorice barriers. It was also recently applied by Hammad et al (12) for the analysis of the volume of voids and gaps present in root canal llings. The present study was, to the best of our knowledge, the rst to use a three-dimensional micro-CT analysis to quantitatively measure the adaptation of the root canal lling material to the walls in the whole canal. As such, it provides far more comprehensive information about the adaptation of the whole root canal lling, which is unaffected by the choice of the plane in which a given section or digital cross-section happens to be.

Conclusions
A micro-CT-based quantitative three-dimensional method for analysis of root canal lling adaptation to the canal walls was presented. It may serve as a useful tool to study and compare the quality of root canal llings. Within the limitations of the present study, the SAFs allowed better cleaning and shaping and better adaptation of the root canal lling than those allowed by rotary les.

Acknowledgment
The contribution of Mr Ofer Shalev to this study is recognized and highly appreciated.

References
1. Ricucci D, Lin LM, Spangberg LSW. Wound healing of apical tissues after root canal therapy: a long-term clinical, radiographic, and histopathologic observation study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108:60921. 2. Whitworth J. Methods of lling root canals: principles and practice. Endod Topics 2005;12:224. 3. Wu M-K, Wesselink PR. A primary observation on the preparation and obturation in oval canals. Int Endod J 2001;34:13741. 4. De-Deus G, Gurgel-Filho ED, Magalhaes KM, et al. A laboratory analysis of guttapercha-lled area obtained using Thermal, System B and lateral condensation. Int Endod J 2006;39:37883. 5. Metzger Z, Teperovich E, Zary R, et al. Respecting the root canal: a new concept of a Self Adjusting File (SAF). J Endod 2010;36:67990. 6. Peters OA, Peters CI, Schonenberger K, et al. ProTaper rotary root canal preparation: effects of root canal anatomy on nal shape analyzed by micro CT. Int Endod J 2003;36:8692. 7. Jung M, Lommel D, Klimek J. The imaging of root canal obturation using micro-CT. Int Endod J 2005;38:61726. 8. Paque F, Laib A, Gautschi H, et al. Hard-tissue debris accumulation analysis by highresolution computed tomography scans. J Endod 2009;35:10447. 9. Ardila CN, Wu M-K, Wesselink PR. Percentage of lled canal area in mandibular molars after conventional root canal instrumentation and after noninstrumentation technique (NIT). Int Endod J 2003;36:5918. 10. Wu M-K, van der Sluis LWM, Wesselink PR. A preliminary study of the percentage of the gutta-percha-lled area in the apical canal lled with vertically compacted warm gutta-percha. Int Endod J 2002;35:52735. 11. Zakizadeh P, Marshal SJ, Hoover CI, et al. A novel approach in assessment coronal leakage of intraorice barriers: a saliva leakage and micro-computed tomography evaluation. J Endod 2008;34:8715. 12. Hammad M, Qualtrough A, Silikas N. Evaluation of root canal obturation: a threedimensional in vitro study. J Endod 2009;35:5414.

JOE Volume 36, Number 9, September 2010

Root Canal Preparation and Obturation in Canals Treated with Rotary vs Self-adjusting Files

1573

Basic ResearchTechnology

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased and Zinc Oxide and Eugenol Sealers
Osvaldo Zmener, DDS,* Cornelis H. Pameijer, DMS, MScD, DSc, PhD, Gabriel A. Kokubu, DDS,* and Daniel R. Grana, DVM*
Abstract
Introduction: An evaluation was made of the connective tissue reaction in rats after subcutaneous implantation of methacrylate resin-based sealers (EndoREZ [Ultradent Products, Inc, South Jordan, UT] with a polymerization accelerator and RealSeal [Sybron Dental Specialties, Orange, CA]) and Pulp Canal Sealer (Sybron Dental Specialties), a zinc oxide and eugenol-based sealer used as the control. Methods: Silicone tubes containing the test materials were implanted in 24 Wistar rats. Solid silicone rods of the same size served as the negative controls. After 10, 30, and 90 days, the animals (n = 8 per period) were euthanized and the implants with surrounding tissues dissected and processed for routine histological evaluation. A four-category evaluation system was used to measure and record the microscopic observations according to the thickness of a brous capsule, the vascular changes, and the various types of inammatory cells. Results: Initially, a severe inammatory reaction was observed of the soft tissues in direct contact with both EndoREZ/Accelerator and Real Seal. The severity decreased over time and was resolved at the end of the experiment. Pulp Canal Sealer showed a severe tissue reaction for all observation periods. The negative controls showed an initial mild to moderate inammatory reaction. After 30 days, healthy brous connective tissue was observed, which increased over time. After 10 days, no statistically signicant differences between the experimental groups were observed. After 90 days, EndoREZ and RealSeal were statistically significantly less toxic than Pulp Canal Sealer (p > 0.05). Conclusions: After 90 days, both methacrylate resinbased sealers were considered biologically acceptable when implanted in subcutaneous connective tissues of the rat. Pulp Canal Sealer remained toxic for the duration of the study. (J Endod 2010;36:15741579)

he current concept among clinicians is that after complete debridement, total obliteration of the root canal space with a biocompatible material constitutes the key factor for successful endodontic therapy (1). Different materials have been advocated for lling root canals; gutta-percha cones complemented with a sealer cement is the most widely used (2). During the last decade, methacrylate resin-based sealers (MRBSs) have gained popularity for root canal obturation (3). Preliminary reports have shown that two well-established MBRSs (ie, EndoRez [ER; Ultradent Products, Inc, South Jordan, UT] and RealSeal (RS; Sybron Dental Specialties, Orange, CA), formerly Epiphany, are both well tolerated by living tissues (47) and have shown promise for in vivo human clinical trials (810). More recently, an ER Accelerator (ACC, Ultradent Products Inc.) has been introduced. The ACC is composed of triethylene glycol dimethacrylate, tertiary amines, and a proprietary ingredient. The technique the manufacturer recommends is the following. When the master guttapercha cone has been placed to length, two or three #20 to #25/.02 taper accessory cones dipped in ACC are harpooned in the sealer and pushed into the canal space as far as possible. The combination of ER and ACC accelerates the polymerization of the sealer, thus allowing for an immediate continuation of the coronal restoration. It also prevents dislodgement of the obturating material when a post space is prepared immediately after obturation, potentially causing early bacterial leakage (11). Previous reports (12, 13) have shown that certain components from methacrylate resin-based materials may remain unpolymerized even after setting and can subsequently be released from the resin matrix. When the sealer is accidentally extruded through the apex or through a lateral canal, which is not an uncommon experience in endodontics (14), the unpolymerized components may be toxic to the periapical tissues. Although the biocompatibility of ER and RS has been investigated (47), the effect of ER/ACC in contact with living tissues and compared with RS has not been reported yet. Therefore, the purpose of this study was to evaluate the biocompatibility of ER/ACC and RS and to compare them with Pulp Canal Sealer (PCS, Sybron Dental Specialties), a zinc oxide and eugenol-based sealer, when implanted subcutaneously in connective tissue of rats.

Materials and Methods

The protocol of this study was approved by the Research Ethics Committee of the Argentine Dental Association. Autoclaved silicone tubes closed at one end (Raholin SRL, V. Madero, BA, Argentina) and 10-mm long with an internal diameter of 1 mm were lled ush with freshly prepared ER/ACC, RS, or PCS (positive control). Solid silicone rods (SIRODs) of the same size as the tubes were used as negative controls. ER alone was not tested because this had been done previously under similar conditions (5). The methacrylate resinbased sealer samples were prepared in such a way that the

Key Words
Biocompatibility, endodontics, sealers, tissue response methacrylate-based

From the *School of Dentistry, University of El Salvador, Buenos Aires, Argentina; and University of Connecticut School of Dental Medicine, Farmington, CT, USA. Supported in part by Ultradent Dental Products Inc. Address requests for reprints to Dr Cornelis H. Pameijer, 10 Highwood, Simsbury, CT 06070. E-mail address: cornelis@pameijer.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.019

1574

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 1. Severity of Tissue Reaction to the Test Materials ER/ACC Days
10 30 90

RS SE
8 0 0

PCS MO
0 8 0

SIROD MO
0 0 0

n
8 8 8

NO
0 0 7

MI
0 0 1

MO
0 8 0

NO
0 0 6

MI
0 0 2

SE
8 0 0

NO
0 0 0

MI
0 0 0

SE
8 8 8

NO
0 8 8

MI
6 0 0

MO
2 0 0

SE
0 0 0

NO, no reaction; MI, mild reaction; MO, moderate reaction; SE, severe reaction.

formation of an oxygen-inhibited layer was prevented. The sealers were prepared under aseptic conditions according to the following method. In group ER/ACC (n = 8), the experimental design necessitated a slight modication of the manufacturers recommendations. ER was injected through an automixing tip in a glass syringe measuring 25-mm long with a 5-mm internal diameter (De Luca SA, Buenos

Aires, Argentina). Two size #40 gutta-percha cones dipped in ACC were subsequently inserted in the sealer and left for 3 seconds each after which they were removed. The sealer was then immediately injected into the silicone tubes through a 30-G needle. The procedure was repeated using a new syringe for each animal. In group RS (n = 8), the sealer was extruded through an automixing tip attached to the two-barrel delivery syringe directly into a plastic

Figure 1. (A-F) Representative specimens of ER/ACC, RS, and PCS at the 10-day observation period. (A) ER/ACC: a low-power magnication of tissue/material contact (hematoxylin and eosin (H&E), original magnication 40). (A) A higher magnication of the outlined area in A. A thin band of necrotic tissue in direct contact with the sealer (black arrow) and a severe granulomatous tissue reaction with dark material particles and newly formed capillaries (white arrow) can be seen (H&E, original magnication 100). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. A thick layer of necrotic tissue in direct contact with the sealer (black arrow) is present. Below it, a severe granulomatous tissue reaction containing many newly formed capillaries (white arrow) can be seen (H&E, original magnication 100). (E) PCS: low magnication of the area of tissue/material contact (H&E, original magnication 40). (F) A higher magnication of the area of tissue/material contact in E showing extruded dark particles surrounded by a severe granulomatous tissue reaction (H&E, original magnication 150). (This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1575

Basic ResearchTechnology
syringe, which in turn delivered the mixture through a 30-G needle into the silicone tube. In group PCS (n = 8), the sealer was prepared on a glass slab according to the manufacturers recommendations. The mixture was loaded in a plastic syringe and injected into the silicone tube though a 30-G needle. Excess material was removed from the open end of the tubes with a sterile plastic instrument. After preparation, the samples were immediately implanted in the subcutaneous connective tissues of white male Wistar rats weighing approximately 200 g each. The husbandry and management of the animals met the requirements of the ISO 10993 1 (1992) and ISO 10993-2 (1992) standards (15, 16) as well as the International Regulatory Requirements for the Care and Use of Laboratory Animals (17). All surgical procedures were performed under strict aseptic conditions. After anesthesia through the intraperitoneal administration of ketamine chloride (14 mg/kg body weight) and acepromazine (10mg/kg body weight), the dorsal skin was shaved and disinfected with 10% iodine-povidone solution (Phoenix SAIC, BA, Argentina). Four approximately 18-mm-long incisions were made through the dermis with a scalpel and further prepared by blunt dissection. Subsequently, one sample of each of the four groups was carefully placed into the pocket. A distance of at least 20 mm between the samples was present to avoid interference of tissue response between two materials. Finally, the wounds were closed with silk sutures, and the animals were maintained on a regular diet and water ad libitum. Eight animals were euthanized after 10, 30, and 90 days with an anesthetic overdose resulting in eight samples per group per time period. The implants with surrounding tissues were dissected and xed in 10% neutral buffered formalin (pH 7.4). After 48 hours of xation, the samples were processed for routine histological evaluation. Parafn blocks were oriented parallel to the long axis of the tubes and longitudinal serial sections of approximately 7-mm thick were cut from the middle of the implants and stained with hematoxylin and eosin. To evaluate the tissue response, three sections of the center of each specimen

Figure 2. (A-F) A representative sample of ER/ACC, RS, and PCS of the 30-day observation period. (A) ER/ACC: a low magnication of tissue/material contact (H&E, original magnication 80). (B) A higher magnication of the outlined area in A. In direct contact with the sealer, there is a dense brogranulomatous tissue concentration (black arrow). Below it is an artifact (ART) within an area of fat cells (H&E, original magnication 150). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. A thick granulomatous zone containing many material particles can be seen in direct contact with the sealer (black arrow). Below it is a number of wide newly formed capillaries (white arrow) (H&E, original magnication 150). (E) PCS: a low power of tissue/material contact (H&E, original magnication 40). (F) A higher magnication of E showing material particles within a severe granulomatous tissue reaction. Note the presence of numerous wide newly formed capillaries (arrows) (H&E, original magnication 150). (This gure is available in color online at www.aae.org/joe/.)

1576

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
were analyzed and digitally photographed at different magnications under a light microscope. All sections were analyzed independently, and two trained evaluators who were blind to the study scored the tissue reactions using the following criteria: NO: no reaction, brous-capsule formation, and absence of inammatory cells; MI: a mild reaction and the presence of a brous-capsule formation with few inammatory cells; MO: a moderate reaction and brous-capsule formation with the presence of polymorphonuclear leukocytes, lymphocytes, plasmocytes, and macrophages; and SE: a severe reaction and the presence of large accumulations of polymorphonuclear leukocytes, lymphocytes, plasmocytes, macrophages, foreign-body giant cells, and congested capillaries. Before the analysis, both evaluators were calibrated by having them analyze a set of 70 similar but unrelated slides displaying various types of inammatory cells to endodontic sealers. In case of a disagreement between the evaluators, the sample under discussion was analyzed jointly until a consensus was reached. Data were analyzed by the Wilcoxon signed rank test to determine if there was a statistically signicant difference between materials at each observation period. The total effect of time and material upon the tissue reaction was analyzed using the Kruskal-Wallis and the Dunn test. The signicance level was set at p < 0.05. A sealer was considered to be biologically acceptable when tissue reactions were recorded as NO to MI.

Results
One animal from the 10-day time period had to be excluded from the study and was replaced with another undergoing the same implantation procedures. Macroscopic examination showed that wound healing

Figure 3. (A-F) A representative sample of ER/ACC, RS, and PCS at the 90-day observation period. (A) ER/ACC: an overview of the area of tissue/material contact showing extrusion (ES) of the sealer (H&E, original magnication 40). (B) A higher magnication of A. In contact with the extruded material, a dense brous connective tissue encapsulation free of inammatory cells is present (black arrow). The white arrow indicates a wide newly formed vessel (H&E, original magnication 850). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. Note the presence of a thick brous connective tissue capsule (arrow). A few inammatory cells can be seen within the capsule as well as in the surrounding tissues (H&E, original magnication 850). (E) PCS: an overview of the area of tissue/material contact showing numerous material particles (white arrow) surrounded by granulomatous tissue slightly invaginated within the lumen of the tube (black arrow) (H&E, original magnication 40). (F) A higher magnication of E showing a brous connective tissue (arrow) that appeared to isolate the inammatory reaction from the surrounding tissues (H&E, original magnication 850).(This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1577

Basic ResearchTechnology
was satisfactory at all observation periods. Histological evaluation showed that the implants were surrounded by brous connective tissue of irregular thickness. It could be easily distinguished from the tissue reaction at the site where the tissues were in direct contact with the test material. The severity of tissue reaction is presented in Table 1. After 10 days, the tissue reaction to ER/ACC, RS, and PCS was scored severe and was extensively dispersed around the end of the tubes (Figs. 1A-F). The majority of ER/ACC samples with direct tissue contact presented with an inammatory reaction with slight invagination into the lumen of the tubes. Some ER/ACC and RS samples exhibited a thin necrotic zone in the direct contact area. Each group displayed many newly formed vessels and randomly dispersed dark particles, which appeared to have been released from the implanted materials. A mild to moderate reaction was observed in SIROD implants. After 30 days, the intensity of the inammatory reaction to ER/ACC and RS decreased slightly and was scored as moderate (Fig. 2A-D). In direct tissue contact, both materials displayed a brogranulomatous tissue, which appeared less dispersed and showed a tendency to be surrounded by a brous tissue containing many inammatory cells. In contact with PCS, a persistent severe granulomatous tissue invagination containing many dark particles was observed. This granulomatous reaction was surrounded by an incipient layer of brous tissue, which was free of inammatory cells (Fig. 2E and F). In contact with SIRODs, there was a thin brous connective tissue containing a few inammatory cells. After 90 days, some ER/ACC samples showed extrusion of the material into the surrounding tissues. In seven samples, a thick dense and mature brous tissue capsule (approximately 70- to 90-mm thick) free of inammatory cells was present (Fig. 3A and B), whereas only one sample had a mild tissue reaction. The RS samples showed a dense brous capsule (50- to 80-mm thick) without inammatory cells, whereas two cases were scored as mild with a few inammatory cells still persisting within the capsule (Fig. 3C and D). PCS samples exhibited a severe persistent granulomatous tissue reaction containing necrotic areas as well as randomly distributed inammatory cells. Particles of the material were observed at the end and within the lumen of the tubes in all samples. Higher magnications showed the granulomatous tissues to be isolated by a thick brous tissue encapsulation. Many newly formed capillaries and a high concentration of acute and chronic inammatory cells were also seen (Fig. 3E and F). A thick dense brous connective tissue, free of inammatory cells, was observed in all SIROD samples (Fig. 4). The Wilcoxon signed rank test showed no statistically signicant differences (p > 0.05) between the reaction to ER/ACC, RS, and PCS at the 10-day observation period. After 30 days, no statistical differences were found between ER/ACC and RS (p > 0.05), whereas both ER/ACC and RS signicantly differed from PCS (p < 0.05). For both periods, ER/ ACC, RS, and PCS differed signicantly from SIRODs. After 90 days, ER/ ACC and RS differed signicantly from PCS (p < 0.05) but not from SIRODs, whereas PCS remained signicantly different from SIRODs. The effect of time for ER/ACC and RS showed that the 10 and 30 days differed signicantly (p < 0.05) from the 90-day observation period. However, no signicant differences (p > 0.05) were observed for PCS between all time intervals.

Figure 4. (A) A representative specimen of SIRODs after the 10-day observation period. In contact with the material (empty space), there is a moderate brogranulomatous tissue (black arrow) with inammatory cells and newly formed capillaries (H&E, original magnication 100). (B) A representative specimen of SIRODs after 90 days. In contact with the material (empty space), a thick dense brous connective encapsulation (arrow) without inammatory cells can be seen (H&E, original magnication 850).(This gure is available in color online at www.aae.org/joe/.)

Discussion
The implantation of endodontic lling materials into subcutaneous connective tissue of rats is a valid screening method for testing biocompatibility (18, 19). PCS was used as the positive control because its toxicity has been previously determined by in vitro (20, 21) and in vivo experiments (22, 23). SIRODs were used as the negative control because they have been proven to be biocompatible (5, 6, 24, 25), a nding that was conrmed in this study. The initial inammatory 1578

reaction of SIRODs may be a consequence of the surgical trauma. This subsided rapidly, and after 90 days a well-organized healthy dense brous tissue free of inammatory cells was observed. The biocompatibility of ER has previously been tested (46); however, the incorporation of the ACC changes not only the original chemical composition of ER but also gives the sealer a shorter setting time; an evaluation of the toxicity is therefore essential to determine its safety. The preparation of ER/ACC and RS samples was performed in such a way that no oxygen-inhibited layer was formed. Oxygen inhibits freeradical polymerization of resin-based materials yielding an uncured surface layer (26, 27), which is of particular concern for MBRS because it greatly affects the outcome in toxicity tests. Elution of uncured chemical components from the oxygen-inhibited layer is one of the main causes for tissue damage. However, severe reactions were observed for ER/ACC and RS at the short-term period, revealing that even after setting they still are an irritant. As has been shown (28), unreacted monomer persists in polymerized methacrylate resins. These unreacted monomers (28) undergo a rapid elution and the leaching of these unbound molecules (29), and other components caused the severe inammatory reactions observed after 10 days. These ndings are consistent with those of Costa et al (30, 31) and are supported by previous experiments of Ferracane and Condon (12)

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
who found that the loss of components from a methacrylate resin-based material in a water-based medium in vitro is initially rapid and, therefore, a similar behavior can be expected in vivo. It is signicant to note that the initial severe reaction decreased over time and had been resolved at the end of the experiment although a few persistent inammatory cells still remained adjacent to RS. After the rapid initial loss, the material depletes progressively slower over time (12), thus causing less irritation. Therefore, the setting time is signicant with respect to irritation. ER/ACC sets in 7 minutes, RS in 25 minutes, and PCS in 260 minutes (Zmener and Pameijer, unpublished data, 2009). The faster setting of ER/ACC could be a contributory factor in reducing the release of components, in effect locking in potential irritants (12). The tissue reaction to PCS was severe at all observation periods and is most likely caused by unreacted eugenol, which is highly toxic to tissues (32). Our ndings for PCS are in agreement with others who reported that, like other zinc oxide and eugenol-based sealers (3335), PCS reacted toxic in cell cultures (20, 21, 36) and in contact with living tissues (22, 3739). However, the results for ER/ ACC or RS disagree with those of Scarparo et al (39) who showed intense tissue reactions to ER. Their observations did not extend beyond 60 days, however. In the study reported here, a 90-day observation period was used. Longer time periods allow more time for elution of components causing a depletion of chemical components that are potential irritants. On the other hand, shortening of setting time may contribute to a lower output of uncured components from ER as shown here. Within the limitations of this study, it was concluded that ER/ACC and RS exhibited similarly and were both well tolerated by the subcutaneous connective tissue of rats after 90 days of implantation. PCS, however, remained toxic even after a 90-day observation period.
12. Ferracane JL, Condon JR. Rate of elution of leachable components from composite. Dent Mat 1990;6:2827. 13. Ferracane JL. Elution of leachable components from composites. J Oral Rehabil 1994;21:44152. 14. Augsburger RA, Peters DD. Radiographic evaluation of extruded obturation materials. J Endod 1990;16:4927. 15. ISO 10993-1:1992, Biological Evaluation of Medical DevicesPart 1: Guidance on Selection of Tests. Geneva: ISO; 1992. 16. ISO 10993-2:1992, Biological Evaluation of Medical DevicesPart 2: Animal Welfare Requirements. Geneva: ISO; 1992. 17. Bayne K. Developing guidelines of the care and use of animals. Ann N Y Acad Sci 1998;30:10510. 18. Olsson B, Sliwkowsky A, Langeland K. Subcutaneous implantation for the biological evaluation of endodontic materials. J Endod 1981;7:35569. 19. FDI. Federation Dentaire Internationale. Recommended standard practices for the biological evaluation of dental materials. Int Dent J 1980;30:1746. 20. Zmener O, Cabrini RL. Monocyte-lymphocyte adhesion to three zinc oxide and eugenol-based endodontic cements: an in vitro assay. Acta Odontol Latinoam 1986;3:2732. 21. Pinna L, Brackett MG, Lockwood PE, et al. In vitro cytotoxicity evaluation of a selfadhesive methacrylate resin-based root canal sealer. J Endod 2008;34:10858. 22. Pascon EA, Leonardo MR, Safavi K, et al. Tissue reaction to endodontic materials: methods, criteria, assessment and observations. Oral Surg Oral Med Oral Pathol 1991;72:22237. 23. Kolokouris I, Economides N, Beltes P, et al. In vivo comparison of the biocompatibility of two root canal sealers implanted into the subcutaneous connective tissue of rats. J Endod 1998;24:825. 24. Zmener O, Guglielmotti B, Cabrini RL. Biocompatibility of two calcium hydroxidebased endodontic sealers: a quantitative study in the subcutaneous connective tissue of the rat. J Endod 1988;14:22935. 25. Zmener O, Guglielmotti B, Cabrini RL. Tissue response to an experimental calcium hydroxide-based endodontic sealer: a quantitative study in the subcutaneous connective tissue of the rat. Endod Dent Traumatol 1990;6:6672. 26. Andrzejewska E, Linden LA, Rabek JF. The role of oxygen in camphorquinone-initial photopolymerization. Macromol Chem Phys 1998;199:4419. 27. Versiani MA, Carvalho-Junior JR, Padilha MIAF, et al. A comparative study of physicochemical properties of AH Plus and Epiphany root canal sealants. Int Endod J 2006;39:46471. 28. Rueggeberg FA, Margeson DH. The effect of oxygen inhibition on an unlled/lled composite system. J Dent Res 1990;69:16528. 29. Rathbun MA, Craig RG, Hanks CT, et al. Cytotoxicity of a BIS-GMA dental composite before and after leaching in organic solvents. J Biomed Mater Res 1991;25: 44357. 30. Costa CAS, Hebling J, Teixeira MF. Preliminary study of the biological compatibility of the dentine adhesives All-bond 2 and Scotchbond MP. Histological evaluation of subcutaneous implants in rats. Rev Odontol USP 1997;11:118. 31. Costa CAS, Teixeira HM, Lopes Nascimento AB, et al. Biocompatibility of two current adhesive resins. J Endod 2000;26:5126. 32. Molnar EJ. Residual eugenol from zinc-oxide-eugenol compounds. J Dent Res 1967; 46:6459. 33. Rodrigues H, Spangberg L, Langeland K. Biologic effects of dental materials. 9. Effect of zinc oxide-eugenol cements on He La cells in vitro. Estomat Cult 1975;9:1914. 34. Zmener O, Cabrini RL. Effects of three calcium hydroxide-based materials on human blood monocytes and lymphocytes. Endod Dent Traumatol 1987;3:2832. 35. Zmener O, Goldberg F, Cabrini RL. Effects of two gutta-percha formulations and one zinc oxide-eugenol and Canada balsam mixture on human blood monocytes and lymphocytes. Endod Dent Traumatol 1989;5:737. 36. Geurtsen W, Leyhausen G. Biological aspects of root canal lling materialshistocompatibility, cytotoxicity and mutagenicity. Clin Oral Invest 1997;1:511. 37. Tagger M, Tagger E. Subcutaneous reactions to implantation of tubes with AH26 and Grosmans sealers. Oral Surg Oral Med Oral Pathol 1986;62:43440. 38. Yesilsoy C, Koren LZ, Morse DR, et al. A comparative tissue toxicity evaluation of established and newer root canal sealers. Oral Surg Oral Med Oral Pathol 1988; 65:45967. 39. Scarparo RK, Grecca FS, Fachin EVF. Analysis of tissue reactions to methacrylate resin-based, epoxy resin-based, and zinc oxide-eugenol endodontic sealers. J Endod 2009;35:22932.

Acknowledgment
Dr. Pameijer is a consultant for Ultradent Products Inc.

References
1. Nguyen TN. Obturation of the root canal system. In: Cohen S, Burns RC. Pathways of the Pulp. 4th ed. St Louis: Mosby; 1987:18394. 2. Taintor JF, Ross PN. Opinions and practices of American Endodontic Diplomates. Dent J 1978;44:3215. 3. ADA Council of Scientic affairs. Statement on posterior resin-based composites. ADA Council on Dental Benet Programs. J Am Dent Assoc 1998;129:16278. 4. Louw NP, Pameijer CH, Norval G. Histopathological evaluation of a root canal sealer in subhuman primates. J Dent Res 2001;80:654. 5. Zmener O. Tissue response to a new methacrylate-based root canal sealer: preliminary observations in the subcutaneous connective tissue of rats. J Endod 2004;30:34851. 6. Zmener O, Banegas G, Pameijer CH. Bone tissue response to a methacrylate-based endodontic sealer: a histological and histometric study. J Endod 2005;31:4579. 7. Sousa CJA, Montes CRM, Pascon EA, et al. Comparison of the intraosseous biocompatibility of AH Plus, EndoREZ and Epiphany root canal sealers. J Endod 2006;32: 65662. 8. Zmener O, Pameijer CH. Clinical and radiographic evaluation of a resin-based root canal sealer. Am J Dent 2004;17:1922. 9. Zmener O, Pameijer CH. Clinical and radiographical evaluation of a resin-based root canal sealer: a 5-year follow-up. J Endod 2007;33:6769. 10. Cotton TP, Schindler WG, Schwartz SA, et al. A retrospective study comparing clinical outcomes after obturation with Resilon/Epiphany or gutta-percha/Kerr sealer. J Endod 2008;34:78997. 11. Zmener O, Pameijer CH, Alvarez Serrano S. Effect of immediate and delayed post space preparation on coronal bacterial microleakage in root canals obturated with a methacrylate-based sealer with an without accelerator. An ex vivo study. Am J Dent 2010;23:11620.

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1579

Basic ResearchTechnology

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased and Zinc Oxide and Eugenol Sealers
Osvaldo Zmener, DDS,* Cornelis H. Pameijer, DMS, MScD, DSc, PhD, Gabriel A. Kokubu, DDS,* and Daniel R. Grana, DVM*
Abstract
Introduction: An evaluation was made of the connective tissue reaction in rats after subcutaneous implantation of methacrylate resin-based sealers (EndoREZ [Ultradent Products, Inc, South Jordan, UT] with a polymerization accelerator and RealSeal [Sybron Dental Specialties, Orange, CA]) and Pulp Canal Sealer (Sybron Dental Specialties), a zinc oxide and eugenol-based sealer used as the control. Methods: Silicone tubes containing the test materials were implanted in 24 Wistar rats. Solid silicone rods of the same size served as the negative controls. After 10, 30, and 90 days, the animals (n = 8 per period) were euthanized and the implants with surrounding tissues dissected and processed for routine histological evaluation. A four-category evaluation system was used to measure and record the microscopic observations according to the thickness of a brous capsule, the vascular changes, and the various types of inammatory cells. Results: Initially, a severe inammatory reaction was observed of the soft tissues in direct contact with both EndoREZ/Accelerator and Real Seal. The severity decreased over time and was resolved at the end of the experiment. Pulp Canal Sealer showed a severe tissue reaction for all observation periods. The negative controls showed an initial mild to moderate inammatory reaction. After 30 days, healthy brous connective tissue was observed, which increased over time. After 10 days, no statistically signicant differences between the experimental groups were observed. After 90 days, EndoREZ and RealSeal were statistically significantly less toxic than Pulp Canal Sealer (p > 0.05). Conclusions: After 90 days, both methacrylate resinbased sealers were considered biologically acceptable when implanted in subcutaneous connective tissues of the rat. Pulp Canal Sealer remained toxic for the duration of the study. (J Endod 2010;36:15741579)

he current concept among clinicians is that after complete debridement, total obliteration of the root canal space with a biocompatible material constitutes the key factor for successful endodontic therapy (1). Different materials have been advocated for lling root canals; gutta-percha cones complemented with a sealer cement is the most widely used (2). During the last decade, methacrylate resin-based sealers (MRBSs) have gained popularity for root canal obturation (3). Preliminary reports have shown that two well-established MBRSs (ie, EndoRez [ER; Ultradent Products, Inc, South Jordan, UT] and RealSeal (RS; Sybron Dental Specialties, Orange, CA), formerly Epiphany, are both well tolerated by living tissues (47) and have shown promise for in vivo human clinical trials (810). More recently, an ER Accelerator (ACC, Ultradent Products Inc.) has been introduced. The ACC is composed of triethylene glycol dimethacrylate, tertiary amines, and a proprietary ingredient. The technique the manufacturer recommends is the following. When the master guttapercha cone has been placed to length, two or three #20 to #25/.02 taper accessory cones dipped in ACC are harpooned in the sealer and pushed into the canal space as far as possible. The combination of ER and ACC accelerates the polymerization of the sealer, thus allowing for an immediate continuation of the coronal restoration. It also prevents dislodgement of the obturating material when a post space is prepared immediately after obturation, potentially causing early bacterial leakage (11). Previous reports (12, 13) have shown that certain components from methacrylate resin-based materials may remain unpolymerized even after setting and can subsequently be released from the resin matrix. When the sealer is accidentally extruded through the apex or through a lateral canal, which is not an uncommon experience in endodontics (14), the unpolymerized components may be toxic to the periapical tissues. Although the biocompatibility of ER and RS has been investigated (47), the effect of ER/ACC in contact with living tissues and compared with RS has not been reported yet. Therefore, the purpose of this study was to evaluate the biocompatibility of ER/ACC and RS and to compare them with Pulp Canal Sealer (PCS, Sybron Dental Specialties), a zinc oxide and eugenol-based sealer, when implanted subcutaneously in connective tissue of rats.

Materials and Methods

The protocol of this study was approved by the Research Ethics Committee of the Argentine Dental Association. Autoclaved silicone tubes closed at one end (Raholin SRL, V. Madero, BA, Argentina) and 10-mm long with an internal diameter of 1 mm were lled ush with freshly prepared ER/ACC, RS, or PCS (positive control). Solid silicone rods (SIRODs) of the same size as the tubes were used as negative controls. ER alone was not tested because this had been done previously under similar conditions (5). The methacrylate resinbased sealer samples were prepared in such a way that the

Key Words
Biocompatibility, endodontics, sealers, tissue response methacrylate-based

From the *School of Dentistry, University of El Salvador, Buenos Aires, Argentina; and University of Connecticut School of Dental Medicine, Farmington, CT, USA. Supported in part by Ultradent Dental Products Inc. Address requests for reprints to Dr Cornelis H. Pameijer, 10 Highwood, Simsbury, CT 06070. E-mail address: cornelis@pameijer.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.019

1574

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
TABLE 1. Severity of Tissue Reaction to the Test Materials ER/ACC Days
10 30 90

RS SE
8 0 0

PCS MO
0 8 0

SIROD MO
0 0 0

n
8 8 8

NO
0 0 7

MI
0 0 1

MO
0 8 0

NO
0 0 6

MI
0 0 2

SE
8 0 0

NO
0 0 0

MI
0 0 0

SE
8 8 8

NO
0 8 8

MI
6 0 0

MO
2 0 0

SE
0 0 0

NO, no reaction; MI, mild reaction; MO, moderate reaction; SE, severe reaction.

formation of an oxygen-inhibited layer was prevented. The sealers were prepared under aseptic conditions according to the following method. In group ER/ACC (n = 8), the experimental design necessitated a slight modication of the manufacturers recommendations. ER was injected through an automixing tip in a glass syringe measuring 25-mm long with a 5-mm internal diameter (De Luca SA, Buenos

Aires, Argentina). Two size #40 gutta-percha cones dipped in ACC were subsequently inserted in the sealer and left for 3 seconds each after which they were removed. The sealer was then immediately injected into the silicone tubes through a 30-G needle. The procedure was repeated using a new syringe for each animal. In group RS (n = 8), the sealer was extruded through an automixing tip attached to the two-barrel delivery syringe directly into a plastic

Figure 1. (A-F) Representative specimens of ER/ACC, RS, and PCS at the 10-day observation period. (A) ER/ACC: a low-power magnication of tissue/material contact (hematoxylin and eosin (H&E), original magnication 40). (A) A higher magnication of the outlined area in A. A thin band of necrotic tissue in direct contact with the sealer (black arrow) and a severe granulomatous tissue reaction with dark material particles and newly formed capillaries (white arrow) can be seen (H&E, original magnication 100). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. A thick layer of necrotic tissue in direct contact with the sealer (black arrow) is present. Below it, a severe granulomatous tissue reaction containing many newly formed capillaries (white arrow) can be seen (H&E, original magnication 100). (E) PCS: low magnication of the area of tissue/material contact (H&E, original magnication 40). (F) A higher magnication of the area of tissue/material contact in E showing extruded dark particles surrounded by a severe granulomatous tissue reaction (H&E, original magnication 150). (This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1575

Basic ResearchTechnology
syringe, which in turn delivered the mixture through a 30-G needle into the silicone tube. In group PCS (n = 8), the sealer was prepared on a glass slab according to the manufacturers recommendations. The mixture was loaded in a plastic syringe and injected into the silicone tube though a 30-G needle. Excess material was removed from the open end of the tubes with a sterile plastic instrument. After preparation, the samples were immediately implanted in the subcutaneous connective tissues of white male Wistar rats weighing approximately 200 g each. The husbandry and management of the animals met the requirements of the ISO 10993 1 (1992) and ISO 10993-2 (1992) standards (15, 16) as well as the International Regulatory Requirements for the Care and Use of Laboratory Animals (17). All surgical procedures were performed under strict aseptic conditions. After anesthesia through the intraperitoneal administration of ketamine chloride (14 mg/kg body weight) and acepromazine (10mg/kg body weight), the dorsal skin was shaved and disinfected with 10% iodine-povidone solution (Phoenix SAIC, BA, Argentina). Four approximately 18-mm-long incisions were made through the dermis with a scalpel and further prepared by blunt dissection. Subsequently, one sample of each of the four groups was carefully placed into the pocket. A distance of at least 20 mm between the samples was present to avoid interference of tissue response between two materials. Finally, the wounds were closed with silk sutures, and the animals were maintained on a regular diet and water ad libitum. Eight animals were euthanized after 10, 30, and 90 days with an anesthetic overdose resulting in eight samples per group per time period. The implants with surrounding tissues were dissected and xed in 10% neutral buffered formalin (pH 7.4). After 48 hours of xation, the samples were processed for routine histological evaluation. Parafn blocks were oriented parallel to the long axis of the tubes and longitudinal serial sections of approximately 7-mm thick were cut from the middle of the implants and stained with hematoxylin and eosin. To evaluate the tissue response, three sections of the center of each specimen

Figure 2. (A-F) A representative sample of ER/ACC, RS, and PCS of the 30-day observation period. (A) ER/ACC: a low magnication of tissue/material contact (H&E, original magnication 80). (B) A higher magnication of the outlined area in A. In direct contact with the sealer, there is a dense brogranulomatous tissue concentration (black arrow). Below it is an artifact (ART) within an area of fat cells (H&E, original magnication 150). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. A thick granulomatous zone containing many material particles can be seen in direct contact with the sealer (black arrow). Below it is a number of wide newly formed capillaries (white arrow) (H&E, original magnication 150). (E) PCS: a low power of tissue/material contact (H&E, original magnication 40). (F) A higher magnication of E showing material particles within a severe granulomatous tissue reaction. Note the presence of numerous wide newly formed capillaries (arrows) (H&E, original magnication 150). (This gure is available in color online at www.aae.org/joe/.)

1576

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
were analyzed and digitally photographed at different magnications under a light microscope. All sections were analyzed independently, and two trained evaluators who were blind to the study scored the tissue reactions using the following criteria: NO: no reaction, brous-capsule formation, and absence of inammatory cells; MI: a mild reaction and the presence of a brous-capsule formation with few inammatory cells; MO: a moderate reaction and brous-capsule formation with the presence of polymorphonuclear leukocytes, lymphocytes, plasmocytes, and macrophages; and SE: a severe reaction and the presence of large accumulations of polymorphonuclear leukocytes, lymphocytes, plasmocytes, macrophages, foreign-body giant cells, and congested capillaries. Before the analysis, both evaluators were calibrated by having them analyze a set of 70 similar but unrelated slides displaying various types of inammatory cells to endodontic sealers. In case of a disagreement between the evaluators, the sample under discussion was analyzed jointly until a consensus was reached. Data were analyzed by the Wilcoxon signed rank test to determine if there was a statistically signicant difference between materials at each observation period. The total effect of time and material upon the tissue reaction was analyzed using the Kruskal-Wallis and the Dunn test. The signicance level was set at p < 0.05. A sealer was considered to be biologically acceptable when tissue reactions were recorded as NO to MI.

Results
One animal from the 10-day time period had to be excluded from the study and was replaced with another undergoing the same implantation procedures. Macroscopic examination showed that wound healing

Figure 3. (A-F) A representative sample of ER/ACC, RS, and PCS at the 90-day observation period. (A) ER/ACC: an overview of the area of tissue/material contact showing extrusion (ES) of the sealer (H&E, original magnication 40). (B) A higher magnication of A. In contact with the extruded material, a dense brous connective tissue encapsulation free of inammatory cells is present (black arrow). The white arrow indicates a wide newly formed vessel (H&E, original magnication 850). (C) RS: a low magnication of tissue/material contact (H&E, original magnication 40). (D) A higher magnication of the outlined area in C. Note the presence of a thick brous connective tissue capsule (arrow). A few inammatory cells can be seen within the capsule as well as in the surrounding tissues (H&E, original magnication 850). (E) PCS: an overview of the area of tissue/material contact showing numerous material particles (white arrow) surrounded by granulomatous tissue slightly invaginated within the lumen of the tube (black arrow) (H&E, original magnication 40). (F) A higher magnication of E showing a brous connective tissue (arrow) that appeared to isolate the inammatory reaction from the surrounding tissues (H&E, original magnication 850).(This gure is available in color online at www.aae.org/joe/.)

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1577

Basic ResearchTechnology
was satisfactory at all observation periods. Histological evaluation showed that the implants were surrounded by brous connective tissue of irregular thickness. It could be easily distinguished from the tissue reaction at the site where the tissues were in direct contact with the test material. The severity of tissue reaction is presented in Table 1. After 10 days, the tissue reaction to ER/ACC, RS, and PCS was scored severe and was extensively dispersed around the end of the tubes (Figs. 1A-F). The majority of ER/ACC samples with direct tissue contact presented with an inammatory reaction with slight invagination into the lumen of the tubes. Some ER/ACC and RS samples exhibited a thin necrotic zone in the direct contact area. Each group displayed many newly formed vessels and randomly dispersed dark particles, which appeared to have been released from the implanted materials. A mild to moderate reaction was observed in SIROD implants. After 30 days, the intensity of the inammatory reaction to ER/ACC and RS decreased slightly and was scored as moderate (Fig. 2A-D). In direct tissue contact, both materials displayed a brogranulomatous tissue, which appeared less dispersed and showed a tendency to be surrounded by a brous tissue containing many inammatory cells. In contact with PCS, a persistent severe granulomatous tissue invagination containing many dark particles was observed. This granulomatous reaction was surrounded by an incipient layer of brous tissue, which was free of inammatory cells (Fig. 2E and F). In contact with SIRODs, there was a thin brous connective tissue containing a few inammatory cells. After 90 days, some ER/ACC samples showed extrusion of the material into the surrounding tissues. In seven samples, a thick dense and mature brous tissue capsule (approximately 70- to 90-mm thick) free of inammatory cells was present (Fig. 3A and B), whereas only one sample had a mild tissue reaction. The RS samples showed a dense brous capsule (50- to 80-mm thick) without inammatory cells, whereas two cases were scored as mild with a few inammatory cells still persisting within the capsule (Fig. 3C and D). PCS samples exhibited a severe persistent granulomatous tissue reaction containing necrotic areas as well as randomly distributed inammatory cells. Particles of the material were observed at the end and within the lumen of the tubes in all samples. Higher magnications showed the granulomatous tissues to be isolated by a thick brous tissue encapsulation. Many newly formed capillaries and a high concentration of acute and chronic inammatory cells were also seen (Fig. 3E and F). A thick dense brous connective tissue, free of inammatory cells, was observed in all SIROD samples (Fig. 4). The Wilcoxon signed rank test showed no statistically signicant differences (p > 0.05) between the reaction to ER/ACC, RS, and PCS at the 10-day observation period. After 30 days, no statistical differences were found between ER/ACC and RS (p > 0.05), whereas both ER/ACC and RS signicantly differed from PCS (p < 0.05). For both periods, ER/ ACC, RS, and PCS differed signicantly from SIRODs. After 90 days, ER/ ACC and RS differed signicantly from PCS (p < 0.05) but not from SIRODs, whereas PCS remained signicantly different from SIRODs. The effect of time for ER/ACC and RS showed that the 10 and 30 days differed signicantly (p < 0.05) from the 90-day observation period. However, no signicant differences (p > 0.05) were observed for PCS between all time intervals.

Figure 4. (A) A representative specimen of SIRODs after the 10-day observation period. In contact with the material (empty space), there is a moderate brogranulomatous tissue (black arrow) with inammatory cells and newly formed capillaries (H&E, original magnication 100). (B) A representative specimen of SIRODs after 90 days. In contact with the material (empty space), a thick dense brous connective encapsulation (arrow) without inammatory cells can be seen (H&E, original magnication 850).(This gure is available in color online at www.aae.org/joe/.)

Discussion
The implantation of endodontic lling materials into subcutaneous connective tissue of rats is a valid screening method for testing biocompatibility (18, 19). PCS was used as the positive control because its toxicity has been previously determined by in vitro (20, 21) and in vivo experiments (22, 23). SIRODs were used as the negative control because they have been proven to be biocompatible (5, 6, 24, 25), a nding that was conrmed in this study. The initial inammatory 1578

reaction of SIRODs may be a consequence of the surgical trauma. This subsided rapidly, and after 90 days a well-organized healthy dense brous tissue free of inammatory cells was observed. The biocompatibility of ER has previously been tested (46); however, the incorporation of the ACC changes not only the original chemical composition of ER but also gives the sealer a shorter setting time; an evaluation of the toxicity is therefore essential to determine its safety. The preparation of ER/ACC and RS samples was performed in such a way that no oxygen-inhibited layer was formed. Oxygen inhibits freeradical polymerization of resin-based materials yielding an uncured surface layer (26, 27), which is of particular concern for MBRS because it greatly affects the outcome in toxicity tests. Elution of uncured chemical components from the oxygen-inhibited layer is one of the main causes for tissue damage. However, severe reactions were observed for ER/ACC and RS at the short-term period, revealing that even after setting they still are an irritant. As has been shown (28), unreacted monomer persists in polymerized methacrylate resins. These unreacted monomers (28) undergo a rapid elution and the leaching of these unbound molecules (29), and other components caused the severe inammatory reactions observed after 10 days. These ndings are consistent with those of Costa et al (30, 31) and are supported by previous experiments of Ferracane and Condon (12)

Zmener et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
who found that the loss of components from a methacrylate resin-based material in a water-based medium in vitro is initially rapid and, therefore, a similar behavior can be expected in vivo. It is signicant to note that the initial severe reaction decreased over time and had been resolved at the end of the experiment although a few persistent inammatory cells still remained adjacent to RS. After the rapid initial loss, the material depletes progressively slower over time (12), thus causing less irritation. Therefore, the setting time is signicant with respect to irritation. ER/ACC sets in 7 minutes, RS in 25 minutes, and PCS in 260 minutes (Zmener and Pameijer, unpublished data, 2009). The faster setting of ER/ACC could be a contributory factor in reducing the release of components, in effect locking in potential irritants (12). The tissue reaction to PCS was severe at all observation periods and is most likely caused by unreacted eugenol, which is highly toxic to tissues (32). Our ndings for PCS are in agreement with others who reported that, like other zinc oxide and eugenol-based sealers (3335), PCS reacted toxic in cell cultures (20, 21, 36) and in contact with living tissues (22, 3739). However, the results for ER/ ACC or RS disagree with those of Scarparo et al (39) who showed intense tissue reactions to ER. Their observations did not extend beyond 60 days, however. In the study reported here, a 90-day observation period was used. Longer time periods allow more time for elution of components causing a depletion of chemical components that are potential irritants. On the other hand, shortening of setting time may contribute to a lower output of uncured components from ER as shown here. Within the limitations of this study, it was concluded that ER/ACC and RS exhibited similarly and were both well tolerated by the subcutaneous connective tissue of rats after 90 days of implantation. PCS, however, remained toxic even after a 90-day observation period.
12. Ferracane JL, Condon JR. Rate of elution of leachable components from composite. Dent Mat 1990;6:2827. 13. Ferracane JL. Elution of leachable components from composites. J Oral Rehabil 1994;21:44152. 14. Augsburger RA, Peters DD. Radiographic evaluation of extruded obturation materials. J Endod 1990;16:4927. 15. ISO 10993-1:1992, Biological Evaluation of Medical DevicesPart 1: Guidance on Selection of Tests. Geneva: ISO; 1992. 16. ISO 10993-2:1992, Biological Evaluation of Medical DevicesPart 2: Animal Welfare Requirements. Geneva: ISO; 1992. 17. Bayne K. Developing guidelines of the care and use of animals. Ann N Y Acad Sci 1998;30:10510. 18. Olsson B, Sliwkowsky A, Langeland K. Subcutaneous implantation for the biological evaluation of endodontic materials. J Endod 1981;7:35569. 19. FDI. Federation Dentaire Internationale. Recommended standard practices for the biological evaluation of dental materials. Int Dent J 1980;30:1746. 20. Zmener O, Cabrini RL. Monocyte-lymphocyte adhesion to three zinc oxide and eugenol-based endodontic cements: an in vitro assay. Acta Odontol Latinoam 1986;3:2732. 21. Pinna L, Brackett MG, Lockwood PE, et al. In vitro cytotoxicity evaluation of a selfadhesive methacrylate resin-based root canal sealer. J Endod 2008;34:10858. 22. Pascon EA, Leonardo MR, Safavi K, et al. Tissue reaction to endodontic materials: methods, criteria, assessment and observations. Oral Surg Oral Med Oral Pathol 1991;72:22237. 23. Kolokouris I, Economides N, Beltes P, et al. In vivo comparison of the biocompatibility of two root canal sealers implanted into the subcutaneous connective tissue of rats. J Endod 1998;24:825. 24. Zmener O, Guglielmotti B, Cabrini RL. Biocompatibility of two calcium hydroxidebased endodontic sealers: a quantitative study in the subcutaneous connective tissue of the rat. J Endod 1988;14:22935. 25. Zmener O, Guglielmotti B, Cabrini RL. Tissue response to an experimental calcium hydroxide-based endodontic sealer: a quantitative study in the subcutaneous connective tissue of the rat. Endod Dent Traumatol 1990;6:6672. 26. Andrzejewska E, Linden LA, Rabek JF. The role of oxygen in camphorquinone-initial photopolymerization. Macromol Chem Phys 1998;199:4419. 27. Versiani MA, Carvalho-Junior JR, Padilha MIAF, et al. A comparative study of physicochemical properties of AH Plus and Epiphany root canal sealants. Int Endod J 2006;39:46471. 28. Rueggeberg FA, Margeson DH. The effect of oxygen inhibition on an unlled/lled composite system. J Dent Res 1990;69:16528. 29. Rathbun MA, Craig RG, Hanks CT, et al. Cytotoxicity of a BIS-GMA dental composite before and after leaching in organic solvents. J Biomed Mater Res 1991;25: 44357. 30. Costa CAS, Hebling J, Teixeira MF. Preliminary study of the biological compatibility of the dentine adhesives All-bond 2 and Scotchbond MP. Histological evaluation of subcutaneous implants in rats. Rev Odontol USP 1997;11:118. 31. Costa CAS, Teixeira HM, Lopes Nascimento AB, et al. Biocompatibility of two current adhesive resins. J Endod 2000;26:5126. 32. Molnar EJ. Residual eugenol from zinc-oxide-eugenol compounds. J Dent Res 1967; 46:6459. 33. Rodrigues H, Spangberg L, Langeland K. Biologic effects of dental materials. 9. Effect of zinc oxide-eugenol cements on He La cells in vitro. Estomat Cult 1975;9:1914. 34. Zmener O, Cabrini RL. Effects of three calcium hydroxide-based materials on human blood monocytes and lymphocytes. Endod Dent Traumatol 1987;3:2832. 35. Zmener O, Goldberg F, Cabrini RL. Effects of two gutta-percha formulations and one zinc oxide-eugenol and Canada balsam mixture on human blood monocytes and lymphocytes. Endod Dent Traumatol 1989;5:737. 36. Geurtsen W, Leyhausen G. Biological aspects of root canal lling materialshistocompatibility, cytotoxicity and mutagenicity. Clin Oral Invest 1997;1:511. 37. Tagger M, Tagger E. Subcutaneous reactions to implantation of tubes with AH26 and Grosmans sealers. Oral Surg Oral Med Oral Pathol 1986;62:43440. 38. Yesilsoy C, Koren LZ, Morse DR, et al. A comparative tissue toxicity evaluation of established and newer root canal sealers. Oral Surg Oral Med Oral Pathol 1988; 65:45967. 39. Scarparo RK, Grecca FS, Fachin EVF. Analysis of tissue reactions to methacrylate resin-based, epoxy resin-based, and zinc oxide-eugenol endodontic sealers. J Endod 2009;35:22932.

Acknowledgment
Dr. Pameijer is a consultant for Ultradent Products Inc.

References
1. Nguyen TN. Obturation of the root canal system. In: Cohen S, Burns RC. Pathways of the Pulp. 4th ed. St Louis: Mosby; 1987:18394. 2. Taintor JF, Ross PN. Opinions and practices of American Endodontic Diplomates. Dent J 1978;44:3215. 3. ADA Council of Scientic affairs. Statement on posterior resin-based composites. ADA Council on Dental Benet Programs. J Am Dent Assoc 1998;129:16278. 4. Louw NP, Pameijer CH, Norval G. Histopathological evaluation of a root canal sealer in subhuman primates. J Dent Res 2001;80:654. 5. Zmener O. Tissue response to a new methacrylate-based root canal sealer: preliminary observations in the subcutaneous connective tissue of rats. J Endod 2004;30:34851. 6. Zmener O, Banegas G, Pameijer CH. Bone tissue response to a methacrylate-based endodontic sealer: a histological and histometric study. J Endod 2005;31:4579. 7. Sousa CJA, Montes CRM, Pascon EA, et al. Comparison of the intraosseous biocompatibility of AH Plus, EndoREZ and Epiphany root canal sealers. J Endod 2006;32: 65662. 8. Zmener O, Pameijer CH. Clinical and radiographic evaluation of a resin-based root canal sealer. Am J Dent 2004;17:1922. 9. Zmener O, Pameijer CH. Clinical and radiographical evaluation of a resin-based root canal sealer: a 5-year follow-up. J Endod 2007;33:6769. 10. Cotton TP, Schindler WG, Schwartz SA, et al. A retrospective study comparing clinical outcomes after obturation with Resilon/Epiphany or gutta-percha/Kerr sealer. J Endod 2008;34:78997. 11. Zmener O, Pameijer CH, Alvarez Serrano S. Effect of immediate and delayed post space preparation on coronal bacterial microleakage in root canals obturated with a methacrylate-based sealer with an without accelerator. An ex vivo study. Am J Dent 2010;23:11620.

JOE Volume 36, Number 9, September 2010

Subcutaneous Connective Tissue Reaction to Methacrylate Resinbased Sealer

1579

Basic ResearchTechnology

Efcacy of Ultrasonic versus Laser-activated Irrigation to Remove Articially Placed Dentin Debris Plugs
Roeland J.G. De Moor, DDS, MSc, PhD,* Maarten Meire, DDS, MSc,* Kawe Goharkhay, DMD, MD, Andreas Moritz, DMD, DM, PhD, and Jacques Vanobbergen, DDS, PhD
Abstract
Introduction: The study assessed the efcacy of laser activated irrigation (LAI) with Erbium: Yttrium Aluminum Garnet (Er:YAG) and Erbium Chromium: Yttrium Scandium Gallium Garnet (Er,Cr:YSGG) wavelengths as compared with passive ultrasonic irrigation (PUI). Previously proposed irrigation times were used for LAI (4 5 seconds) and the intermittent ush technique (3 20 seconds). Methods: We used a split root model with an articial root canal wall groove. Roots were prepared to an apical size # 40 with ProFiles 0.06 (Dentsply Maillefer, Baillaigues, Switzerland). Five groups of 20 straight canine roots were evaluated as follows: Group 1: hand irrigation for 20 s with 2.5% NaOCl (CI); Group 2: PUI performed once for 20 s with the #20 Irrisafe (Satelec Acteon group, Merignac, France) (PUI 1); Group 3: PUI for 3x 20 s with the Irrisafe (PUI 2); Group 4: LAI with the Er,Cr:YSGG laser and Z2 (200 mm) Endolase tip (Biolase, San Clemente, USA) at 75 mJ for 4x 5 s (LAI 1); Group 5: LAI with the Er:YAG laser (HoYa Versawave, Cortaboeuf, France) and a 200 mm endodontic ber at 75 mJ for 4x 5 s (LAI 2). Images from the groove were taken before and after irrigation. The quantity of dentin debris in the groove after the experimental protocols was evaluated. Results: Statistically signicant differences (p < 0.05) were found between CI and all other groups and between PUI 1 and the other groups. Conclusion: LAI techniques using erbium lasers (Er:YAG or Er,Cr:YSGG) for 20 seconds (4 5 seconds) are as efcient as PUI with the intermittent ush technique (3 20 seconds). (J Endod 2010;36:15801583)

Key Words
Irrigation, laser, root canal, ultrasound

ffective endodontic treatment requires the combination of physical and chemical agents to eradicate soft-tissue debris, smear layer, and microorganisms because buildup of debris in the root canal system makes effective cleaning and disinfection impossible. The use of lasers at different wavelengths has been proposed to supplement conventional endodontic cleaning procedures (14). A considerable limitation, however, is the unidirectional emission of the laser beam, which makes it difcult to access the entire root canal wall with the laser. The laser ber must be moved repeatedly in a spiraling motion along the root canal walls in order to maximize the area exposed to the laser beam, but even this is not completely efcient and the entire root canal wall will not be exposed to the laser beam (2, 4). Alternative approaches such as side-ring tips have limited use because of their size (4) or require further investigation before clinical application (5, 6). Laser-activated irrigation (LAI) with an erbium laser has been introduced as a method for activating the irrigant (510). The effect is based on cavitation; in water, activation of the laser at subablative settings may result in the formation of large elliptical vapor bubbles, which expand and implode. These vapor bubbles may cause a volumetric expansion of 1,600 times the original volume, which increases pressure and drives uid out of the canal. When the bubble implodes after 100 to 200 microseconds, an underpressure develops and sucks uid back into the canal, inducing secondary cavitation effects. Therefore, the laser works as a uid pump. Another technique, passive ultrasonic irrigation (PUI), is also based on the principle of cavitation and acoustic streaming (11). The ultrasonic activation of irrigants therefore plays a pivotal role in contemporary endodontics (12, 13). The removal of dentin debris from the root canal using LAI has been investigated in only two studies (10, 11). Both studies, de Moor et al (10) with an Er,Cr:YSGG laser (2,780 nm) and de Groot et al (11) with an Er:YAG laser (2,940 nm), have shown that LAI is signicantly more effective in removing dentin debris from the apical part of the root canal than PUI or hand irrigation when the irrigant was activated for 20 seconds. It remains unknown (1) whether the use of PUI for more than 20 seconds (3 20 seconds according to van der Sluis et al [14]) is as effective as 20 seconds of LAI , and (2) whether there is a difference between the efcacy of LAI performed with an Er:YAG laser or Er,Cr:YSGG laser (both erbium lasers, with different wavelengths, 2,780 nm and 2,940 nm, respectively). Therefore, the aim of the present study was to evaluate ex vivo the removal of articially placed dentin debris in standardized root canals by (1) active hand irrigation for 20 seconds, (2) PUI with an Er:YAG or Er,Cr:YSGG laser, and (3) LAI for 20 seconds and 3 20 seconds.

From the Departments of *Operative Dentistry and Endodontology and Community Dentistry and Oral Public Health, Dental School, Ghent University, Ghent University Hospital, Ghent, Belgium; and Department of Conservative Dentistry, Dental School, Bernhard Gottlieb University Clinic of Dentistry, Vienna, Austria. Address requests for reprints to Dr Roeland J.G. De Moor, Department of Restorative Dentistry and Endodontology, Ghent University Hospital, Dental School, De Pintelaan 185, B-9000 Gent, Belgium. E-mail address: Roeland.DeMoor@UGent.be. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.06.007

Material and Methods

Sample Selection and Preparation For the setup of this study, an experimental root canal model described by Lee et al (15) was used (Fig. 1). One hundred maxillary canines with straight roots were selected. These roots were mounted and prepared as described in de Moor et al (9). After verication of the location of the apical foramen with an ISO size 15 le through the apical foramen, the teeth were decapitated at 19 mm of the location of the apical foramen with a diamond disc (Horico, Berlin, Germany). The coronal 3 mm of the canals were enlarged by a round bur with a diameter of 2.3 mm (Komet, Dusseldorf, Germany, 340.202.001.001.023, American size 8) and simulating the

1580

De Moor et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology

Figure 1. A schematic representation of the specimen preparation. On one half of the instrumented root canal, a groove was cut 2 to 6 mm from the apex. In the coronal 3 mm, an articial pulp chamber was prepared over a length of 3 mm (diameter = 2.3 mm). (This gure is available in color online at www.aae.org/joe/.)

pulp chamber (Fig. 1). The root canals were then instrumented with the ProFile system (Dentsply Maillefer, Baillaigues, Switzerland) until size 40, taper 0.06 (master apical le) 1 mm short of the apical foramen (working length). The standardization of the root canal outline was checked; pictures of the canals (3,040 2,040 pix) were taken, and the diameter of 10 randomly chosen models was measured at 2, 6, and 10 mm from the apical end of the canal using image analysis software (Sigmascan Pro Image Analysis Version 5.0; SPSS Inc., Chicago, IL) (9). After instrumentation, the roots were longitudinally split through the canal. A standard groove 4 mm in length, 0.2 mm in width, and 0.5 mm in depth was cut in one canal wall 2 to 6 mm from the apex; the dimension of the groove is comparable to an apical oval root canal (15). Each groove was lled with dentin debris that had been mixed for 5 minutes with 2% NaOCl to simulate a situation in which dentin debris accumulates in uninstrumented canal extensions during root canal preparation. This model was introduced to standardize the root canal anatomy and the amount of dentin debris present in the root canal before the irrigation procedure, and it was intended to increase the reliability of the evaluation of dentin debris removal. The methodology is sensitive, and the data are reproducible (14). So, a total of 100 blocks with root canalprepared roots and plugged root canal wall grooves were used for the experiment and randomly divided in ve groups of 20.

Irrigation Protocols Five irrigation protocols were investigated: conventional irrigation (CI) with 2.5% sodium hypochlorite (NaOCl), two protocols using PUI, and two protocols using LAI. In group 1 (n = 20), hand irrigation with 4 mL 2.5% NaOCl was performed with a 10-mL syringe (Terumo, Leuven, Belgium) and a 27-G endodontic needle (Monoject; Sherwood Medical, St Louis, MO). For 20 seconds, the needle was inserted 1 mm short of the working length and moved slowly over a distance of 4 mm up and
JOE Volume 36, Number 9, September 2010

down in the apical half of the root canal. The ow rate was approximately 0.3 mL/s, and the total irrigant volume was 6 mL. In group 2 (n = 20) (PUI 1), a stainless steel noncutting wire (#20) (Irrisafe, Satelec Acteongroup) was used driven by an ultrasonic device (Suprasson Pmax Newtron; Satelec, Acteongroup, Merignac, France) at power setting blue 4 (frequency 30 KHz, displacement amplitude about 30 mm according to the manufacturer) for 20 seconds. The tip of the Irrisafe was kept 1 mm from the apical stop. After this procedure, the canal was ushed with 2 mL 2.5% NaOCl with a syringe and a 27G endodontic needle (Monoject). In group 3 (n = 20) (PUI 2), the same protocol as in group 2 was performed but with a 3 20 seconds activation of the irrigant. The irrigant was ushed out and renewed after each activation cycle. In group 4 (n = 20) (LAI 1), the NaOCl was activated by laser irradiation (Er,Cr:YSGG laser; Waterlase Millenium, Biolase, San Clemente, CA) using an endodontic ber (Z2, Endolase Tip, Biolase) with a diameter of 200 mm and 25-mm length, with pulse energies of 75 mJ at 20 Hz. The ber was kept 5 mm away from the most apical preparation and then kept stationary for 5 seconds. A mark was put on the ber with a black marker at 13 mm in order to position it in the root canal at this depth. This protocol was repeated four times, without removing the tip from the root canal. At the end of this procedure, the canal was ushed with 2 mL 2.5% NaOCl using a syringe with a endodontic needle. In group 5 (n = 20) (LAI 2), the same protocol as in group 4 was followed but with an Er:YAG laser (HoYa Versawave, Cortaboeuf, France) and a 200-mm endodontic ber at 75 mJ. The irrigant was also activated for 4 5 seconds. After irrigation, the canals were carefully dried with paper points (size 35).

Evaluation of Dentin Debris Removal After unlocking the two halves, the amount of remaining debris in the articial grooves was evaluated. Digital images of the groove were
Ultrasonic vs Laser-activated Irrigation

1581

Basic ResearchTechnology
taken before and after irrigation at 40 magnication. These images were scored by two dentists who were unaware of the irrigation techniques used. Scores were assigned based on the following system: 0: the groove is empty, 1: less than half of the groove is lled with dentin debris, 2: more than half of the groove is lled with dentin debris, and 3: the groove is completely lled with dentin debris. 7, 8, 10). This was also shown in the present study. Apparently, there is no difference in the efcacy of both wavelengths in terms of smear layer removal at the settings used in this study. The present study is the rst to compare the effects of Er:YAG and Er,Cr:YSGG wavelengths on laser-induced cavitation for smear layer removal. Both wavelengths have a good absorption in water and sodium hypochlorite (18). The laser light is used here at subablative settings, which does not damage the root canal wall and hence the formation of ledges is avoided (2, 4). The ber is also inserted centrally in the root canal without contact with the root canal wall and kept stationary during emission. Also, no spiral motion is made in the irrigant, which is needed when the laser ber is used in the conventional way and when the whole root canal wall has to be exposed directly to the laser light. In this respect, the risk of ledge creation appeared to be greater with Er:YAG than with Er,Cr:YSGG (4, 19). Care must be taken, however, when using the laser bers in the root canal, as apical extrusion of the irrigant after laser activation has been described at the present power settings (6). A previous study by George et al (6) showed that there was twice as much dye penetration through the apical constriction with the ber tip at 4 mm than at 5 mm. Therefore, a distance of 5 mm from the apical stop to the ber tip was used for the present evaluation. All apical stops were also controlled under magnication 40 (Pico Opmi, Zeiss) by two dentists and no damage or widening of the unprepared last 1 mm up (from the apical stop to the outer surface of the root) was seen. These ndings conrm the ndings of our previous study (9). Another interesting nding is that the intermittent ush technique (PUI 2) was as effective as both LAI protocols (12). Therefore, within the connes of the present study (removal of debris from a root canal wall groove), the use of an expensive laser device to obtain comparable results is not necessary. In the study by de Groot et al (10), however, more efcacy was attributed to LAI with an Er:YAG laser because of the formation of secondary cavitation bubbles. The uid ow associated with such an inertial collapse, combined with acoustic streaming resulting from the oscillations of smaller bubbles, was thought to explain the cleaning efcacy of LAI. The secondary cavitation bubbles allowed better cleaning of the root canal wall because they are excited by the bubble collapse of the consecutive laser pulse. Hence, new research should be focused not only on the removal of debris from the root canal wall but also on the cleanliness of the entire root canal, especially in the apical third, and the debridement of anastomes and isthmuses. In addition, the Er:YAG used at subablative settings is also efcient in removing biolms, even those with Escherichia faecalis (19). Thus, the combination of cavitation and direct interaction with the biolm, with the possibility of sight ring (6) and overcoming the limitations of a straight forwarded

Statistical Analysis The interrater agreement was determined (Cohen kappa), and data were initially analyzed using explorative data analysis. Chi-square tests were performed to determine whether the observed frequencies (counts) markedly differ from the frequencies expected by chance. A correction for multiple testing was performed using the Holm procedure, a variant of the Bonferroni adjustment. It is a step-down procedure in that one starts with the most extreme (ie, the smallest) p value and continues with less extreme p values in the successive rejection decisions. Because of this adjustment, the level of signicance was set between a < 0.05 and a < 0.005.

Results
Table 1 shows the debris scores after irrigation with the ve different techniques. Highly signicant differences were found between CI and both PUI and LAI protocols (p < 0.001). Statistically signicant differences were also observed between PUI 1 and both groups of LAI, LAI 1 and LAI 2 (p < 0.001). The Cohen kappa coefcient of interrater agreement was 0.85.

Discussion
The model chosen for this study on the effectiveness of irrigation in removing dentin debris in articial irregularities and extensions allows the comparison of the presence of debris before and after irrigation. In most studies, the amount of debris is evaluated only after preparation and irrigation. Furthermore, these studies did not report how much debris was present before irrigation and are therefore unable to establish the extent of removal using the different irrigation procedures. In the present study, the groove cut in the root canal wall is made in order to simulate uninstrumented extensions in the apical half. Canines were used because they have wide canals (16, 17), and, hence, the root can be more easily split in the mesiodistal direction. The generation of shockwaves by dental lasers inside root canals can play an important role in smear layer removal (510). Similarly, smear layer removal can be achieved when water is activated in root canals using erbium lasers (Er,Cr:YSGG or Er:YAG) (6, 9, 10), causing the formation of vapor bubbles that expand and implode (5,

TABLE 1. Dentin Debris Scores After Conventional Irrigation With 2.5% NaOCl, Passive Ultrasonic Irrigation, Er:YAG LaserActivated Irrigation, and Er,Cr:YSGG LaserActivated Irrigation Debris scores (%) Irrigation technique
CI* PUI 1*, PUI 2*, LAI 1*, LAI 2*,

0
6 (30) 12 (60) 15 (75) 16 (80)

1
7 (35) 6 (30) 5 (25) 4 (20)

2
5 (25) 7(35) 2 (10)

3
15 (75)

N (total)
20 20 20 20 20

Debris scores: 0: the groove is empty, 1: less than half of the groove is lled with dentin debris, 2: more than half of the groove is lled with dentin debris, 3: the groove is completely lled with dentin debris. CI, conventional hand irrigation during 20 seconds with 2.5% NaOCl; PUI 1, PUI during 20 s with the #20 Irrisafe (Satelec Acteon Group, Merignac, France); PUI 2, PUI during 3 20 seconds with the Irrisafe; LAI 1, LAI with the Er,Cr:YSGG laser and Z2 (200 mm) Endolase tip (Biolase, San Clemente, CA) at 75 mJ during 4 5 seconds; LAI 2, LAI with the Er:YAG laser (HoYa Versawave, Cortaboeuf, France) and a 200mm endodontic ber at 75 mJ during 4 5 seconds. *Statistically signicant differences between groups: CI-PUI 1, CI-PUI 2, CI-LAI 1, and CI-LAI 2 (p < 0.001). Statistically signicant differences between groups: PUI 1-LAI 1 and PUI 1-LAI 2 (p < 0.001).

1582

De Moor et al.

JOE Volume 36, Number 9, September 2010

Basic ResearchTechnology
laser beam (4), might nally result in an added value for erbium lasers in root canal treatment when used at subablative settings (20).
9. De Moor RJG, Blanken J, Meire M, et al. Laser activated irrigation or cavitation causing laser driven irrigation. Part 2: evaluation of the efcacy. Lasers Surg Med 2009;41:5203. 10. de Groot SD, Verhaagen B, Versluis M, et al. Laser-activated irrigation within root canals: cleaning efcacy and ow visualization. Int Endod J 2009;42:107783. 11. van der Sluis LMW, Wu M-K, Versluis M, et al. Passive ultrasonic irrigation of the root canal: a review of the literature. Int Endod J 2007;40:41526. 12. Plotino G, Pameijer CH, Grande NM, et al. Ultrasonics in endodontics: a review of the literature. J Endod 2007;33:8195. 13. Gu LS, Kim JR, Ling J, et al. Review of contemporary irrigant agitation techniques and devices. J Endod 2009;35:791804. 14. van der Sluis LMW. Passive ultrasonic irrigation of the root canal [PhD thesis]. ACTA (Academisch Centrum Tandheelkunde Amsterdam), Amsterdam, The Netherlands; 2007. 15. Lee SJ, Wu MK, Wesselink PR. The effectiveness of syringe irrigation and ultrasonics to remove debris from simulated irregularities within prepared root canal walls. Int Endod J 2004;37:6728. 16. Kerekes K, Tronstad L. Morphometric observations on root canals of human anterior teeth. J Endod 1977;3:249. 17. Wu M-K, Roris A, Barkis D, et al. Prevalence and extent of long oval canals in the apical third. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;89: 73943. 18. De Moor RJG, Delme KIM. Laser-assisted cavity preparation and adhesion to erbium-lased tooth structure: part 1. Laser-assisted cavity preparation. J Adhes Dent 2009;11:42738. 19. Varella CH, Pileggi R. Obturation of root canal system treated by Cr, Er:YSGG laser irradiation. J Endod 2007;33:10913. 20. Meire M, De Prijck K, Coenye T, et al. Evaluation of Nd:YAG and Er:YAG irradiation and antibacterial photodynamic therapy on Enterococcus faecalis biolms on dentin disks. Int Endod J 2009;42:3519.

Conclusion
LAI techniques with erbium lasers (Er:YAG or Er,Cr:YSGG) for 20 seconds (4 5 seconds) are as efcient as passive ultrasonic irrigation with the intermittent ush technique (3 20 seconds).

References
1. Kimura Y, Wilder-Smith P, Matsumoto K. Lasers in endodontics: a review. Int Endod J 2000;33:17385. 2. Stabholz A, Sahar-Helft S, Moshonov J. Laser in endodontics. Dent Clin North Am 2004;48:80932. 3. Meire M, De Moor RJG. Lasers in endodontics: laser disinfection, an added value? Endod Pract Today 2007;1:15972. 4. De Moor RJG, Torbeyns D, Meire M. Lasers in endodontics. Part 2: root canal wall cleanliness and modication. Endod Pract Today 2009;3:1933. 5. George R, Walsh LJ. Apical extrusion of root canal irrigants when using Er:YAG and Er, Cr:YSGG lasers with optical bers: an in vitro dye study. J Endod 2008;34:7068. 6. George R, Meyers IA, Walsh LJ. Laser activation of endodontic irrigants using improved conical laser ber tips for removing smear in the apical third of the root canal. J Endod 2008;34:15247. 7. Blanken JW, Verdaasdonk RM. Cavitation as a working mechanism of the Er, Cr:YSGG laser in endodontics: a visulatisation study. J Oral Laser Applic 2007;7:97106. 8. Blanken J, Meire M, De Moor RJG, et al. Laser activated irrigation or cavitation causing laser driven irrigation. Part 1: a visualisation study. Lasers Surg Med 2009;41:5149.

JOE Volume 36, Number 9, September 2010

Ultrasonic vs Laser-activated Irrigation

1583

Case Report/Clinical Techniques

Cone-Beam Computed Tomographic Evaluation of Spontaneously Healed Root Fracture

Kaan Orhan, DDS, PhD,* Umut Aksoy, DDS, and Atakan Kalender, DDS, PhD
Abstract
Introduction: Dental trauma can lead to injuries in teeth and their supporting structures, which occurs most commonly in young patients and varies in severity from enamel fractures to avulsions. Root fractures are relatively uncommon among dental traumas, mostly affecting the permanent dentition. It has been reported that root fractures can undergo healing, whereas pulp necrosis can also occur. Methods: In this report, we present a case of the utilization of cone-beam computed tomography (CBCT) in the detection of possible cervical or internal resorption secondary to a mid-horizontal fracture in a 36-year-old male patient. The fractured teeth spontaneously healed and were diagnosed radiographically after 28 years. Initially, conventional radiographs showed fractured fragments with radiolucent lines. Because of possible invasive cervical resorption and doubt over internal resorption in the conventional images, it was decided to examine the teeth with cone-beam computed tomography with three-dimensional reconstructions. Results: The images clearly showed the displaced fragments of the root fracture. Despite the appearance in the conventional radiograph, no evidence of cervical or internal resorption was detected in the teeth other than the healed root fracture. Conclusion: It was concluded that CBCT images should be obtained for root fractures, especially those in which cervical or internal resorption is suspected from routine conventional radiographs. (J Endod 2010;36:15841587)

raumatic injuries occur more commonly in young patients and vary in severity from enamel fractures to avulsions, which are the cause of emergency treatment in dental practice. Root fractures of permanent teeth can be cited as an example of these kinds of injuries. Horizontal root fractures have a relatively low incidence, ranging from 0.5% to 7% when compared with other dental impact injuries. They are frequently seen in the middle third of the root followed by apical and coronal third fractures; the most affected teeth are the maxillary central incisors of male patients (15). Although root fractures can be generally detected shortly after the injury, they are identied occasionally at subsequent routine dental examinations (1, 68). The diagnosis of root fractures is accomplished with clinical and radiographic examination. Clinical examination includes evaluation of the mobility, presence or absence of tenderness and pain to palpation of the soft tissues, and percussion of the teeth and pulp testing. Radiographic examination can establish the diagnosis of root fracture but must be performed carefully. Two or three radiographs taken at various angles may be needed because of the angulation of the fracture. If the x-ray beam does not pass directly through the fracture line, it usually cannot be seen on the radiographs (913). However, in the last decade, with the development of cone-beam computed tomography systems (CBCT), dentoalveolar imaging can be performed with less radiation and greater accuracy (14). In recent studies, it was shown that dental CBCT can be recommended for the assessment root canal systems (15), apical periodontitis (16), and inammatory root resorptions (17) and has been shown to be useful in the diagnosis and management of these kinds of dentoalveolar trauma patients (1316, 18). In this report, we aimed to present a case regarding the use of CBCT in the detection of possible cervical or internal resorption secondary to a spontaneous healed mid-horizontal fracture and to discuss the potential use of CBCT in root fractures.

Case Report
A 36-year-old man was referred to our clinic with the chief complaint of pain in the mandibular right second molar. A routine periapical radiograph revealed suspicion of a horizontal root fracture in the maxillary right central incisor (Fig. 1). The patients medical history was unremarkable, but the dental history revealed that he had been hit in his mouth by his friends head during a game while he was 8 years old. The patient also stated that his central incisors had displaced palatinally, and he had replaced them alone by his hand and did not seek any professional dental care at that time. Although the clinical examination showed palatinally displaced teeth and slight discolorations of the central incisors, the patient had no complaints except for the pain in the molar tooth. Vitality testing with solid carbon dioxide (CO2 ice) and an electric pulp tester (Digitest Parkell, Farmingdale, NY) elicited no response from both the maxillary right and left central incisors. There was no mobility, stulae, or pain to percussion or palpation. However, in the periapical radiograph, slight radiolucent areas in the fracture line were seen, which were interpreted to be possible resorption areas in the teeth (Fig. 1). It was decided to obtain CBCT imaging for further evaluation. CBCT analysis was performed in all three dimensions, axial, sagittal, and cross-sectional images, with a 0.4-mm slice thickness (Newtom 3G; Quantitative Radiology SRL, Verona, Italy). The axial images revealed horizontal root fractures of the maxillary right central and left central incisors. Sagittal and cross-sectional views also showed horizontal root fractures in the middle third with pulp obliterations of the teeth. However, there was no evidence of periradicular pathology and no cervical or internal resorption, except

Key Words
cone-beam computed tomography, diagnosis, endodontic problems, healing, management of endodontic problems, root fracture

From the Departments of *Oral Diagnosis and Radiology and Endodontics, Near East University, Mersin, Turkey. Address requests for reprints to Dr Kaan Orhan, Near East University, Faculty of Dentistry, Department of Oral Diagnosis and Radiology, Mersin 10, Turkey. E-mail address: call53@ yahoo.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.04.004

1584

Orhan et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

Figure 1. (a) A panoramic radiograph and (b) periapical radiograph of the patient showing root fractures. Note the arrows showing radiolucent possible resorption area in the fracture line.

for the spontaneously healed root fracture of the incisors (Fig. 2). Three-dimensional reconstructed images also showed the healed fractures in detail (Fig. 3).

Discussion
Traumatic dental injuries are everyday occurrences in children and adolescents (2, 19, 20). Maxillary central incisors are the most vulnerable to injury, sustaining approximately 80% of all dental injuries, followed by the maxillary lateral and the mandibular incisors (21). The literature indicates that many factors may inuence the type of healing that occurs for root fractures. For root fractures, the

length of time between the trauma and the treatment, the stage of root development, and any associated signs and symptoms of mobility and pain may inuence the type of healing (22). The healing of root fractures with or without initial treatment is reported to occur in up to 80% of cases, whereas pulp necrosis occurred in 20% (2325). When the fracture line allows communication with the oral cavity, immobilization is difcult, and microbial contamination of the pulp with subsequent pulpal necrosis is almost inevitable (4, 7). Root fractures in the apical and middle thirds usually require no immediate treatment, but because of the pathological changes that can occur several years after injuries, long-term follow-up of patients with

Figure 2. Axial, sagittal, and cross-sectional CBCT images showing a healed root fracture of the incisors without evidence of periradicular pathology or cervical or internal resorption.

JOE Volume 36, Number 9, September 2010

CBCT Evaluation of Spontaneously Healed Root Fracture

1585

Case Report/Clinical Techniques

Although in the present case the initial diagnosis showed that CBCT images were superior in diagnostic efcacy to conventional radiographic images, for follow-up the CBCT images should not necessarily replace conventional methods. CBCT studies cause higher radiation exposures (4-20 times greater). From the standpoint of radiation risk, CBCT appears to have three to seven times the risk of a panoramic examination depending on the area examined, the degree of collimation, and the acquisition software version. Thus, the decision to select an imaging modality for diagnostic purposes, as in this case, should be based on the diagnostic yield expected and in accordance with the ALARA (as low as reasonably achievable) principle (31, 33). In conclusion, from the radiation protection point of view, the diagnostic information of CBCT must improve the treatment results; without such a benet, this technique should not be recommended.
Figure 3. A three-dimensional image of root fractures using volumetric rendering software.

References
1. Andreasen FM, Andreasen JO. Crown fractures. Textbook and Color atlas of traumatic injuries to the teeth. 3rd ed. Copenhagen: Munksgaard; 1994:21956. 2. Gomes AP, de Araujo EA, Goncalves SE, et al. Treatment of traumatized permanent inci sors with crown and root fractures: a case report. Dent Traumatol 2001;17:2369. 3. Davidovich E, Heling I, Fuks AB. The fate of a mid-root fracture: a case report. Dent Traumatol 2005;21:1703. 4. Andreasen FM, Andreasen JO. Root fractures. Textbook and Color atlas of traumatic injuries to the teeth. 3rd ed. Copenhagen: Munksgaard; 1994:279313. 5. Caliskan MK. Prognosis of large cyst-like periapical lesions following nonsurgical root canal treatment: a clinical review. Int Endod J 2004;37:40816. 6. Yates JA. Root fractures in permanent teeth: a clinical review. Int Endod J 1992;25: 1507. 7. Trope M, Chivian N, Sigurdsson A, et al. Traumatic injuries. In: Cohen S, Burns RC, eds. Pathways of the pulp. 8th ed. St. Louis: Mosby; 2002:60349. 8. Tziafas D, Margelos I. Repair of untreated root fracture: a case report. Endod Dent Traumatol 1993;9:403. 9. Patel S, Dawood A, Ford TP, et al. The potential applications of cone beam computed tomography in the management of endodontic problems. Int Endod J 2007;40: 81830. 10. Cotton TP, Geisler TM, Holden DT, et al. Endodontic applications of cone-beam volumetric tomography. J Endod 2007;33:112132. 11. Patel S. New dimensions in endodontic imaging: part 2. Cone beam computed tomography. Int Endod J 2009;42:46375. 12. Hassan B, Metska ME, Ozok AR, et al. Detection of vertical root fractures in endodontically treated teeth by a cone beam computed tomography scan. J Endod 2009;35:71922. 13. Patel S, Dawood A, Whaites E, et al. New dimensions in endodontic imaging: part 1. Conventional and alternative radiographic systems. Int Endod J 2009;42:44762. 14. Mozzo P, Procacci C, Tacconi A, et al. A new volumetric CT machine for dental imaging based on the cone-beam technique: preliminary results. Eur Radiol 1998;8:155864. 15. Matherne RP, Angelopoulos C, Kulild JC, et al. Use of cone-beam computed tomography to identify root canal systems in vitro. J Endod 2008;34:879. 16. Estrela C, Bueno MR, Leles CR, et al. Accuracy of cone beam computed tomography and panoramic and periapical radiography for detection of apical periodontitis. J Endod 2008;34:2739. 17. Estrela C, Bueno MR, De Alencar AH, et al. Method to evaluate inammatory root resorption by using cone beam computed tomography. J Endod 2009;35:14917. 18. Bornstein MM, Wolner-Hanssen AB, Sendi P, et al. Comparison of intraoral radiography and limited cone beam computed tomography for the assessment of root-fractured permanent teeth. Dent Traumatol 2009;25:5717. 19. de Blanco LP. Treatment of crown fractures with pulp exposure. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;82:5648. 20. Garcia-Godoy F, Pulver F. Treatment of trauma to the primary and young permanent dentitions. Dent Clin North Am 2000;44:597632. 21. Andreasen FM, Andreasen JO. Crown fractures. Textbook and Color atlas of traumatic injuries to the teeth. 3rd ed. Copenhagen: Munksgaard; 1994. 25777. 22. Feely L, Mackie IC, Macfarlane T. An investigation of root-fractured permanent incisor teeth in children. Dent Traumatol 2003;19:524. 23. Artvinli LB, Dural S. Spontaneously healed root fracture: report of a case. Dent Traumatol 2003;19:646. 24. Camp JH. Management of sports-related root fractures. Dent Clin North Am 2000; 44:95109. 25. Rintaro T, Kiyotaka M, Minoru K. Conservative treatment for root fracture located very close to gingiva. Dent Traumatol 2005;21:1114.

traumatic injuries is very crucial (4, 2628). Davidovich et al (3) reported such a trauma case, which resulted in invasive cervical resorption 13.5 years after the trauma. In that case, early signs of cervical resorption were missed at 8 years in the follow-up radiographic examinations. Because periapical radiographs are two-dimensional images of three-dimensional structures, the superimposition of adjacent structures may obscure the visibility of fractures. A three-dimensional imaging protocol should be used in detecting fractures, especially as an only choice for vertical root fractures (12, 29). Youssefzadeh et al (30) reported an in vivo study that was conducted with medical computed tomography (MDCT) on the detection of vertical root fractures and found out that MDCT was far better than conventional periapical radiographs. However, the radiation dose, the limited availability, and the increased cost impede its use in dentistry (10, 12, 31). CBCT produces a cone shaped x-ray beam, and this makes it possible to capture the image in a single shot, rather than capturing slices separately, as in MDCT. The most important advantage of this imaging modality is acquiring comparative images with a much lower radiation dose than MDCT (1013). In recent studies, it was concluded that CBCT scans are more accurate than periapical radiographs and effective in a safe way to detect both horizontal and vertical root fractures (9, 11, 12, 32). It was also concluded that currently CBCT should be considered when conventional radiographic techniques fail to provide information for diagnosing horizontal root fractures (32). In our case, vitality testing and the electric pulp tester elicited no response from both the maxillary right and left central incisors. There was no mobility, stulae, or pain to percussion or palpation. However, in the periapical radiograph, radiolucent areas in the fracture line were seen, which were interpreted as possible cervical or internal resorption in the teeth. Therefore, it was decided to perform CBCT so as not to the miss any delayed complications of trauma, like resorptive processes that may lead to tooth loss. CBCT images showed no sign of pathology in the teeth in our case. Estrela et al (17) reported in a recent study that inammatory root resorption was determined more accurately and at earlier stages by using CBCT scans than with conventional radiographic images because of the three-dimensional view potential. It was also stated that inammatory root resorption might be underestimated when evaluated with only periapical radiographs. We suggest obtaining three-dimensional CBCT images for the detection of possible cervical or internal resorption in cases similar to ours. 1586

Orhan et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

26. Cheung SP, Walker RT. Root fractures:a case of dental non-intervention. Endod Dent Traumatol 1988;4:1868. 27. Cohen S, Burns RC. Pathways of the pulp. 6th ed. St. Louis: Mosby; 1994: 452. 28. Simon S, Lumley PJ, Cooper PR, et al. Trauma and dentinogenesis: a case report. J Endod 2010;36:3424. 29. Hassan B, Metska ME, Ozok AR, et al. Comparison of ve cone beam computed tomography systems for the detection of vertical root fractures. J Endod 2010;36: 1269. 30. Youssefzadeh S, Gahleitner A, Dorffner R, et al. Dental vertical root fractures: value of CT in detection. Radiology 1999;210:5459. 31. Ludlow JB, Davies-Ludlow LE, Brooks SL. Dosimetry of two extraoral direct digital imaging devices: NewTom cone beam CT and Orthophos Plus DS panoramic unit. Dentomaxillofac Radiol 2003;32:22934. 32. Kamburoglu K, Ilker Cebeci AR, Grondahl HG. Effectiveness of limited cone-beam computed tomography in the detection of horizontal root fracture. Dent Traumatol 2009;25:25661. 33. Angelopoulos C, Thomas SL, Hechler S, et al. Comparison between digital panoramic radiography and cone-beam computed tomography for the identication of the mandibular canal as part of presurgical dental implant assessment. J Oral Maxillofac Surg 2008;66:21305.

JOE Volume 36, Number 9, September 2010

CBCT Evaluation of Spontaneously Healed Root Fracture

1587

Case Report/Clinical Techniques

Cytotoxic Chemotherapy-induced Odontalgia: A Differential Diagnosis for Dental Pain

Yehuda Zadik, DMD, MHA,* Vladimir Vainstein, MD, Ilana Heling, DMD, Tzahi Neuman, MD, Scott Drucker, BSc,k and Sharon Elad, DMD, MSc*
Abstract
Introduction: Peripheral neurotoxicity and neuropathic pain are well-known complications of several anti-cancer chemotherapeutic agents. Such pain might cause an impairment of the patients quality of life and is a common limiting factor of anti-cancer chemotherapy. Neurotoxicity in orofacial structures has been previously described as diffuse jaw pain or numbness. Currently, localized pulpal pain is not listed as a possible complication of cytotoxic therapy. The aim of this report was to suggest cytotoxic-induced neurotoxicity as a differential diagnosis for toothache during anti-cancer therapy. Methods: We described the diagnostic process in a patient suffering from severe pulpal pain in apparently intact teeth during cytotoxic therapy. A non-Hodgkins lymphoma patient complained of 2 episodes of excruciating dental pain evoked by mouth breathing, which caused nocturnal awakenings. Results: Both episodes developed immediately after administrations of cyclophosphamide as part of an anti-cancer chemotherapy protocol. Clinical parameters and radiographic characteristics eliminated other possible dental and facial etiologies. Pulp extirpation (pulpectomy) resulted in immediate pain relief. In both episodes, cytologic evaluation of the extirpated pulp tissue failed to show inammation or an inltration of lymphoma cells. Conclusions: This case presented a circumstantial relation between the clinical presentation of dental pain, with associated signicant impairment of the patients quality of life, and the timing of administrations of high-dose cyclophosphamide. It suggests that chemotherapy-induced toxicity might manifest as pulpitis-like toothache, which might present a diagnostic challenge for the dental practitioner. (J Endod 2010;36:15881592)

rofacial pain in oncology patients is quite common, affecting up to 70% of patients with cancer (1). Non-dental orofacial pain related to malignancy can be attributed to a variety of etiologies including space-occupying lesion, leukemic inltration, secondary anemia, chemotherapy or radiotherapy mucositis, acute and chronic graft-versus-host disease, secondary infection, postsurgery pain, osteoradionecrosis, bisphosphonate-related osteonecrosis of jaws, and nerve damage caused by cytotoxic chemotherapy and/or radiotherapy (2). In general, the vast majority of such pain is directly associated with the tumor itself (in 87%93% of the cases), although much less frequently it is associated with anti-cancer therapy (17%21%) or both (1). Neuropathic pain is induced by a primary lesion or dysfunction of the nervous tissues and might be triggered by local or systemic conditions (such as local injury or diabetes mellitus, respectively) that affect structures along the neuraxis (3). Approximately 50% of pain in cancer patients can be categorized as exclusively or partly neuropathic (4). The nerve damage associated with the administration of chemotherapeutic agents might be manifested as purely sensory neuropathy, mixed sensorimotor neuropathy (with or without the involvement of autonomic nervous system), and/or mental status changes. These effects characteristically result from accumulation of the cytotoxic agent throughout multiple cycles of therapy; symptoms disappeared within days of each administration of chemotherapy, but their severity and duration increased over subsequent cycles (5). Neuropathic pain induced by chemotherapy is difcult to control with analgesic medications (4) and is thus a well-recognized clinical problem. In addition to its inuence on the patients quality of life, many patients are unable to complete their necessary treatment cycles because of this neuropathy (6, 7). McCarthy and Skillings (8) reported orofacial complications among chemically treated breast cancer patients. As high as 65% of their patients suffered from orofacial neurotoxicity during a 7-week course; of these patients, 86% suffered from orofacial pain. The aim of our study was to report a case of severe dental pain in 2 teeth occurring on 2 separate episodes in a lymphoma patient and to discuss the diagnosis of pulpal neuropathic pain caused by anti-cancer drugs.

Case Report
A 54-year-old man bedridden with non-Hodgkins lymphoma presented with 2 episodes of intense dental pain. He had been diagnosed 6 months earlier with large B-cell lymphoma primarily involving bone marrow, mediastinum, and the central nervous system. The lymphoma had manifested as paraparesis, transverse myelitis, and mediastinal lymphadenopahy. An anti-lymphoma protocol of rituximab, cyclophos-

Key Words
Cancer, chemotherapy, cyclophosphamide, dental, neuropathy, pain, tooth

From the *Department of Oral Medicine, Hebrew University-Hadassah School of Dental Medicine, Jerusalem, Israel; Department of Hematology, Hadassah University Medical Center, Jerusalem, Israel; Department of Endodontics, Hebrew University-Hadassah School of Dental Medicine, Jerusalem, Israel; and Department of Pathology, Hadassah University Medical Center, Jerusalem, Israel; and kSchool of Dental Medicine, University of Pennsylvania, Philadelphia, Pennsylvania. Address requests for reprints to Dr Yehuda Zadik, Hebrew University-Hadassah School of Dental Medicine, Department of Oral Medicine, P.O. Box 91120, Jerusalem, Israel. E-mail address: yzadik@gmail.com. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.004

1588

Zadik et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

Figure 1. Timing of dental pain in relation to anti-lymphoma protocol.

phamide, and doxorubicin was administrated to the patient, accompanied by intrathecal injections of cytarabine infused through an Ommaya catheter (Fig. 1). Other medications taken by the patient were aspirin, folic acid, lactulose, mesna, leucovorin, methenamine, simvastatin, sulfamethoxazole/trimethoprim, and tamsulosin. Fig. 1 presents the timing of the 2 episodes of dental pain with relation to the chemotherapeutic treatments. The patients medical history included hyperlipidemia, nonalcoholic fatty liver, and benign prostatic hyperplasia.
TABLE 1. Description of the 2 Dental Pain Episodes Episode
Pain timing Symptoms

Dental pain was located initially in the right mandibular molar region and later in the left maxillary molar region. The pain was rated by the patient as 9 and 10 on a 010 pain scale in the rst and second episodes, respectively (Table 1). During the 2 episodes, pain was evoked by mouth breathing and by drinking cold beverages. The pain caused nocturnal awakenings. Eating did not evoke pain. There was no history of dental pain during the year preceding chemotherapy. The patient did not report any other (extraoral) neuropathic pain.

First episode
On the third day of the rst administration of cyclophosphamide (8/2008). Intense pain (VAS 9).* The pain awakened him from sleep and intensied with drinking cold liquids and air stimulation. Normal mucosal tissue. Second right mandibular molar: No dental decay, intact pre-existing restoration, adequate contact area with adjacent tooth, no occlusal abrasion, no cervical lesion. Periodontal status: probing depth up to 3 mm, non-tender. Cold stimuli yielded pain that remained for 3.5 minutes after removal of stimulus. Percussion (vertical/horizontal) and vestibular palpation tests yielded intense pain. (1) Malignant pulpal inltrate (2) Chemotherapy-induced neurotoxicity Second mandibular right molar: Pulp tissue was sampled to cytology evaluation. Pulpectomy (tetracycline/triamcinolone intracanal dressing). Pain resolved immediately. Cytologic examination did not reveal B cells in the amputated pulp or inammatory cells.

Second episode
On the rst day of the second administration of cyclophosphamide (9/2008). Intense pain (VAS 10).* The pain awakened him from sleep. No tenderness over the maxillary sinuses. Normal mucosal tissue. First left maxillary molar: No dental decay, intact pre-existing restoration, no occlusal abrasion, no cervical lesion. Periodontal status: probing depth up to 3 mm, non-tender. Cold stimuli yielded intense pain (which subsided on removal of stimulus). Percussion (vertical/horizontal) and vestibular palpation tests yielded intense pain. (1) Malignant pulpal inltrate (2) Chemotherapy-induced neurotoxicity First maxillary left molar: Pulp tissue was sampled to cytology evaluation. Pulpectomy (tetracycline/triamcinolone intracanal dressing). Pain resolved immediately. Cytologic examination did not reveal B cells in the amputated pulp or inammatory cells.

Oral evaluation

Differential diagnosis Management

Follow-up

*Visual analogue scale of 010, with 0 as no pain and 10 as the most intense pain.

JOE Volume 36, Number 9, September 2010

Cytotoxic Chemotherapy-induced Odontalgia

1589

Case Report/Clinical Techniques

Figure 2. Periapical radiographs of the affected teeth in the rst (a) and second (b) episodes.

Examination included clinical (vestibular) palpation and percussion, cold tests, periodontal probing, and radiography (Fig. 2; Table 1). No dental caries lesions were observed in the affected teeth. Physical examination eliminated other possible etiologies such as caries with pulp involvement, lateral periodontal abscess or perio-endo lesion, cracked tooth, vertical root fracture, and food impaction as well as a mucosal origin for the pain. Pain from the opposing teeth was ruled out as well. Periodontal probing depth up to 3 mm was measured around the tooth. Percussion (vertical/horizontal) and vestibular palpation tests yielded intense pain. Cold stimuli resulted in pain lasting 3.5 minutes after removal of the stimuli. Results of blood tests taken on the day of initial presentation (August 10th) were the following: white blood cell count, 10.7 109/L (normal limits, 410 109/L) with 90.2% (9.7 109/L) neutrophils; red blood cell count, 3.42 1012/L (normal, 4.56 1012/L); hemoglobin level, 9.7 g/dL (normal, 1418 g/dL); platelet count, 292.0 109/L (normal, 140400 109/L); alanine aminotransferase level, 25 U/L (normal, 653 U/L); aspartate aminotransferase level, 20 U/L (normal, 260 U/L); fasting glucose level, 4.9 mmol/L (normal, 46 mmol/L); and blood creatinine level, 49 mmol /L (normal, 58110 mmol/L). Blood tests taken on September 21st after his second episode yielded similar results. During the rst episode no denitive diagnosis was reached, but a working diagnosis of direct inltration of lymphoma cells into the pulp chamber causing local pressure and pain was made. This led to the assumption that the chemotherapy would reduce the presumed lymphomatous inltrate and would therefore help resolve the pressure and pain. On the basis of this initial diagnosis and the fact that the pain was not attributed to a known dental pathology, the decision was made by the patient and practitioners not to perform an irreversible intervention for pain relief (ie, pulpectomy), but rather to manage the pain by powerful analgesics. The patient received oxycodone 30 mg every 4 hours (Oxycod syrup; Rafa Pharmaceutics, Jerusalem, Israel) for pain relief and was evaluated daily. Despite the chemotherapy, there was no improvement in dental pain during a 2-week course. Consequently, because of the severity of the symptoms the decision was made to take a more radical treatment approach. Under local anesthesia (inferior alveolar nerve block with 3.6 mL of 2% lidocaine with 1:100,000 epinephrine), the dental pulp of the affected tooth was extirpated, and the root canal system was dressed with combined triamcinolone acetonide (1.0%) and demeclocycline calcium (3.021%) paste (Ledermix paste; Haupt Pharma GmbH, Wolfratshausen, Germany). 1590

This treatment resulted in immediate pain relief. It should be mentioned that despite a regional nerve block, additional intrapulpal anesthesia was needed as a result of pain during procedure. Similarly, in the second episode, which happened a month after the beginning of the rst episode, the pain characteristics were not specic and resembled those of the rst episode (Table 1). Periodontal probing depth up to 3 mm was measured around the tooth. Percussion (vertical/horizontal) and vestibular palpation tests yielded intense pain. Cold stimuli yielded intense pain; however, the pain receded on removal of the stimulus. The same rst line supportive treatment was recommended to the patient by using powerful analgesics. On follow-up 3 days later, no symptomatic improvement was reported. It was agreed to repeat the same ultimate dental treatment as in the rst episode. Pulp extirpation (after buccal and palatal inltration of 3.6 mL of 2% lidocaine with 1:100,000 epinephrine) resulted in immediate pain relief in this episode as well. In both episodes cytologic evaluation (by T.N.) was performed on pulp content that was collected by 19-gauge needle as well as from debris on the endodontic le. No lymphoma, inammatory process, or infection were seen (Fig. 3). Whole body positron emission tomography scan after injection of uorodeoxyglucose (18F), performed

Figure 3. Cytology of the extirpated pulp (rst episode); normal blood smear with no sign of malignancy or inammation. Original magnication, 630. (This gure is available in color online at www.aae.org/joe/.)

Zadik et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

between the rst and second pain episodes, revealed diffuse absorption in bone marrow and spleen (standard uptake value, 3.8) and anteriorsuperior part of mediastinum (standard uptake value, 2.6). No involvement of the head and neck region was detected. Endodontic treatments and rehabilitation of the extirpated teeth were not completed as a result of the patients poor medical condition. Eight months after the second episode the patient died as a result of decubitus-induced sepsis. therapy, extraction, or apical surgery) might result in temporary relief, but the pain characteristically returns in days or weeks and might migrate after the dental procedure (including midline crossing and involving both jaws) (22, 23). This similarity strengthens the diagnosis of neuropathic dental pain in the presented patient because atypical odontalgia mostly attributed to nerve deafferentation pain resulted from peripheral nerve injury (23). The anti-cancer agent most likely to have induced the pulp neuropathy of our patient is cyclophosphamide; in the rst episode, the pain appeared on the third day of cyclophosphamide administration, and in the second episode the pain appeared on the rst day of administration of cyclophosphamide (Fig. 1). Moreover, in the rst episode the other agents (cytarabine, doxorubicin, mesna, and rituximab) were administrated a week before the beginning of the pain, and only cyclophosphamide was administrated in proximity to the time of pain. Generally, cyclophosphamide and its isomer ifosfamide can induce central manifestations of encephalopathy and changes in mental status (6) and peripheral sensorimotor and autonomic neuropathy (24). However, the fact that cyclophosphamide is not a well-known sensory neuropathy causative is a limitation of this hypothesis. Of the 4 remaining agents (cytarabine, doxorubicin, mesna, and rituximab) the only agent that is known to have neurotoxicity as an adverse effect is cytarabine (cytosine arabinoside). This agent was reported to induce several central neurologic toxicities such as encephalopathy, seizures, cerebellar dysfunction, aseptic meningitis, intrathecal myelopathy (25). However, it is unlikely that cytarabine is the causative agent in this case because it was not given during the second pain episode. This case presents a circumstantial relation between the clinical presentation of dental pain and the timing of high-dose cyclophosphamide administrations. It suggests that cytotoxic therapy-induced neurotoxicity might manifest as a pulpal-type toothache that can present a diagnostic challenge for the dental practitioner. As mentioned above, the induction of peripheral neuropathy is a clinical problem because it impairs quality of life and is a common limiting factor of anti-cancer chemotherapy. Pulpal extirpation, although it irreversibly impairs dental pulp vitality and requires later endodontic treatment and rehabilitation of the tooth, provided immediate relief for the reported pain. More research is needed, however, for the establishment of this entity of chemotherapy-induced neuropathic dental pain and determination of its optimal management.

Discussion
In the medically competent patient, the differential diagnosis of acute severe dental pain includes irreversible pulpitis, acute periapical periodontitis or lateral periodontal abscess, and the related diagnosis of perio-endo lesion and vertical root fracture, as well as food impaction (9). All these potential diagnoses were ruled out in the 2 episodes by a comprehensive clinical-radiologic evaluation (10) as described above. Other conditions that might mimic severe dental pain include neurovascular orofacial pain, (pre-) trigeminal neuralgia, referred pain from maxillary sinus, ear, carotid artery, heart pathology, and pain caused by intracranial tumor (11). However, other manifestations needed for diagnosis, such as suitable systemic signs, ear, nose, throat, or chest symptoms, were not present. Moreover, in all of these conditions, although the pulp is the site of the referred pain, pulp extirpation has not relieved the pain, thus ruling out these conditions in our patient. Pain caused by central nervous system involvement of lymphoma was also ruled out in our patient because the pain was localized and affected bilateral dentition and because of the timing of pain in relation to the therapy. After the exclusion of all the above possible conditions, a diagnosis of inltration by lymphoma cells into the pulp chamber (1214) causing local pressure and pain was made. This assumption was ruled out retrospectively by histologic examination of the extirpated pulp tissue. After a comprehensive search of the literature, the possibility of neuropathic dental pain caused by chemotherapy was raised (15, 16). Neurotoxicity in the orofacial structures has been described as diffuse jaw pain or numbness. However, localized toothache as a manifestation of such neurotoxicity has not been extensively studied and is not listed as a possible orofacial complication of administration of cytotoxic agents (1, 2, 17, 18). Peripheral neurotoxicity and neuropathic pain including thermal (cold or heat) allodynia and hyperalgesia are well-known complications of several anti-cancer chemotherapeutic agents (19). Facial (oral) cold sensitivity is the most commonly reported form of the acute transient neuropathic symptoms of oxaliplatin-treated colorectal cancer patients, experienced while eating and drinking. This sensitivity has previously been described as causing symptoms of short duration that resolve and do not interfere with function and only rarely was debilitating because patients can adapt to avoid cold stimuli (20). The present patient, however, experienced spontaneous intense pain that was relieved only by pulpectomy. In their series of 34 breast cancer patients treated by chemotherapy, McCarthy and Skillings (8) mentioned that one patient undergoing breast cancer chemotherapy suffered from dental sensitivity and stated that this can be a complication of chemotherapy. In an additional series of vincristine-treated patients, 33% and 28% of the patients suffered from dental pain in their maxillary and mandibular teeth, respectively (21). The characteristics of the presented pain case were similar to those of the so-called atypical odontalgia, ie, a severe throbbing pain localized to teeth or supporting tissues with no clinical or radiographic pathologic ndings, dental procedures (most commonly endodontic

References
1. Benoliel R, Epstein J, Eliav E, Jurevic R, Elad S. Orofacial pain in cancer: part I mechanisms. J Dent Res 2007;86:491505. 2. Elad S, Epstein J, Klasser G, Sroussi H. Orofacial pain in the medically complex patient. In: Sharav Y, Benoliel R, eds. Orofacial pain and headache. Edinburgh: Mosby Elsevier; 2008:32148. 3. Benoliel R, Eliav E. Neuropathic orofacial pain. Oral Maxillofac Surg Clin North Am 2008;20:23754. 4. Clark GT, Saravanan R. Orofacial pain and neurosensory disorders and dysfunction in cancer patients. Dent Clin N Am 2008;52:183202. 5. Pasetto LM, DAndrea MR, Rossi E, Monfardini S. Oxaliplatin-related neurotoxicity: how and why? Crit Rev Oncol Hematol 2006;59:15968. 6. Sioka C, Kyritsis AP. Central and peripheral nervous system toxicity of common chemotherapeutic agents. Cancer Chemother Pharmacol 2009;63:7617. 7. Stengel M, Baron R. Oxaliplatin-induced painful neuropathy: icker of hope or hopeless pain? Pain 2009;144:2256. 8. McCarthy GM, Skillings JR. Orofacial complications of chemotherapy for breast cancer. Oral Surg Oral Med Oral Pathol 1992;74:1728. 9. Rossman L, Hasselgren G, Wolcott J. Diagnosis and management of orofacial dental pain emergencies. In: Cohen S, Hargreaves K, eds. Pathways of the pulp. 9th ed. St Louis, MO: Mosby Elsevier; 2006:4058. 10. Bender IB. Pulpal pain diagnosis: a review. J Endod 2000;26:1759. 11. Sharav Y, Benoliel R. Acute orofacial pain. In: Sharav Y, Benoliel R, eds. Orofacial pain and headache. Edinburgh: Mosby Elsevier; 2008:7590.

JOE Volume 36, Number 9, September 2010

Cytotoxic Chemotherapy-induced Odontalgia

1591

Case Report/Clinical Techniques

12. Biggs JT, SabalaZebra C XII. part Ilarge-cell lymphoma. J Endod 1992;18:5701. 13. Biggs JT, SabalaZebra C XII. part 2large-cell lymphoma. J Endod 1992;18:6324. 14. Payne M, al-Damouk JD. Gingival swelling as a manifestation of non-Hodgkins lymphoma. Br Dent J 1993;175:2934. 15. Peterson DE, Sonis ST. Oral complications of cancer chemotherapy: present status and future studies. Cancer Treat Rep 1982;66:12516. 16. Rosenberg SW. Oral complications of cancer chemotherapy: a review of 398 patients. J Oral Med 1986;41:937. 17. Epstein JB, Elad S, Eliav E, Jurevic R, Benoliel R. Orofacial pain in cancer: part II clinical perspectives and management. J Dent Res 2007;86:50618. 18. Fischer DJ, Klasser GD, Epstein JB. Cancer and orofacial pain. Oral Maxillofac Surg Clin North Am 2008;20:287301. 19. Binder A, Stengel M, Maag R, et al. Pain in oxaliplatin-induced neuropathy: sensitisation in the peripheral and central nociceptive system. Eur J Cancer 2007;43:265863. 20. Leonard GD, Wright MA, Quinn MG, et al. Survey of oxaliplatin-associated neurotoxicity using an interview-based questionnaire in patients with metastatic colorectal cancer. BMC Cancer 2005;5:116. 21. McCarthy GM, Skillings JR. Jaw and other orofacial pain in patients receiving vincristine for the treatment of cancer. Oral Surg Oral Med Oral Pathol 1992;74:299304. 22. Benoliel R, Heir GM, Eliav E. Neuropathic orofacial pain. In: Sharav Y, Benoliel R, eds. Orofacial pain and headache. Edinburgh: Mosby Elsevier; 2008:25594. 23. Blasberg B, Eliav E, Greenberg MS. Orofacial pain. In: Greenberg MS, Glick M, Ship JA, eds. Burkets oral medicine. 11th ed. Hamilton, Ontario, Canada: BC Decker Inc; 2008:25788. 24. JOE editorial board. Evidence-based review of clinical studies on pharmacology (non-anesthetic studies). J Endod 2009;35:11239. 25. Cavaliere R, Schiff D. Neurologic toxicities of cancer therapies. Curr Neurol Neurosci Rep 2006;6:21826.

1592

Zadik et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

Metastasis of Hepatocellular Carcinoma into the Mandible with Radiographic Findings Mimicking a Radicular Cyst: A Case Report
Hisako Fujihara, DDS, PhD,* Daichi Chikazu, DDS, PhD, Hideto Saijo, DDS, PhD, Hideyuki Suenaga, DDS, PhD, Yoshiyuki Mori, DDS, PhD, Mitsuyoshi Iino, DDS, PhD, Yoshiki Hamada, DDS, PhD,* and Tsuyoshi Takato, MD, PhD
Abstract
Introduction: Hepatocellular carcinoma (HCC) is a common neoplasm worldwide, with more than half of the tumors associated with regional metastasis. Extrahepatic metastasis is also common, and the most frequently affected sites are the lungs, abdominal lymph nodes, diaphragm, and bone. However, HCC metastasis to the mandible is rare, with approximately 50 cases reported in the literature. Methods: In this report, we describe a case of HCC metastasis to the mandible at the apex of #18 root in a 62-year-old man. This patient had already been diagnosed with metastasis to pancreatic caput lymph node. The radiographic features of the mandible resembled radicular cyst and did not show typical ndings of malignancy. Results: Under the rst diagnosis of radicular cyst, root canal treatment was initially performed, and then surgical treatment of the removal of the cystic lesion and #18 extraction were performed. Finally, the lesion was diagnosed as HCC metastasis from pathological examination. Consequently, he received constitutional chemotherapy in the hepatitis unit and is now in remission. Conclusion: This case shows the importance of considering the differential diagnosis of malignancy. (J Endod 2010;36:15931596)

epatocellular carcinoma (HCC) is the fth most common cancer. It is estimated that 8,500 to 11,500 new cases of HCC occur annually in the United States, with a relatively higher frequency reported in Southeast Asia including Japan (1, 2). Among all HCC patients, more than 25% of patients are reported to have extrahepatic metastasis, with 10.1% of patients showing bone metastasis (3). The preferred site is the vertebrae, followed by the ribs, sternum, and pelvis in decreasing order. The mandible is an uncommon site of extrahepatic metastasis of HCC. Here, we report a rare case of HCC metastasis to the mandible with characteristic radiographic ndings mimicking radicular cyst at the apex of the molar root.

Case Report
A 62-year-old Japanese man was referred to the Department of Oral-maxillofacial Surgery, Dentistry and Orthodontics, the University of Tokyo Hospital, with a chief complaint of slight swelling and discomfort of the left mandible. The symptoms were noticed 1 month previously. He did not suffer from any paralysis, pain, or dyskinesia. He had a history of HCC caused by hepatitis C infection from a transfusion of coagulation factor VIII blood product for hemophilia A. For the treatment of HCC, he had undergone transcatheter arterial embolization and radiofrequency ablation when he was 57 years old. He later underwent resection of the S8 liver area when he was 61 years old. The clinical examination showed a soft, painless slight swelling in the left mandibular angle region, with no trismus or neurologic problems such as mental nerve paresthesia or facial nerve paralysis. Intraorally, he had healthy dentition with good oral hygiene and a normal mucous membrane. There was slight swelling of the buccal side of the #18 gingiva. An old composite resin restoration was noted on the occlusal aspect of #18. The tooth responded negatively to electric pulp vitality test, and no discomfort or pain was noted on percussion. On the mesiobuccal aspect of the tooth, the probing depth measured 4 mm, whereas the mobility of the tooth was within normal limits. On panoramic x-ray examination, a radiolucent, well-dened lesion about 20 mm in diameter consistent with the appearance of a radicular cyst was observed at the apex of #18 (Fig. 1). Root absorption of #18 was unclear. On computed tomographic (CT) images, a smooth circular lesion was also observed in the same region. A slight resorption of the buccal cortical bone was observed on horizontal slices (Fig. 2A and B). The correlation between the apex of #18 and the circular lesion was not clear even on the CT images, and, unfortunately, dental x-ray examination was impossible because of his vomiting reex. There were also small circular radiolucent lesions in the right mandibular body in the #29 and #30 regions. On CT images, these several small circular lesions were also observed on the relatively lingual side of the mandible. Laboratory data revealed slight anemia (hemoglobin = 11.2 g/dL) and a decreased platelet count (8.9 104 /mm), but other blood cell counts were within normal limits. Coagulation factors were also within normal limits, and his coagulability was as follows: prothrombin time = 12.5 seconds and 71.7%, international normalized ratio = 1.42, activated partial thromboplastin time = 54.7 seconds, and brinogen = 239 units. Hepatic-related factors were as follows: lactate dehydrogenase = 264 IU/L, glutamic

Key words
Carcinoma, diagnosis, hepatocellular carcinoma, metastasis to mandible, radicular cyst

From the *Department of Oral and Maxillofacial Surgery, School of Dental Medicine, Tsurumi University, Kanagawa, Japan; and Department of Oral-maxillofacial Surgery, Dentistry and Orthodontics, The University of Tokyo Hospital, Tokyo, Japan. Address requests for reprints to Dr Hisako Fujihara, Department of Oral and Maxillofacial Surgery, School of Dental Medicine, Tsurumi University, 2-1-3, Tsurumi, Tsurumi-ku, Yokohama, Kanagawa 230-8501, Japan. E-mail address: fujihara-h@tsurumi-u.ac.jp. 0099-2399/$0 - see front matter Copyright 2010 American Association of Endodontists. doi:10.1016/j.joen.2010.05.009

JOE Volume 36, Number 9, September 2010

A Case of Metastasis of Hepatocellular Carcinoma into the Mandible

1593

Case Report/Clinical Techniques

also observed at the mandibular level (Fig. 2D). The apex of #18 showed accumulation, and stronger accumulation was observed in the #28 region. The standard uptake value (SUV) of the apex of #18 was 4.7, and the SUV of the #28 region was 7.9 or 23.7 with artifact. On comparison with pantomography, periodontitis and the metalcast prosthetic of #28 were considered to be the cause of the accumulation. The SUV of the caput pancreas lymph node was 8.0 and that of the liver was 5.4. These PET-CT images were consistent with other radiologic ndings. On the basis of these ndings, this lesion was diagnosed as a #18 radicular cyst with little possibility of malignant tumor or other pathology unrelated to the tooth. Because his systemic condition was poor, planned surgical treatment of #18 was amended to root canal treatment. To start the root canal treatment, a conventional endodontic access opening was made with a rubber dam placement. An access cavity was prepared, and the root canal was instrumented with stainless steel hand les until an apical stop of ISO #40 could be created. Persistent seepage of dark red, muddy exudate through the root canal diminished gradually with instrumentation. The root canal was irrigated frequently with 1.3% sodium hypochlorite followed by a nal rinse with 5 mL sterile saline. Subsequently, sterilized cotton with iodine was placed, and the access cavity was temporarily sealed with sterile cotton with Sandarac Vanish (G.C., Tokyo, Japan). At the next visit, the root canal was reentered and irrigated alternately with 1.3% sodium hypochlorite and sterile saline with a crown-down technique.

Figure 1. Radiologic ndings: pantomography at the rst referral. The red arrows indicate a radiolucent circular lesion at the apex of #18. There were also radiolucent lesions in the body of the right mandible below teeth #29 and #30.

oxaloacetic transaminase = 87 IU/L, glutamic pyruvic transaminase = 74 IU/L, gamma-glutamyl transpeptidase = 66 IU/L, and alpha-fetoprotein (ALP) = 346 IU/L. Tumor markers showed relatively high AFP and AFP-L3 (lectin 3) levels of 158.5 ng/mL and 41.1%, respectively. At his rst visit to our department, HCC metastasis to a caput pancreatic lymph node had already been diagnosed based on positron emission tomography-CT images (Fig. 2C), which was consistent with the high AFP level. On the PET-CT images, high accumulations were

Figure 2. CT and CT-PET images: (A) sagittal CT image. The red arrow indicates a lesion with a clear, smooth margin resembling a radicular cyst of #18. (B) A horizontal CT image. The red arrow indicates slight bone absorption of the outer side of the cortical bone. (C) A PET image of the abdomen. The black arrow indicates pancreatic head lymph node metastasis. This diagnosis had already been made before the rst referral. (D) A PET image of the oral region. The red spot in the right premolar region was suggested to be periodontitis of #28 by clinical symptoms. The red arrow points to yellow spot that shows less suggestion of malignancy compared with red spot. This lesion was nally diagnosed as HCC metastasis.

1594

Fujihara et al.

JOE Volume 36, Number 9, September 2010

Case Report/Clinical Techniques

During root canal treatment for 2 months, the swelling of the buccal mucosa of #18 decreased. However, it suddenly became prominent. Therefore, conservative root canal treatment was thought to be ineffective for this cystic lesion, so extraction of #18 was inevitable. At that time, his general condition had improved, and surgical removal of the cystic lesion concomitant with extraction of #18 was performed under general anesthesia with replenishment of coagulation factor VIII (Advate; Baxter, Deereld, IL). Surprisingly, the cystic lesion was composed of soft solid tissue suspected to be a tumorous lesion. The specimen was pathologically analyzed with hematoxylin-eosin staining, and immunostained for hepatocytes, glypican-3 (GPC-3), alphafetoprotein (AFP), and cytokeratin (CAM 5.2). Hematoxylin-eosing staining showed the lesion to have an alveolar or trabecular pattern. Eosinophilic granules in abundant cytoplasm, mild nuclear pleomorphism, prominent nucleoli, and numerous mitotic gures were observed in the thick-walled lesion (Fig. 3A and B). Additional immunohistochemical study revealed that the tumor cells were positive for hepatocytespecic antigen (Fig. 3C), GPC-3 (Fig. 3D), and CAM 5.2 (Fig. 3E) but negative for AFP (Fig. 3F). Considering these ndings, this specimen was diagnosed as HCC metastasis to the oral region. Immediately after the diagnosis was made, he was transferred to the hepatitis unit and underwent systemic chemotherapy with uorouracil (5-FU) and irradiation treatment. Since the completion of this treatment, he has been in remission for 6 months.

Discussion
HCC is the most common malignant tumor of the liver. The number of new HCC patients is approximately 500,000 to 1 million per year and is increasing (2). HCC usually undergoes intrahepatic metastasis in the early stages, and only 25% of HCC patients develop extrahepatic metastasis. The preferred sites are the lungs (34%-70%), regional lymph nodes (16%-45%), and bone (6% in vertebrae, ribs, and long bones) (4 9). However, HCC metastasis to the mandible is very rare. Since Dick et al (10) rst reported a case of HCC metastasis to the mandible in 1957, approximately 50 cases have been documented in the literature to our knowledge (1123). Patients ranged from 25 to 88 years, with a mean age of 63.9 years; 47 were male, and 2 were female. The mandible is the most commonly affected area in the oral and maxillofacial region. This age and sex predilection resembles that of primary HCC, which frequently occurs in men and takes several years to metastasize to the oral and maxillofacial region. Several pathways of HCC metastasis to the oral region have been postulated. One of them is hematogenous pathway; the tumor reaches the circulation through invasion of hepatic arterial and/or portal venous branches. Most oral metastases are associated with lung metastases, and they possibly occur by this route (24). Another possible pathway of metastasis is an anastomotic network of paravertebral veins that bypasses the pulmonary, inferior caval and portal venous circulations (25, 26). This pathway may be responsible for metastasis to the vertebral bodies, which are the preferred site of bony HCC metastasis. This could be the most likely pathway from HCC without pulmonary metastasis, as observed in our patient (27, 28). On the other hand, the mandibular angle and body have a rich blood supply, and blood ow slows down in this area, allowing deposition of metastatic cells (29). This is the main reason that the mandible is the most commonly involved region. Moreover, it could also be hypothesized that the pulp in tooth #18 became necrotic with a developing periapical lesion. The increased area of inammation would result in an increased blood ow with more permeable vessel walls, which may allow deposition of metastatic cells. Although the exact pathway of HCC metastasis to the apex of #18 in our patient is still unclear, the previously mentioned pathways would be reasonable because the apex of #18 is positioned near the mandibular angle and body. Radiographically, all reported cases of HCC metastasis to the mandible showed ndings suspicious of tumor, such as tumor mass formation in the bone (1014, 1719, 21, 23, 24), osteolytic lesions (14, 16, 20, 22), and ill-dened borders (3, 10, 11, 15). The present case was quite different from these reported patients in that (1) the site of metastasis was the apex of #18 and (2) the margin of the lesion was clearly dened in x-rays. Moreover, the SUV of the PET of the apex of #18 was only 4.7 although that of the pancreatic caput lymph node metastasis was 8.0. This result did not suggest malignancy of the #18 apex lesion because the SUV of malignant lesions ranges from 7.0 to 9.0. Consequently, this patient was not diagnosed as having metastatic HCC to the mandible, and malignancy was excluded initially. The relation between the #18 root and the lesion was still obscure; therefore, the lesion was not claried as being odontogenic or not. The rst period of root canal treatment seemed to be effective because swelling 1595

Figure 3. Histopathological ndings of the specimen. (A) Hematoxylin-eosin staining: the lesion was composed of atypical cells showing an alveolar or fascicular pattern, with eosinophilic granules in the cytoplasm. There are well-developed trabecular patterns and cells with abundant cytoplasm, mild nuclear pleomorphism, prominent nucleoli, and numerous mitotic gures. The bar indicates 50 mm. (B) A higher magnication of the same section. The bar indicates 10 mm. (C-F) Immunohistochemical reaction was positive for (C) hepatocyte, (D) GPC-3, and (E) CAM 5.2 and negative for (F) AFP. Bars from C to F indicate 50 mm.

JOE Volume 36, Number 9, September 2010

A Case of Metastasis of Hepatocellular Carcinoma into the Mandible

Case Report/Clinical Techniques

of the buccal mucosa decreased. Fortunately, his general condition had recovered when the lesion became prominent again, and extraction of #18 and removal of the lesion were performed. Finally, the pathological diagnosis of HCC metastasis to the mandible was made. There were also several small circular radiolucent lesions in the right mandible body in the #29 and #30 region. On CT images, they were observed on the lingual side in horizontal slices. Moreover, the patient had never had objective or subjective symptoms in the right mandible from the rst time of his referral. Before starting the root canal treatment for #18, these lesions were compared with the ndings of pantomography performed by his general practitioner 5 years before his treatment. Similar lesions in the right mandible were also observed in the old pantomograph, and their diameters and positions were almost identical; these lesions were diagnosed as static bone cysts or other benign lesions. There was no cystic lesion at the apex of #18. After the cystic lesion of #18 was diagnosed as HCC metastasis, we considered rediagnosing with a pathological examination of the right mandibular lesions because they might possibly have had the same diagnosis. However, we dared to postpone the pathological examination for several reasons: (1) the lesions had been there before the cystic lesion of #18 apex occurred; (2) they were deep below the buccal cortical bone, and biopsy would be more invasive and more difcult to coagulate than the removal of #18 cystic lesion and extraction; and (3) in the case that they were malignant, chemotherapy would be effective. Six months after chemoradiotherapy, the lesions of the right mandible had not increased and did not cause any symptoms. In fact, the denitive diagnosis of the right mandibular radiolucent lesions was still unclear. Patients with HCC have a relatively good survival rate, with a 5-year cumulative survival rate of more than 50% when surgical treatment is performed (30). However, when there is bone metastasis, the survival rate decreases sharply; 1-year survival is 15% to 20%, and 2-year survival is around 4% (3133). The present patient was still alive and in remission 6 months after the diagnosis of HCC metastasis to the mandible after effective chemotherapy in the hepatitis unit of our hospital. In conclusion, this report describes the unusual features of a case of oral metastatic HCC in which the lesion was located at the apex of a root with the appearance of a radicular cyst. This case shows that clinicians should consider the possibility of malignancy in the differential diagnosis, even if a malignant lesion is not suggested by the clinical ndings, especially in patients with a history of a malignant tumor.
6. Katyal S, Oliver JH, Peterson MS, et al. Extrahepatic metastases of hepatocellular carcinoma. Radiology 2000;216:698703. 7. Tunc B, Filik L, Tezer-Filik I, et al. Brain metastasis of hepatocellular carcinoma: a case report and review of the literature. World J Gastroenterol 2004;10:16889. 8. Chang JY, Ka WS, Chao TY, et al. Hepatocellular carcinoma with intra-atrial tumor thrombi. A report of three cases responsive to thalidomide treatment and literature review. Oncology 2004;67:3206. 9. Chang L, Chen YL, Kao MC. Intracranial metastasis of hepatocellular carcinoma: review of 45 cases. Surg Neurol 2004;62:1727. 10. Dick A, Mead SG, Mensh M, et al. Primary hepatoma with metastasis to the mandible. Am J Surg 1957;94:84650. 11. Pires FR, Sagarra R, Correa ME, et al. Oral metastasis of a hepatocellular carcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004;97:35968. 12. Ramon Ramirez J, Seoane J, Montero J, et al. Isolated gingival metastasis from hepatocellular carcinoma mimicking a pyogenic granuloma. J Clin Periodontol 2003;30: 9269. 13. Arai R, Otsuka T, Mori K, et al. Metastasis of hepatocellular carcinoma to the supramaxillary gingiva and right ventricle. Hepatogastroenterology 2004;51:115961. 14. Junquera L, Rodriguez RC, Torre A, et al. Hepatocellular carcinoma metastatic to the mandible: a case involving severe hemorrhage. Med Oral 2004;9:3459. 15. Niedzielska I, Langowska AH, Pajak J, et al. Mandible metastasis of hepatocellular carcinoma. Wiad Lek 2004;57:3924. 16. Chen SY, Cheng PW, Tsai CC. Pathology quiz case: mandibular metastasis of hepatocellular carcinoma. Arch Otolaryngol Head Neck Surg 2005;131:7356. 17. Teshigawara K, Kakizaki S, Sohara N, et al. Solitary mandibular metastasis as an initial manifestation of hepatocellular carcinoma. Acta Med Okayama 2006;60:2437. 18. Han L, Bhan R, Zak I, et al. Metastatic hepatocellular carcinoma to the mandible masquerading as a parotid gland mass: a potential pitfall in the diagnosis by ne needle aspiration biopsy. Diagn Cytopathol 2007;35:6746. 19. Huang SF, Wu RC, Chang JT, et al. Intractable bleeding from solitary mandibular metastasis of hepatocellular carcinoma. World J Gastroenterol 2007;13:45268. 20. Kamatani T, Tatemoto Y, Tateishi Y, et al. Isolated metastasis from hepatocellular carcinoma to the mandibular condyle with no evidence of any other metastases: a case report. Br J Oral Maxillofac Surg 2008;46:499501. 21. Li R, Walvekar RR, Nalesnik MA, et al. Unresectable hepatocellular carcinoma with a solitary metastasis to the mandible. Am Surg 2008;74:3469. 22. Piot B, Adam P, Sagan C, et al. Mandibular metastasis of hepatocellular carcinoma. Apropos of 3 cases. Rev Stomatol Chir Maxillofac 1994;95:4314. 23. Zanirato S, Porta C, Moroni M, et al. Jaw bone metastasis from hepatocellular carcinoma. Ital J Gastroenterol Hepatol 1997;29:4789. 24. Takinami S, Yahata H, Kanoshima A, et al. Hepatocellular carcinoma metastatic to the mandible. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1995;79:64954. 25. Batson OV. The vertebral system of veins as a means for cancer dissemination. Prog Clin Cancer 1967;3:118. 26. Batson OV. The function of the vertebral veins and their role in the spread of metastases. 1940. Clin Orthop Relat Res 1995;312:49. 27. Hirshberg A, Leibovich P, Buchner A. Metastases to the oral mucosa: analysis of 157 cases. J Oral Pathol Med 1993;22:38590. 28. Hirshberg A, Shnaiderman-Shapiro A, Kaplan I, et al. Metastatic tumours to the oral cavity - pathogenesis and analysis of 673 cases. Oral Oncol 2008;44:74352. 29. Romanas MM, Cherian R, McGregor DH, et al. Hepatocellular carcinoma diagnosed by ne-needle aspiration of the parotid gland. Diagn Cytopathol 2004;30:4015. 30. Schwarz RE, Smith DD. Trends in local therapy for hepatocellular carcinoma and survival outcomes in the US population. Am J Surg 2008;195:82936. 31. Seong J, Koom WS, Park HC. Radiotherapy for painful bone metastases from hepatocellular carcinoma. Liver Int 2005;25:2615. 32. Taki Y, Yamaoka Y, Takayasu T, et al. Bone metastases of hepatocellular carcinoma after liver resection. J Surg Oncol 1992;50:128. 33. Kaizu T, Karasawa K, Tanaka Y, et al. Radiotherapy for osseous metastases from hepatocellular carcinoma: a retrospective study of 57 patients. Am J Gastroenterol 1998;93:216771.

References
1. Kuhlman JE, Fishman EK, Leichner PK, et al. Skeletal metastases from hepatoma: frequency, distribution, and radiographic features. Radiology 1986;160:1758. 2. Anthony PP. Hepatocellular carcinoma: an overview. Histopathology 2001;39:10918. 3. Yoshimura Y, Matsuda S, Naitoh S. Hepatocellular carcinoma metastatic to the mandibular ramus and condyle: report of a case and review of the literature. J Oral Maxillofac Surg 1997;55:297306. 4. Natsuizaka M, Omura T, Akaike T, et al. Clinical features of hepatocellular carcinoma with extrahepatic metastases. J Gastroenterol Hepatol 2005;20:17817. 5. Cha C, Fong Y, Jarnagin WR, et al. Predictors and patterns of recurrence after resection of hepatocellular carcinoma. J Am Coll Surg 2003;197:7538.

1596

Fujihara et al.

JOE Volume 36, Number 9, September 2010