CONTINUING EDUCATION

Predictive variables for the outcome of early functional treatment of Class III malocclusion
Lorenzo Franchi, DDS, a Tiziano Baccetti, DDS, PhD, b and Isabella Tollaro, MD, DDS c

Firenze, Italy
The aim of this study was to select a model of cephalometric and occlusal predictive variables for the results of early treatment of Class III malocclusion, due to mandibular protrusion in the deciduous dentition. Lateral cephalograms and dental casts of 45 subjects (22 boys and 23 girls) with Class III malocclusion in the deciduous dentition were analyzed at the start of treatment (mean age 5.57 _+ 0.85 years). The patients were treated with a functional appliance (removable mandibular retractor). All subjects were reevaluated after a mean period of 9.54 _ 2.28 years comprising active treatment plus retention. At this time, the sample was divided into two groups according to occlusal criteria: successful group (23 subjects) and unsuccessful group (22 subjects). Stepwise variable selection on the measurements at the time of first observation identified three predictive variables: the inclination of the condylar axis in relation to the stable basicranial line (CondAx-SBL), the inclination of the nasal line to the mandibular line (NL-ML), and the transverse width of the mandibular arch measured at the first deciduous molars. Discriminant analysis assigned a classificative power of 95.55% to the predictive model. On the basis of the equation generated by the multivariate statistical method, outcome of early treatment for each new case with Class III malocclusion in the deciduous dentition can be accordingly predicted. The important role of both vertical and transverse relationships in Class III prognosis was emphasized. (Am J Orthod Dentofac Orthop 1997;112:80-6.)

Prognosis of Class III malocclusion still represents one of the major challenges in contemporary orthodontics. Two intimately interconnected facets of this clinical problem are (1) the appraisal of long-term results of early treatment, and (2) the early identification of "surgical" cases. Adequate statistical methods on samples of subjects starting treatment in the very early ages have to be implemented to predict success or failure of interceptive therapy of developing Class III malocclusion. Multivariate approach to cephalometric data for predictive purposes has been recommended by Johnston 1 and used for differential diagnosis by Kowalski.2,3 Discriminant analysis was used by Battagel 4 to forecast relapse in subjects with Class III malocclusion treated by nonextraction techniques in the mixed and permanent dentitions. Stensland et al.5 used the same statistical method to predict short-term outcome of early orthopedic therapy of Class III malocclusion. Different kinds of functional or orthopedic apFrom the Department of Orthodontics, The University of Florence. aPhD program in Preventive Orthodontics. bPost-Doctoral Fellow. CProfessor, Head and Chairman. Reprint requests to: Dr. Lorenzo Franchi, Istituto di Odonto-GnatoStomatologia, Universit~ degli Studi di Firenze, Via del Ponte di Mezzo, 46-48, 50127, Firenze, Italy. Copyright 1997 by the American Association of Orthodontists. 0889-5406/97/$5.00 + 0 8/1/72686

pliances are available for treatment of Class III malocclusion associated with mandibular protrusion in the early developmental phases? -11 Our previous studies showed significant favorable craniofacial changes induced by a functional appliance (removable mandibular retractor, RMR) in children with Class III malocclusion, treated in the deciduous dentition.12, a3 The aim of the current study was to investigate whether any dentoskeletal characteristics in the deciduous dentition can be used as predictive variables for long-term results of early functional treatment of Class III malocclusion because of mandibular protrusion.
SUBJECTS AND METHODS

A sample of 45 subjects, 22 boys and 23 girls, with Class III malocclusion, was selected from the files of the Department of Orthodontics at the University of Florence. At the time of first observation, the subjects presented with Class III malocclusion in the deciduous dentition (concomitant presence of anterior crossbite, Class III deciduous canine relationship, and mesial step deciduous molar relationship). Class III malocclusion was due to mandibular protrusion in all the subjects, as revealed by cephalometric floating norms. 14,a5Functional analysis16 showed a postural rest position anterior to the occlusal position in all the patients. Mean age at the time of first observation was 5.57 -+ 0.85 years.

80

American Journal of Orthodontics and Dentofacial Orthopedics Volume 112, No. 1

Franchi, Baccetti, and Tollaro

81

Fig. 1. Removable mandibular retractor.

Early treatment of the malocclusion was accomplished by means of a functional appliance, removable mandibular retractor (RMR) (Fig. 1). Mean active treatment period was 2.79 _+ 1.24 years. During this period, the children wore RMR at least 14 hours a day (nighttime included). The degree of cooperation was acceptable in all the patients. Active treatment was discontinued when anterior crossbite had been completely corrected. Thereafter, all patients underwent a retention period by means of RMR, which was applied only during the night. Construction features and mode of action of RMR have been extensively reported elsewhere. 12 At the time of first observation lateral cephalograms and dental casts of all patients were analyzed. Lateral cephalograms of the 45 children with Class III malocclusion were taken with the same x-ray device and by a single technician. Focus-median plane distance was 152 cm and film-median plane distance was 10 cm, with an enlargement of 7%. Cephalometric analysis was based on a previously described basicranial reference system comprising two lines perpendicular to each other12: Stable Basicranial Line, SBL (passing through the most superior point of the anterior wall of the sella turcica at the junction with tuberculum sellae, point T, 17 and tangent to the lamina cribrosa of the ethmoid); Vertical T, VertT (perpendicular to SBL and passing through point T). The following landmarks were used: point A (A); point B (B); prosthion (Pr); infradentale (Id); pogonion (Pg); menton (Me); gonial intersection (Goi); articulare (At); condylion (Co); center of the condyle (Cs), i.e., a

SBL.

G ~

Fig. 2. Cephalometric linear measurements.

point equidistant from the anterior, posterior, and superior borders of the condylar head; basion (Ba); anterior nasal spine (ANS); and posterior nasal spine (PNS). The definitions of all these landmarks correspond to those given by Bj6rk, 18 ~degaard, 19 and Riolo et al. 2 The following measurements were performed: Linear measurements for the assessment of sagittal relationships (Fig. 2): A-VertT, B-VertT, Pr-VertT, IdVertT, Pg-VertT, Goi-VertT, Co-VertT.

82 Franchi, Baccetti, and Tollaro

AmericanJournalof Orthodonticsand DentofacialOrthopedics July 1997

NL ANS

Fig. 3. Cephalometdc angular measurements.

Fig, 4. Transverse measurements on dental casts (see text for explanations).

Linear measurements for the assessment of mandibular dimensions (Fig. 2): Co-Pg, Co-Goi, Goi-Pg. Angular measurements for the assessment of cranial base angulation (Fig. 3): Ba-T-VertT, Ar-T-VertT. Angular measurements for the assessment of vertical relationships (Fig. 3): Nasal line (NL)-SBL, mandibular line (ML)-SBL, nasal line-mandibular line (NL-ML), gonial angle (Ar-Goi-Me). Angular measurement for the assessment of condyle inclination (Fig. 3): Condylar axis (CondAx)-SBL. Condylar axis is a line passing through point condylion and point Cs.
The assessment of the method error for the cephalometric measurements was performed with the Dahlberg's formula 21 on 20 repeated measurements randomly selected from the total of the observations. The error ranged between 0.13 and 0.81 mm for the linear measurements, and between 0.19 and 0.93 for the angular measurements. On the dental casts of all the subjects at the time of the first observation, the following measurements for the assessment of interarch transverse relationships were carried out: Maxillary deciduous intermolar width, intended as the distance between the central fossae of right and left first maxillary deciduous molars (measurement 1 in Fig. 4). Mandibular deciduous intermolar width, intended as the distance between the tips of the distobuccal cusps of right and left first mandibular deciduous molars (measurement 2 in Fig. 4). Transverse discrepancy (TD), i.e., the difference between maxillary and mandibular deciduous inter-

molar widths (measurement 1 subtracted from measurement 2 in Fig. 4). Method error for dental cast measurements 21 on 20 repeated measurements randomly selected from the total of the observations was of 0.14 mm. The subjects were revaluated after a mean period of 9.54 + 2.28 years (mean age: 15.76 _+ 1.67 years) on the basis of their postretention occlusions. At this time, failure of treatment was defined as "the concomitant presence of Class III permanent molar relationship, Class I l l permanent canine relationship and anterior crossbite (of at least one incisor)." According to this rationale, the sample was divided into two groups: successful group (SG) and unsuccessful group (USG). The SG consisted of 23 subjects, 12 boys and 11 girls, mean age at the time of first observation 5.5 -+ 0.8 years; the USG comprised 22 subjects, 10 boys and 12 girls, mean age at the time of first observation 5.65 + 0.91 years. Reevaluation occurred 9.32 -+ 2.43 years after first observation for SG, and 9.78 -+ 2.14 years after first observation for USG.

Data Analysis
Discriminant analysis was applied to cephalometric and dental cast values of the 45 subjects at the time of first observation. Discriminant analysis is a multivariate statistical technique that is specifically designed to separate as widely as possible two groups of subjects taken from the same population. 22 In the current study, this statistical procedure was used to identify those

American Journal of Orthodontics and Dentofacial Orthopedics Volume 112, No. 1

Franchi, Baccetti, a n d T o l l a r o

83

Table I. Descriptive statistics for all cephalometric and dental cast variables
Total sample n = 45 Successful groupn = 2 3 Min 43 36 40 41 29 2 9 85 40.5 57 27 21 -2.5 19.5 12 119 58 Mean 53.87 49.98 52.87 52.32 47.32 12.65 16.78 97.24 46.26 65.54 40.80 35.84 6.06 29.06 23.00 128.39 71.27 SD 2.96 4.67 3.92 4.76 5.77 3.84 2.06 4.43 2.47 3.34 4.00 4.66 3.70 4.41 4.14 4.80 4.01 SE 0.62 0.97 0.82 0.99 1.20 0.80 0.43 0.92 0.51 0.69 0.83 0.97 0.77 0.92 0.86 1.00 0.84 Max 59 60 61 63 63 18 12 108 50.5 73 48 45 13 37 31.5 136 79 Min 49 40 44 42 36 3 20 89 41 60 34 26 -2 21 12 119 65 Mean 54.25 49.91 53.43 52.61 46.18 11.95 16.18 98.29 45.68 64.95 40.34 34.81 4.36 33.77 29.41 133.32 64.32 Unsuccessfulgroupn = 22 SD 4.04 6.07 4.73 5.43 7.78 5.10 2.83 5.65 3.31 4.30 5.42 5.81 4.68 6.96 5.63 4.76 3.21 SE 0.86 1.29 1.00 1.16 1.66 1.09 0.60 1.20 0.70 0.91 1.15 1.24 1.00 1.48 1.20 1.01 0.68 Max 61 58 60 60 58 20 21 108 52 71.5 49 46 16 44 39 143 69.5 Min 43 36 40 41 29 2 9 85 40.5 57 27 21 -2.5 19.5 18 122 58

Cephalometric variables
A-VertT (mm) B-VertT (mm) Pr-VertT (mm) Id-VertT (mm) Pg-VertT (ram) Goi-VertT (mm) Co-VertT (mm) Co-Pg (mm) Co-Goi (ram) Goi-Pg (ram) Ba-T-VertT () Ar-T-VertT () NL-SBL () ML-SBL () NL-ML () Ar-Goi-Me () CondAx-SBL ()

Mean
54.05 49.94 53.14 52.46 46.76 12.31 16.49 97.75 45.98 65.25 40.58 35.34 5.23 31.37 26.13 130.58 67.87

SD 3.50 5.34 4.30 5.04 6.77 4.46 2.46 5.04 2.89 3.80 4.70 5.22 4.25 6.20 5.85 5.36 5.03

SE 0.52 0.79 0.64 0.75 1.00 0.66 0.37 0.75 0.43 0.57 0.70 0.78 0.63 0.92 0.87 0.80 0.75

Max 61 60 61 63 63 20 21 108 52 73 49 46 16 44 39 143 79

Dental casts measurements Maxillary deciduous intermolar width (ram) Mandibular deciduous intermolar width (mm)
TD (mm) Max, Maximum; Min, minimum.

33.59 35.24 -1.62

1.8i 2.42 2.00

0.27 0.36 0.29

37.3 40.8 3.5

30 30 -6

33.37 33.71 -0.34

1.81 1.75 1.50

0.38 0.36 0.31

37.3 37.3 3.5

30 30 -3

33.83 36.84 -2.96

1.83 1.97 1.55

0.39 0.42 0.33

37.2 40.8 0

30 33.2 -6

dentoskeletal variables in the deciduous dentition most responsible for the prediction of success or failure of early Class III treatment. This multivariate approach was allowed by exploratory comparison between the values for all variables in SG and in U S G with one-way analysis of variance at the time of first observation. 22'23 Null hypothesis that all group means were equal was rejected at the significant level o f p < 0.01. To arrive at the best model for discrimination, the first phase of the analysis was to select those variables which were most important for group separation between SG and USG. Therefore stepwise variable selection was used to identify "good" predictor variables. Forward selection procedure with F-to-enter and F-to r e m o v e equal to 4 was chosen. 23 When the smallest set of significant discriminant variables was selected, predictive power (classificative power) of the model was tested with discriminant analysis. The aim of this last procedure is to provide a method to forecast which group a new case is most likely to fall into. Unstandardized discriminant function coefficients are calculated for each previously selected variable, along with a constant. 23 This leads to an equation that will assign a score to each patient. A "mean score" for each of the two groups is given. Halfway between these scores is the dividing value (critical score), which establishes to which of the two groups an individual case belongs. 4

Table II, Stepwise variable selection procedure

Variables in model CondAx-SBL F-Remove Variables not in model F-Enter 26.9855 14.3599 12.7936 Ar-Goi-Me NL-SBL ML-SBL A-VertT Ba-T-VertT Co-Pg Co-VertT B-VertT TD Id-VertT Pg-VertT Pr-VertT 1.5067 1.1273 1.1273 0.5584 0.5296 0.2139 0.1311 0.1217 0.0951 0.0639 0.0585 0.0520

Mandibular Intermolar Width

ML-NL

Step # 3. Stepwise variable selection R-squared = 0.7395, F-to-enter = 4, F-to-remove = 4, d.f. = 41.

RESULTS

Descriptive statistics for all c e p h a l o m e t r i c and dental cast variables both for total sample and for successful and unsuccessful groups at the t i m e of first observation (deciduous dentition) are listed in Table I. Stepwise variable selection g e n e r a t e d a threevariable m o d e l that p r o d u c e d the m o s t efficient separation b e t w e e n the two groups ( S G vs U S G ) .

84

Franchi, Baccetti, a n d Tollaro

American Journal of Orthodontics and Dentofacial Orthopedics July 1997

Table IlL D i s c r i m i n a n t a n a l y s i s
Classification results Predicted group membership 1 Actual group
G r o u p 1 (success) G r o u p 2 (failure)

2 %
95.65 4.55

Number of cases
23 22

No.
22 1

No:
1 21

%
4.35 95.45

Predictive variables
CondAx-SBL ML-NL Mandibular arch width

Unstandardized canonical discriminant function coefficients

0.20809 -0.11904 -0.31604 0.10603

Constant
Percentage of cases correctly classified: 95.55% Individual score = 0.20809(CondAx_SBL) -- 0.11904(ML_NL ) -- 0.31604(Mand. Dec. Interrnolar Width) ~1- 0.10603. Discriminant scores for group means (group centroids): Successful G r o u p = 1.61094. Unsuccessful G r o u p = - 1.68416.

Critical score = - 0.03651.

The variables selected were the inclination of the condylar axis, in relation to the stable basicranial line (CondAx-SBL), the inclination of the nasal line to the mandibular line (NL-ML), and the transverse width of the mandibular arch measured at the first deciduous molars (Table II). The classificative power of the selected three-variable model was 95.55% (Table III). Only one case in each group was not correctly classified. Unstandardized discriminant function coefficients of the selected variables, along with a calculated constant (Table III), lead to the following equation that provides individual scores for the assignment of a new case to U G or to USG: Individual score = 0.20809(CondAx-SBL~ 0.11904(ML_NL) -- 0"31604(Mantt.Dec.Int . . . . larWidth) + 0 . 1 0 6 0 3 .

Class III malocclusion. "Outcome" was considered as the result of active treatment plus retention. Therefore the aim of this study was not to assess relapse of early Class III treatment. All patients underwent a longterm evaluation of treatment results about 9.5 years after the start of treatment (around 15.5 years of age), when a great part of active growth had already occurred. A certain amount of residual craniofacial growth had still to take place in a few subjects at later ages. For all these reasons, it is not convenient to compare the results of the current work to those of previous investigators. Stensland et al.5 studied the short-term outcome of early treatment of Class III malocclusion; Battagel 4 analyzed a model for prediction of relapse in treated children with Class III map occlusion. Other peculiar features of the current study were as follows: Treatment of Class III malocclusion was started in the deciduous dentition. All patients presented with mandibular protrusion and were accordingly treated with a specific appliance. As a consequence, the proposed predictive model is not suitable for children with Class III malocclusion associated with maxillary retrusion and treated with face mask or protraction headgear. Definition of failure of early treatment was based on strict criteria, i.e., the same occlusal features that were used as a clinical rationale for the diagnosis of Class III malocclusion at the start of treatment. As to statistical method, discriminant analysis appears to be one of the most powerful tools for the

The critical score (i.e., the value dividing SG from USG) was -0.03661, that is the mean value of group centroids of the two groups (Table III). Each new case with Class III malocclusion associated with mandibular protrusion in the deciduous dentition that will show an individual score higher than the critical score will be treated successfully by functional appliances. On the contrary, each new case with Class III malocclusion presenting with a more negative individual score than the critical score can be predicted to respond poorly to early treatment. Both predictions can be made with a probability of error of about 5%.
DISCUSSION

The current investigation dealt with the predictability of outcome in early functional treatment of

American Journal of Orthodontics and Dentofacial Orthopedics Volume 112, No. 1

Franchi, Baccetti, and ToUaro 85

identification of predictive variables at the time of first observation. Nevertheless, there are likely to be other elements for the final result of treatment (cephalometric variables, heredity, soft tissue features) that have not been included in the current analysis, but that might improve the model. 1-4,24,25 In the current study, a set of cephalometric and dental cast measurements was chosen as representative of craniofacial and occlusal characteristics in the three dimensions. In fact, our previous research provided evidence that the cephalometric analysis comprising a basicranial reference system and measurements for the inclination of the condylar axis was quite efficient in the description of mandibular resizing and reshaping (anterior morphogenetic rotation), induced by early functional treatment of Class III malocclusion. 12,13Moreover, experimental studies emphasized similar mandibular changes produced by retractive forces on the growing mandible. 26-29 The good discriminant power of only three of the chosen measurements witnessed that the analysis was viable for the identification of predictive variables in early treatment of mandibular protrusion. The three predictive measurements in the deciduous dentition were (1) the orientation of the condylar head, (2) the maxillomandibular vertical relationship, and (3) the transverse size of the mandibular arch. Better prognosis of early treatment of Class III malocclusion is associated with a large CondAx-SBL angle, i.e., with an upwardforward inclination of the condyle. It has been demonstrated that the major effect of R M R therapy in children with Class III malocclusion consists of a more upward and forward direction of condylar growth when compared with an untreated control group. 12,13Therefore it appears that early treatment of Class III malocclusion is more effective when the orientation of the condyle at the start of treatment is concordant with the expected change induced by treatment itself. The two other predictive factors that were selected by multivariate analysis regard facial vertical relationships and transverse dimensions of the mandibular arch. These results claim for a more attentive evaluation of vertical and transverse relationships in early Class III diagnosis and treatment planning. Interestingly enough, no sagittal parameter or dimensional measurement of the mandible on the lateral films was actually included in the predictive model. Skeletal open bite and excessive width of the mandibular arch are to be considered as unfavorable signs in prognosis of Class III malocclusion in the deciduous dentition.

The predictive model that was developed in the current study can identify good or bad responders to early treatment of Class III malocclusion with a functional appliance. This does not necessarily mean that the model is able to automatically recognize "surgical versus not-surgical" cases. In bad responders, the possibilities of alternative orthodontic therapies should be taken into account in comprehensive treatment planning.
CONCLUSIONS

The appraisal of long-term results of early treatment of Class III malocclusion associated with mandibular protrusion was performed by means of discriminant analysis, to identify significant predictive variables in the deciduous dentition. Three predictive measurements were selected: the inclination of the condylar axis in relation to the stable basicranial line (CondAx-SBL), the inclination of the nasal line to the mandibular line (NL-ML), and the transverse width of the mandibular arch measured at the first deciduous molars. The classificative power of the model for the prediction of success or failure of early Class III treatment is 95.55% for each new individual case.
REFERENCES 1. Johnston LE. A statistical evaluation of ccphalometric prediction. Angle Orthod 1968;38:284-304. 2. Kowalski CA, Nasjleti CE, Walker GF. Differential diagnosis of adult male black and white populations. An~e Orthod 1974;44:346-50. 3. Kowalski CJ, Nasjleti CE, Walker GF. Dentofacial variations within and between four groups of adult American males. Angle Orthod 1975;45:146-51. 4. Battagel J. Discriminant analysis: a model for the prediction of relapse in Class III children treated orthodontically by non-extraction technique. Eur J Orthod 1993; 15:199-209. 5. Stensland A, Wisth PJ, B0e OE. Dentofacial changes in children with negative overjet treated by a combined orthodontic and orthopaedic approach. Eur J Orthod 1988;10:39-51. 6. Thilander B. Treatment of Angle Class III malocclusion with chin-cup. Trans Eur Orthod Soc 1963;39:384-98. 7. Matsui Y. Effect of chin cap on the growing mandible. J Jpn Orthod Snc 1965;24:165-81. 8. Graber LW. Chincup therapy for mandibular prognathism. Am J Orthod 1977;72:23-41. 9. Sakamoto T. Effective time for the application of orthopedic force in the skeletal Class III malocclusion. Am J Orthod 1981;80:411-6. 10. Mitani H, Sakamoto T. Chin cap force to a growing mandible: long term clinical reports. Angle Orthod 1984;54:93-122. 11. Allen RA, Connolly IH, Richardson A. Early treatment of Class III incisor relationship using the chincap appliance. Eur J Orthod 1993;15:371-6. 12. Tollaro I, Baccetti T, Franchi L. Mandibular skeletal changes induced by early functional treatment of Class lII malocclusion: a superimposition study. Am J Orthod Dentofac Orthop 1995;108:525-32. 13. Tollaro I, Baccetti T, Franchi L. Craniofacial changes induced by early functional treatment of Class III malocclusion. Am J Orthod Dentofac Orthop 1996;109:310-8. 14. Segner D. Floating norms as a means to describe individual skeletal patterns. Eur J Orthod 1989;11:214-20. 15. Tollaro I, Baccetti T, Franchi L. Floating norms for the assessment of craniofacial pattern in the deciduous dentition. Eur J Orthod 1996;18:359-65. 16. Graber TM, Rakosi T, Petrnvic A. Dentofacial orthopedics with functional appliances. St Louis: CV Mosby Co., 1985:115-8. 17. Viazis AD. The cranial base triangle. J Clin Orthod 1991;25:565-70. 18. Bjrrk A. The face in profile. Sven Tandl/ik Tidskr 1947;40:Suppl 5B. i9. Odegaard I. Growth of the mandible studied with the aid of metal implants. Am J Orthod 1970;57:145-57. 20. Riolo ML, Moyers RE, McNamara JA, Hunter WS. An atlas of craniofacial growth: cephalometric standards from the University School Growth Study.

86 Franchi, Baccetti, and Tollaro

American Journal of Orthodontics and Dentofacial Orthopedics July 1997

26. Janzen EK, Bhiher JA. The cephalometric, anatomic, and histologic changes in Macaca mulatta after application of a continuous-acting retraction force on the mandible. Am J Orthod 1965;5i:823-55. 27. Charlier JP, Petrovic A, Linck G. La fronde mentonni6re et son action sur la croissance mandibulaire. Recherches exp6rimental chez le rat. Orthod Fr 1969;40: 99-I13. 28. Asano T. The effects of mandibular retractive force on the growing rat mandible. Am J Orthod Dentofac Orthop 1986;90:464-74. 29. Petrovic A, Stotzmann J. Effects on the rat mandible of a chincup-type appliance and of partial or complete immobilization. Proc Finn Dent Soc 1991;87:85-9L

21. 22. 23. 24. 25.

Ann Arbor: Center for Human Growth and Development, University of Michigan, 1974. Dahlberg G. Statistical methods for medical and biological students. Neff York: Interscience Publications, 1940. Norusis MJ. SPSS/PC+ Advanced Statistics V2.0. Chicago: SPSS Inc., 1988:B1-B39. Statistical Graphics Corporation. Statgraphics. Vet. 2.6. Rockville (MD): STSC, Inc., 1987. Hirschfeld WJ, Moyers RE. Prediction of craniofacial growth: the state of the art. Am J Orthod 1971;60:435-44. Skieller V, Bj6rk A, Linde-Hansen T. Prediction of mandibular growth rotation evaluated from a longitudinal implant sample. Am J Orthod 1984;86:359-70.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS

Bound volumes of the American Journal of Orthodontics and Dentofacial Orthopedics are available to subscribers (only) for the 1997 issues from the Publisher, at a cost of $79.00 ($93.30 Canada and $90.00 international) for Vol. 111 (January-June) and Vol. 112 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby-Year Book, Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; telephone (314)453-4351 or (800)325-4177. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.