Clinical Research
An Analysis of the Proximity of Maxillary Posterior
Teeth to the Maxillary Sinus Using Cone-beam
Computed Tomography
Xiao-mei Tian, BDS,* Liang Qian, BDS,* Xian-zhen Xin, DDS, MDS,* Bin Wei, DDS, PhD,
and Yao Gong, DDS, MDS
Abstract
Introduction: It is known that the level of the maxillary
sinus floor varies with age. Because few studies have
investigated whether the position of the posterior roots
relative to the maxillary sinus varied with the variation
of the sinus floor or not, the present study assessed
the position according to age. Methods: Cone-beam
computed tomographic images of 848 patients were re-
constructed to evaluate the position of the posterior
roots relative to the sinus floor, which were divided
into 3 types, and quantify the distances between poste-
rior root apexes and the adjacent border of the sinus
floor. Measurements were taken for each root, and
data were correlated with age, which was divided into
4 groups (#20 years, 2140 years, 4160 years, and
>60 years). Results: A total of 3063 premolars and
3095 molars were evaluated. The mean distances from
the root tips to the border of the maxillary sinus floor
increased with increasing age. From the first and second
premolar roots, the mean distances ranged from
3.6  4.1 mm to 8.9  4.6 mm and from
0.7  3.3 mm to 5.3  3.9 mm, respectively. From
the mesiobuccal, distobuccal, and palatinal roots of
the first molars, the mean distances were
0.4  2.8 mm to 4.6  4.0 mm, 0.3  2.4 mm to
4.4  3.8 mm, and 0.4  3.5 mm to 3.9  4.1 mm,
respectively. From the mesiobuccal, distobuccal, and
palatinal roots of the second molars, the mean
distances were 0.5  2.2 mm to 3.4  3.5 mm,
0.3  2.2 mm to 3.9  3.7 mm, and 1.1  3.2 mm to
4.6  4.5 mm, respectively. The frequency of type IS
decreased with increasing age. It was very low in first
premolars (0%8.13%) and higher in second premolars
(0%25.68%), first molars (0%44.75%), and
second molars (0%32.89%). Age significantly influ-
enced the mean distances to the sinus floor and the fre-
quencies of type IS (inside). Conclusions: Cone-beam
computed tomographic imaging is an effective method
From the *Department of Prosthodontics, Stomatology Special Consultation Clinic, and Department of Orthodontics, Ninth Peoples Hospital, Shanghai Jiao Tong
University School of Medicine, Shanghai Key Laboratory of Stomatology, Shanghai, China
Address requests for reprints to Dr Yao Gong, No. 500, Quxi Road, Huangpu District, Shanghai, China. E-mail address: gongyao0328@hotmail.com or Dr Bin Wei,
No. 639, Zhizaoju Road, Huangpu District, Shanghai, China. E-mail address: weibin0328@hotmail.com
0099-2399/$ - see front matter
Copyright 2016 American Association of Endodontists.
http://dx.doi.org/10.1016/j.joen.2015.10.017
JOE Volume 42, Number 3, March 2016
to study the position of the posterior roots to the maxillary sinus floor. Variation in
proximity measurements was found by age, with those under the age of 40
showing a greater likelihood of the position of maxillary roots above/inside the sinus
floor. (J Endod 2016;42:371377)
Key Words
Cone-beam computed tomography, maxillary premolar, maxillary sinus, molar
R ecognition of the anatomic position of the teeth relative to the maxillary sinus, which
is a constant challenge in dentistry, is essential to increase successful nonsurgical
endodontic diagnosis, treatment, and extraction. Close proximity between the root apex
and the adjacent maxillary sinus floor is significant during a clinical operation because it
may result in various complications, such as sinusitis (1, 2), penetration, oroantral
fistulae, endoantral syndrome, or root displacement into the maxillary sinus (35).
However, in periapical surgery of surgical endodontics, although the close
relationship does not affect its prognosis even when there is an exposure into the
sinus during surgery (6, 7), it will establish a channel between the sinus and the
infected periapical tissue and then may cause acute or chronic sinusitis (8). Meanwhile,
this close relationship is also relevant during implant operations. Wehrbein and
Diedrich (9) and Sharan and Madjar (10) found the extraction of the roots protruding
into the sinus could increase the associated risk of pneumatization using panoramic
radiography. Sinus expansion after the loss of maxillary posterior roots can greatly
decrease the bone height available for implant placement and occasionally reaches
the alveolar ridge (3).
In general, the reliability of panoramic radiography in assessing the position be-
tween the maxillary sinus and the posterior roots is doubtful because of superposition
and magnification of the anatomic structure (11). Therefore, recent studies focused on
computed tomographic (CT) and cone-beam computed tomographic (CBCT) imaging.
As a nondestructive tool, CBCT imaging, which provides 3-dimensional images of the
oral and maxillofacial region for disease diagnoses as well as the anatomic position,
not only has a lower effective radiation dose, higher spatial resolution, and cheaper
price compared with CT imaging but also can accurately interpret the relationship of
the maxillary roots relative to the maxillary sinus floor.
It is known that the level of the maxillary sinus floor varies widely with age (12),
which is level with the nasal floor at about 12 years old and reaches the lowest point at
approximately 20 years of age with the eruption of the third molars (10, 13). The
investigators hypothesized that the position of the posterior roots with respect to the
Proximity of Maxillary Posterior Teeth to Maxillary Sinus 371
Clinical Research
maxillary sinus may also constantly vary according to age. However, few
studies revealed the anatomic position changed according to age
accurately. Thus, the primary objective of the present study was to
analyze the vertical relationship between the posterior roots and the
adjacent maxillary sinus floor in a large sample of the Chinese
population using reconstructed CBCT images. The secondary
objective was to correlate the data with age.
Materials and Methods
Subjects
The study materials were composed of dental CBCT
images retrospectively selected from the CBCT database at the
department of Stomatology Special Consultation Clinic, Ninth Peoples
Figure 1. (B and C) Type IS and (A and D) CO were shown (arrows) in the sagittal and coronal planes.
372 Tian et al.
Hospital, Shanghai Jiao Tong University School of Medicine,
Shanghai, China, from January 2013 to December 2013. Because
of the retrospective nature, the study was exempt from
approval by an institutional review board. Subjects were selected, and
a total of 848 patients CBCT images fulfilled the following inclusion
criteria.
Inclusion Criteria
Cases presenting with the following findings were included for
further analysis:
1. Presence of maxillary permanent premolar or molar teeth on scans
2. Fully erupted teeth and fully formed apexes
JOE Volume 42, Number 3, March 2016

3. Maxillary premolar and molar with neither definitive root resorption
nor bony destruction around the teeth
4. Maxillary sinus floor that was complete and not damaged by disease
Exclusion Criteria
Cases presenting with the following findings were excluded:
1. The presence of periapical or periradicular lesions
2. Position of maxillary sinus floor after sinus floor evaluation
3. Teeth had previously undergone orthodontic treatment
4. Images with artifact caused by motion, beam hardening or scatter,
and so on
Radiographic Evaluations
CBCT images were obtained using a CBCT scanner (NewTom VG;
QR srl, Verona, Italy) with a field of view of 500 cm2 (20  25 cm) and a
basic voxel size of 0.16 mm. Operating parameters were set at 110 kV
and 10 mA with a scanning time of 18 seconds and acquired by an
experienced radiologist based on operation instructions. The following
analyses and measurements were performed by NNT software, version
2.21 (ImageWorks, Elmsford, NY). Reformatted images were magnified
by 180% and analyzed with a slice thickness of 0.25 mm to make a clear
radiographic evaluation of the maxillary teeth roots relative to the
adjacent maxillary sinus floor. The distances between posterior roots
and the adjacent sinus floor were measured using tools built into the
NNT software.
To address the research purpose, the vertical relationship between
posterior roots and the maxillary sinus floor was classified into 3
categories (Fig. 1):
1. Type IS: The root tips extending above/inside the maxillary sinus
floor
2. Type CO: The root contacting with the maxillary sinus floor
3. Type OS: The root extending below/outside the sinus floor
Figure 2. Measurement of the distance between the root apex and the maxillary sinus floor in sagittal (A) and coronal (B) planes. A line drawn at the approximate
base of the root intruding into the maxillary sinus is established, and the length is measured using a line from the root apex perpendicular to the baseline.
JOE Volume 42, Number 3, March 2016
Clinical Research
The relationship and distance were evaluated for each side of each
root of the molars. However, for the premolars, only 1 root was
measured, which was close to the sinus floor because the buccal and
palatal roots of most premolars were fused. The methods used to
analyze and measure were as follows:
1. The vertical relationship between the root of the maxillary premolar
and molar and the maxillary sinus floor was comprehensively
evaluated in sagittal (Fig. 1A and B) and coronal (Fig. 1C and D)
CBCT planes simultaneously and recorded. The priority order of
the 3 types was type IS, type CO, and type OS (ie, in sagittal planes,
the relationship was type IS, whereas in coronal planes the
relationship was type CO; finally, type IS was recorded).
2. The shortest vertical/oblique distance of the root apex of the
premolar and molar to the closest border of the maxillary sinus
was measured in sagittal (Fig. 2A) and coronal (Fig. 2B) CBCT
planes simultaneously (negative value if the root was inside the
sinus floor), and then the values that were smaller in these 2
measurements for the same root were recorded.
All data were evaluated again at an interval of 1 month; no
statistical difference was found among replicate measurements using
the t test. In addition, the age of each patient was recorded to correlate
these data. Age was divided into 4 groups: #20 years, 2140 years,
4160 years, and >60 years.
Statistics
Statistical analysis was performed by SAS 9.2 (SAS Institute Inc,
Cary, NC). Logistic regression was used to compare the difference of
the frequency of root relationship with respect to the maxillary sinus
floor between the types of root and the difference of the frequency of
the root above/inside the sinus floor according to age group. The
mixed-effects model was used to analyze the relationship of the mean
distances to the maxillary sinus floor with the age group. The inner
correlation within subjects was considered in both models. P < .05
was considered statistically significant.
Proximity of Maxillary Posterior Teeth to Maxillary Sinus 373
Clinical Research
Results
A total of 3063 maxillary premolars and 3095 molars were
assessed in 848 patients (females: 64.27%, males: 35.73%) who had
a mean age of 34.04  15.13 years (range, 1481 years). Among
the evaluated premolars, 1525 (49.78%) were first premolars, and
1538 (50.22%) were second premolars. Among the evaluated molars,
1564 (50.53%) were first molars, and 1531 (49.47%) were
second molars.
Mean Distance Analysis
The mean distances from the first premolars and second
premolars to the maxillary sinus floor were 5.5  4.8 mm and
2.3  4.0 mm, respectively (Table 1) (P < .0001). The values
for the mesiobuccal (MB), distobuccal (DB), and palatinal (P) roots
of the first molars were 1.3  3.6 mm, 1.4  3.4 mm, and
1.0  3.9 mm, respectively (P > .05 for MB vs DB and
P < .0001 for MB vs P and DB vs P). The values for the MB, DB
,and P roots of the second molars were 0.5  2.9 mm,
1.3  3.1 mm, and 1.9  3.5 mm, respectively (P < .0001 for
MB vs DB, MB vs P, and DB vs P). When comparing the first with
the second molars, significant differences were found in the MB
(P < .0001), DB (P < .05), and P roots (P < .0001).
Frequency Analysis
The majority of the first premolars (83%) were located below the
border of the maxillary sinus floor (Table 2). The corresponding per-
centage was lower in second premolars (56.6%). Significant differ-
ences were found in comparing the relationship of first premolars
versus second premolars with respect to the border of the sinus floor
(P < .0001). Analysis of type OS first molars (38%44.5%) occurred
more frequently. However, except for the mesiobuccal roots in which
type CO dominated (43.2%), type OS occurred more frequently in other
roots of maxillary second molars (43.6% for DB roots and 51.6% for P
roots). Significant differences were found when comparing the relation-
ship of MB roots versus P roots and DB roots versus P roots of first mo-
lars relative to the border of the sinus floor (P < .0001) but not found in
MB roots versus DB roots (P > .05). For the second molars, significant
differences were found when comparing the relationship of MB, DB,
and P roots relative to the border of the sinus floor (P < .0001).
When comparing first molars with second molars, significant
differences were found in the MB, DB, and P roots (P < .05).
Secondary Objectives
According to analysis of the mean distances of first and second
premolars and molars roots according to age (Table 1 and Fig. 3A),

TABLE 1. Measurements between the Maxillary Posterior Teeth and the Adjacent Maxillary Sinus Floor (mm) According to Age
#20 y 2140 y 4160 y >60 y Total
1PM
n 332 704 398 90 1524
Mean (SD) 3.6 (4.1) 4.9 (4.3) 7.6 (4.9) 8.9 (4.6) 5.5 (4.8)
Median 2.9 4.4 7.2 8.7 5.0
Min to Max 6.0 to 19.9 4.0 to 23.2 2.3 to 21.7 0.8 to 19.2 6.0 to 23.2
2PM
n 331 730 389 88 1538
Mean (SD) 0.7 (3.3) 1.8 (3.6) 4.0 (4.1) 5.3 (3.9) 2.3 (4.0)
Median 0 0.8 3.1 4.9 1.3
Min to Max 7.3 to 17.1 6.0 to 16.5 3.8 to 16.4 0.0 to 14.3 7.3 to 17.1
1MO MB
n 362 767 350 84 1563
Mean (SD) 0.4 (2.8) 0.9 (3.0) 3.4 (4.1) 4.6 (4.0) 1.3 (3.6)
Median 0 0 2 3.8 0
Min to Max 8.0 to 15.9 5.8 to 17.3 4.8 to 17.3 0.0 to 17.2 8.0 to 17.3
1MO DB
n 361 750 349 82 1542
Mean (SD) 0.3 (2.4) 0.9 (2.9) 3.4 (4.0) 4.4 (3.8) 1.4 (3.4)
Median 0 0 2 3.4 0
Min to Max 5.5 to 10.5 6.0 to 16.0 3.0 to 18.2 1.0 to 13.9 6.0 to 18.2
1MO P
n 362 764 349 82 1557
Mean (SD) 0.4 (3.5) 0.5 (3.4) 2.7 (4.3) 3.9 (4.1) 1.0 (3.9)
Median 0 0 1.3 3.1 0
Min to Max 8.0 to 14.3 12.3 to 12.6 5.3 to 17.8 2.3 to 14.9 12.3 to 17.8
2MO MB
N 304 764 379 83 1530
Mean (SD) 0.5 (2.2) -0.1 (2.5) 1.9 (3.2) 3.4 (3.5) 0.5 (2.9)
Median 0 0 1 2.5 0
Min to Max 6.0 to 7.0 7.0 to 11.8 6.5 to 15.5 0.0 to 14.0 7.0 to 15.5
2MO DB
n 263 613 330 62 1268
Mean (SD) 0.3 (2.2) 0.6 (2.6) 2.9 (3.6) 3.9 (3.7) 1.3 (3.1)
Median 0 0 2 3 0
Min to Max 6.0 to 9.3 5.8 to 15.1 3.5 to 19.0 0.0 to 16.1 6.0 to 19.0
2MO P
n 290 696 359 75 1420
Mean (SD) 1.1 (3.2) 1.2 (3.0) 3.4 (3.8) 4.6 (4.5) 1.9 (3.5)
Median 0 0 2.3 3.4 0.9
Min to Max 7.3 to 12.9 6.5 to 14.5 4.0 to 18.6 3.3 to 19.0 7.3 to 19.0
1PM, first premolar; 1MO DB, first molar distobuccal; 1MO MB, first molar mesiobuccal; 1MO P, first molar palatinal; 2MO DB, second molar distobuccal; 2MO MB, second molar mesiobuccal; 2MO P, second
molar palatinal; 2PM, second premolar; SD, standard deviation.

374 Tian et al. JOE Volume 42, Number 3, March 2016

 Clinical Research
TABLE 2. Frequency of Relationship between Posterior Roots and the 1. The mean distance of the first premolar was greater than that of the
Maxillary Sinus Floor second premolar and first and second molars even though we took
No. of roots and frequency (%) age into consideration. The greatest difference of all data sets
accounted to 5.7 mm when comparing the maximum measurements
Type of root Type IS Type CO Type OS Total (1PM) and the minimum measurements (2MO MB) in the third age
1PM 46 (3.02) 213 (13.97) 1266 (83.02) 1525 group (7.6 mm vs 1.9 mm), and the smallest difference of all data
2PM 199 (12.97) 470 (30.56) 869 (56.50) 1538 sets accounted to 4.1 mm comparing the maximum measurements
1MO MB 346 (22.14) 558 (35.70) 659 (42.16) 1563
1MO DB 326 (21.14) 529 (34.31) 687 (44.55) 1542
(1PM) and the minimum measurements (2MO MB) in the first age
1MO P 449 (28.84) 515 (33.08) 593 (38.09) 1557 group (3.6 mm vs 0.5 mm).
2MO MB 376 (24.58) 659 (43.07) 495 (32.35) 1530 2. In the P root of the first molar, the frequency of type IS was greater
2MO DB 207 (16.32) 506 (39.91) 555 (43.77) 1268 than that of other roots of premolars and molars even though we
2MO P 172 (12.11) 518 (36.48) 730 (51.41) 1420 took age into consideration (the greatest difference reached
1PM, first premolar; 1MO DB, first molar distobuccal; 1MO MB, first molar mesiobuccal; 1MO P, first 36.62% between 1MO P and 1PM in the first age group [44.75%
molar palatinal; 2MO DB, second molar distobuccal; 2MO MB, second molar mesiobuccal; 2MO P, vs 8.13%]). However, in all roots, the distance of the
second molar palatinal; 2PM, second premolar. second molar MB root was always shortest even though we took
age into account. This might be explained by the fact that the fre-
quency of type OS in the MB root of the second molars was the least.
a consistent pattern (ie, the minimum distances from root tips to the 3. Measurements between root tips and the sinus floor increased with
adjacent border of the sinus floor increased with increasing age) was increasing age. The greatest difference of all data sets accounted to
observed. With respect to analysis of the frequency of type IS of all roots 5.3 mm when comparing the first and fourth age group of first
according to age (Table 3 and Fig. 3B), it showed that the frequency of premolars (3.6 mm vs 8.9 mm). The smallest difference of all
type IS decreased with increasing age. Meanwhile, the differences in the data sets accounted to 3.5 mm when comparing the first and fourth
distances and the frequencies of type IS among each age group were age group of P roots of second molars (1.1 mm vs 4.6 mm) (Fig. 4).
statistically significant (P < .0001). Furthermore, when comparing 4. The frequency of type IS decreased with increasing age. The greatest
the first and second premolars according to age, significant differences difference reached 43.37% between the first and fourth age group of
in the distances and the frequency of type IS were found (P < .05). MB roots of first molars (43.37% vs 0%).
When analyzing the first molars, significant differences were found in
MB versus P roots and DB roots versus P roots but not in MB versus Previous radiographic studies often used panoramic radiographs,
DB roots (P > .05). When analyzing the second molars, significant dif- CT scanning, or CBCT imaging to assess the position of roots with
ferences were found in MB versus DB roots, MB versus P roots, and DB respect to the maxillary sinus floor. However, Sharan and Madjar
versus P roots (P < .05). When comparing the first and second molars, (11) compared panoramic radiographs with CT scanning of 80 subjects
significant differences were found in MB, DB, and P roots (P < .05). regarding root position with respect to the sinus floor. Only 39% of
roots that projected on the sinus cavity in panoramic radiographs
showed protrusion into the sinus with CT imaging. The mean projection
Discussion length in panoramic radiographs was 2.1 times longer compared with
It is known that the level of the border of the maxillary sinus floor CT scanning (P < .001). Howe (14) compared CBCT imaging with gross
varied with age (10, 12, 13). However, whether the position of maxillary dissection regarding measuring the dimensions of the maxillary bone of
posterior roots with respect to the sinus floor varied with the variation of 37 cadaver specimens; he found a Pearson correlation coefficient of
the maxillary sinus floor or not had not attracted attention. Therefore, r = 0.85 with a positive bias for CBCT images. Therefore, CBCT images
the present study aimed at investigating the relationship and measuring in metric measurements between posterior roots and the adjacent sinus
the distances between the posterior roots and the maxillary sinus floor floor are considered accurately.
by using reconstructed CBCT images and correlated the data with age. Regarding the mean distances of posterior roots to the maxillary
Analyzing the data regarding the relationship and distance between sinus floor, only 2 of 4 studies used CBCT images to measure the
the roots of premolars and molars and the maxillary sinus floor, the distance between the roots of the premolars and molars and the sinus
following statements can be made: floor (15, 16), and the other 2 studies only measured the distance of the

Figure 3. Trend charts of distances (A) and frequencies of type IS (B) according to age.

JOE Volume 42, Number 3, March 2016 Proximity of Maxillary Posterior Teeth to Maxillary Sinus 375
Clinical Research
TABLE 3. Frequency of Type IS According to Age
#20 y
1PM 27/332 (8.13)
2PM 85/331 (25.68)
1MO MB 157/362 (43.37)
1MO DB 147/361(40.72)
1MO P 162/362 (44.75)
2MO MB 100/304 (32.89)
2MO DB 55/263 (20.91)
2MO P 52/290 (17.93)
1PM, first premolar; 1MO DB, first molar distobuccal; 1MO MB, first molar mesiobuccal; 1MO P, first molar palatinal; 2MO DB, second molar distobuccal; 2MO MB, second molar mesiobuccal; 2MO P, second
molar palatinal; 2PM, second premolar.
premolars or molars to the sinus floor (17, 18). Kilic et al (15) used
coronal CBCT planes and reported mean distances of 8.42 mm
(right side) and 6.58 mm (left side) for first premolars, 3.75 mm (right
side) and 3.73 mm (left side) for second premolars, 0.7 mm to
1.86 mm (right side) and 0.1 mm to 0.52 mm (left side) for first molars
and 0.25 mm to 1.06 mm (right side) and 0.31 mm to 0.78 mm (left
side) for second molars. Georgescu et al (16) used sagittal CBCT planes
and reported mean distances of 7.56 mm for first premolars, 4.64 mm
for second premolars, 3 mm to 3.08 mm for first molars, and 2.76 mm
to 3.22 mm for second molars. However, von Arx et al (18) used
sagittal, coronal, and axial CBCT planes to measure the distances of pre-
molars to the maxillary sinus and found the mean distances were
different in the 3 planes. Therefore, to measure the minimum vertical
distances of posterior maxillary roots to the adjacent border of the sinus
floor in the present study, the distances of sagittal and coronal CBCT
planes, in which it was regarded as enough to obtain 3-dimensional
data of the vertical distances, were measured and the minimum values
taken. The mean distances were 5.5  4.8 mm in first premolars;
2.3  4.0 mm in second premolars; 1.3  3.6 mm, 1.4  3.4 mm,
and 1.0  3.9 mm in MB, DB, and P roots of first molars; and
0.5  2.9 mm, 1.3  3.1 mm, and 1.9  3.5 mm in MB, DB, and
P roots of second molars, respectively. These findings are in accordance
with the results reported by authors (16, 19) who provided evidence
that the MB root of the second molar was the closest one to the sinus
floor according to mean values and different from the results that the
DB root of the second molar was the closest one to the sinus floor
(15, 20).
In the present study, the frequencies of posterior roots inside
the sinus cavity were 3.02% and 12.97% in first and second
premolars; 22.14%, 21.14%, and 28.84% in the MB, DB, and P roots
of first molars; and 24.58%, 16.32%, and 12.11% in the MB, DB,
and P roots of second molars, respectively. These data could explain
Figure 4. Distance difference between the first and fourth age group.
376 Tian et al.
Frequency (%)
2140 y 4160 y >60 y
18/705 (2.55) 1/398 (0.25) 0/90 (0.00)
104/730 (14.25) 10/389 (2.57) 0/88 (0.00)
174/767 (22.69) 15/350 (4.29) 0/84 (0.00)
168/750 (22.40) 10/349 (2.87) 1/82 (1.22)
248/764 (32.46) 35/349 (10.03) 4/82 (4.88)
249/764 (32.59) 27/379 (7.12) 0/83 (0.00)
134/613 (21.86) 18/330 (5.45) 0/62 (0.00)
108/696 (15.52) 11/359 (3.06) 1/75 (1.33)
why the DB root of the second molar is not the closest one to the
sinus floor. However, our results were different from the previous
findings of Israel, Turkey, Korea, and Brazilian populations (11,
15, 17, 2022) in which the lowest figures were reported for root
protrusions of first (0%) and second premolars (0.9%) and the
MB (3.2%, 12.9%), DB (1.8%, 8.3%), and P roots (5.5%, 4.1%)
of first and second molars (22), and the highest figures were
reported for root protrusions of first (7.2%) and second premolars
(13.6%) (18) and the MB (32.5%, 36.7%), DB (30.1%, 34.3%),
and P roots (34.2%, 30.9%) of first and second molars (17, 21).
Therefore, the investigators suspected that the relationship between
maxillary roots and the sinus floor might have ethnic differences
just as the root canals that are genetically determined and have
definitive importance in anthropology (23, 24), and this aspect is
worth further research.
Regarding age, in the present study, the mean distances from
first and second premolar roots to the maxillary sinus floor
ranged from 3.6  4.1 mm to 8.9  4.6 mm and from
0.7  3.3 mm to 5.3  3.9 mm, respectively. From the MB, DB, and
P roots of the first molars, the mean distances were 0.4  2.8 mm
to 4.6  4.0 mm, 0.3  2.4 mm to 4.4  3.8 mm, and
0.4  3.5 mm to 3.9  4.1 mm, respectively. From the MB, DB,
and P roots of the second molars, the mean distances were
0.5  2.2 mm to 3.4  3.5 mm, 0.3  2.2 mm to 3.9  3.7 mm,
and 1.1  3.2 mm to 4.6  4.5 mm, respectively. We found that the
mean distances of all roots of maxillary premolars and molars to the
adjacent border of the sinus floor increased with increasing age, and
in all age groups, the MB root of the second molar was always the closest
one to the sinus (Fig. 3A). The variation tendencies of the mean
distances of all posterior roots are in accordance with the variation
trend of the level of the maxillary sinus floor. Therefore, we guess the
level of the maxillary sinus floor, which is in the position of all posterior
roots, changes downwardly or upwardly simultaneously with age, and it
can influence the mean distances of posterior roots to the sinus floor in
different age groups. Meanwhile, the frequency of type IS of premolars
and molars roots decreased with increasing age (Table 3 and Fig. 3B).
Type IS occurred mainly in the first and second age group and rarely in
the fourth age group; this reminds clinicians that it is comparatively
dangerous when performing surgical treatment for maxillary posterior
teeth before 40 years old and relatively safe after 60 years old.
Comparing the mean distance and the frequency of type IS with
the secondary study parameters, significant correlations were found.
In all premolar and molar roots, significant differences were found
between the different age groups (P < .0001), which implies that the
endodontic treatment of premolar and molar roots should be
distinguished in different age groups. When comparing the mean dis-
tances and frequencies of type IS between the 2 roots, except for the
MB versus DB root of first molars (P > .05), other comparisons had
JOE Volume 42, Number 3, March 2016

significant differences, which reminds clinicians to discriminate the
position relative to the maxillary sinus floor between different roots
during either root canal treatment or extraction. Furthermore, except
for second premolars, the distance difference between the first and
fourth age group decreased from the first premolar to the
second molar palatinal root (Fig. 4). The investigators suspected
that the morphology of the maxillary sinus floor is different between
the first and fourth age group, and this aspect should be researched
further.
However, although the classification method is the same, partial
results still differ from the conclusions found in other studies. This
may be caused by ethnic differences and differences in the measuring
methods used. Thus, the Chinese population should have its own
research data to guide clinical operation; our research partially fulfilled
that. Because of the limitations of our experiment, more factors and
more data are needed to more explicitly explain this problem.
Conclusions
This study assessed the proximity between the border of the
maxillary sinus floor and the maxillary posterior roots using dental
CBCT images. It can be summarized as follows:
1. The first premolar was always farther and the second molar
mesiobuccal root was closer to the border of the maxillary sinus
floor.
2. The root protruding into the sinus was rare in the first premolar and
dominated in the first molar palatinal root.
3. The root was closer to the border of the maxillary sinus floor before
the age of 20 and farther after the age of 60.
4. Age significantly influenced the mean distances and the frequency of
the root above the maxillary sinus floor.
Acknowledgments
Xiao-mei Tian and Liang Qian contributed equally to this
work and should be considered cofirst authors.
Supported in part by a research grant from 973 Program Grant
2012 CB910401 and Science and Technology Committee of
Shanghai (grant no. 13140902701).
References
1. Lu Y, Liu Z, Zhang L, et al. Associations between maxillary sinus mucosal thickening
and apical periodontitis using cone-beam computed tomography scanning: a
retrospective study. J Endod 2012;38:106974.
2. Nurbakhsh B, Friedman S, Kulkarni GV, et al. Resolution of maxillary sinus
mucositis after endodontic treatment of maxillary teeth with apical periodontitis:
a cone-beam computed tomography pilot study. J Endod 2011;37:150411.
JOE Volume 42, Number 3, March 2016
Clinical Research
3. Hauman CH, Chandler NP, Tong DC. Endodontic implications of the maxillary sinus:
a review. Int Endod J 2002;35:12741.
4. Harrison DF. Oro-antral fistula. Br J Clin Pract 1961;15:16974.
5. Selden HS. Endo-antral syndrome and various endodontic complications. J Endod
1999;25:38992.
6. Ioannides C, Borstlap WA. Apicoectomy on molars: a clinical and radiographical
study. Int J Oral Surg 1983;12:739.
7. Persson G. Periapical surgery of molars. Int J Oral Surg 1982;11:96100.
8. Watzek G, Bernhart T, Ulm C. Complications of sinus perforations and their man-
agement in endodontics. Dent Clin North Am 1997;41:56383.
9. Wehrbein H, Diedrich P. Progressive pneumatization of the basal maxillary sinus
after extraction and space closure [in German]. Fortschr Kieferorthop 1992;53:
7783.
10. Sharan A, Madjar D. Maxillary sinus pneumatization following extractions: a
radiographic study. Int J Oral Maxillofac Implants 2008;23:4856.
11. Sharan A, Madjar D. Correlation between maxillary sinus floor topography and
related root position of posterior teeth using panoramic and cross-sectional
computed tomography imaging. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
2006;102:37581.
12. Jun BC, Song SW, Park CS, et al. The analysis of maxillary sinus aeration
according to aging process; volume assessment by 3-dimensional reconstruc-
tion by high-resolutional CT scanning. Otolaryngol Head Neck Surg 2005;
132:42934.
13. Louie AF. Surgical and Radiologic Anatomy for Oral Implantology, 1st ed.
Hanover, Germany: Quintessence Publishing; 2013:14264.
14. Howe RB. First molar radicular bone near maxillary sinus: a comparison of CBCT
analysis and gross anatomic dissection for small bony measurement. Oral Surg Oral
Med Oral Pathol Oral Radiol Endod 2009;108:2649.
15. Kilic C, Kamburoglu K, Yuksel SP, et al. An assessment of the relationship between
the maxillary sinus floor and the maxillary posterior teeth root tips using dental
cone-beam computerized tomography. Eur J Dent 2010;4:4627.
16. Georgescu CE, Rusu MC, Sandulescu M, et al. Quantitative and qualitative bone
analysis in the maxillary lateral region. Surg Radiol Anat 2012;34:5518.
17. Jung YH, Cho BH. Assessment of the relationship between the maxillary molars and
adjacent structures using cone beam computed tomography. Imaging Sci Dent
2012;42:21924.
18. Von Arx T, Fodich I, Bornstein MM. Proximity of premolar roots to maxillary sinus: a
radiographic survey using cone-beam computed tomography. J Endod 2014;40:
15418.
19. Eberhardt JA, Torabinejad M, Christiansen EL. A computed tomographic study of the
distances between the maxillary sinus floor and the apices of the maxillary posterior
teeth. Oral Surg Oral Med Oral Pathol 1992;73:3456.
20. Kwak HH, Park HD, Kang MK, et al. Topographic anatomy of the inferior wall of the
maxillary sinus in Koreans. Int J Oral Maxillofac Surg 2004;33:3828.
21. Ok E, Gungor E, Colak M, et al. Evaluation of the relationship between the maxillary
posterior teeth and the sinus floor using conebeam computed tomography. Surg
Radiol Anat 2014;36:90714.
22. Pajin O, Centurion BS, Rubira-Bullen IR, et al. Maxillary sinus and posterior teeth:
accessing close relationship by cone-beam computed tomographic scanning in a
Brazilian population. J Endod 2013;39:74851.
23. Neelakantan P, Subbarao C, Ahuja R, et al. Cone-beam computed tomography study
of root and canal morphology of maxillary first and second molars in an Indian
population. J Endod 2010;36:16227.
24. Silva EJ, Nejaim Y, Silva AI, et al. Evaluation of root canal configuration of maxillary
molars in a Brazilian population using cone-beam computed tomographic imaging:
an in vivo study. J Endod 2014;40:1736.
Proximity of Maxillary Posterior Teeth to Maxillary Sinus 377