The removal of permanent teeth has been a

ORIGINAL ARTICLE Cephalometric changes in overbite and vertical facial height after removal of 4 first molars or first premolars Mark G. Hans, a Gordon Groisser, b Clay Damon, c Douglas Amberman, d Suchitra Nelson, e and J. Martin Palomo f Cleveland, Ohio Introduction: The objective of this study was to evaluate the outcome of standard edgewise orthodontic treatment with extraction of 4 first molars (6xT group) or Tweed edgewise treatment with extraction of 4 first premolars (4xT group). Materials and Methods: A cephalometric analysis that isolated tipping and bodily tooth movements of the maxillary and mandibular incisors and measured vertical skeletal changes in the anterior region of the maxilla and mandible was used. Thirty subjects treated by 10 practitioners comprised the 6xT group, whereas 31 subjects treated in the Case Western Reserve University orthodontic clinic were used in the 4xT group. Control groups (6xC and 4xC) were selected from untreated subjects enrolled in the Bolton-Brush Growth Study and were matched on age and gender. Data were collected before (T1) and after (T2) treatment. Results: Analysis of the data showed no statistically significant changes between 6xT and 6xC for any of the variables studied. An increase in overbite of 2.1 mm in the 6xT group was the result of small but clinically significant changes in both tipping and extrusion of maxillary and mandibular incisors. In the 4xT group, statistically and clinically significant changes were observed for intrusion of the maxillary and mandibular incisors, resulting in a 4.1-mm decrease in overbite. Importantly, both the 6xT and the 4xT groups showed no increase in mandibular vertical height during treatment. Conclusion: Both treatment strategies showed good control of vertical mandibular growth. Bodily intrusion of anterior teeth was the main contributor to correction of deep overbite in the Tweed edgewise sample. (Am J Orthod Dentofacial Orthop 2006;130:183-8) The removal of permanent teeth has been a controversial topic throughout orthodontic history, beginning with the great extraction debate between Angle and Case and continuing through Johnston s comparison of extraction and nonextraction 1 outcomes in borderline cases. Although it is difficult to argue against the popularity of nonextraction treatment with patients, parents, and referring dentists, extraction a Professor and chairman, Department of Orthodontics, School of Dental Medicine, Case Western Reserve University, Cleveland, Ohio. b Private practice, Potomac, Md. c Private practice, Spokane, Wash. d Clinical professor, Department of Orthodontics, School of Dental Medicine, Case Western Reserve University, Cleveland, Ohio. e Associate professor, Department of Community Dentistry, Case Western Reserve University, Cleveland, Ohio. f Assistant professor, Department of Orthodontics, Case Western Reserve University, Cleveland, Ohio. Supported by the Case Western Reserve University Orthodontic Alumni Fund. Based in part on Master s Theses submitted to Case Western Reserve University by Gordon Groisser and Clay Damon. Reprint requests to: Dr Mark G. Hans, Department of Orthodontics, School of Dental Medicine, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106-4905; e-mail, mark.hans@case.edu. Submitted, August 2004; revised and accepted, January 2005. 0889-5406/$32.00 Copyright 2006 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2005.01.021 of permanent teeth is still a valuable arrow in the orthodontist s quiver of options. The 2 primary reasons for removal of permanent teeth are to correct a discrepancy between tooth size and arch length, and to reduce bimaxillary protrusion. However, a third and somewhat underappreciated rationale for removing permanent teeth is to control the vertical dimension. Two types of vertical problems are of interest to orthodontists. The first clinical concern is lack of contact between the anterior teeth, or open bite. Several authors have suggested that removing permanent teeth from the posterior buccal segments with subsequent protraction to close spaces corrects the open bite by clockwise 2-4 rotation of the mandible. This rationale for extraction is sometimes referred to as the wedge hypothesis. Although this reasoning is popular with clinicians, case-controlled studies failed to demonstrate an association between decreased vertical facial growth and 5 extraction pattern. For example, Elham et al studied the effects of extraction of the mandibular first permanent molars and found no significant change in vertical facial dimensions, but they observed an increase in dental overbite (OB) because of lingual tipping of the mandibular incisors. Other authors reported increases 183

184 Hans et al American Journal of Orthodontics and Dentofacial Orthopedics August 2006 Table I. Descriptions of 6xT and 6xC groups Table II. Description of 4xT and 4xC groups 6xC 6xT 4xC 4xT Total sample (n) 25 30 Girls (n) 15 18 AgeatT1 12y5mo 2y7mo 12y4mo 2y6mo AgeatT2 15y7mo 2y3mo 15y6mo 2y2mo Time interval 3y2mo 1y5mo 3y2mo 1y6mo Total sample (n) 31 31 Girls (n) 15 15 Age at T1 12 y 11 mo 1y3mo 12y11mo 1y1mo AgeatT2 16y0mo 1y8mo 16y1mo 1y7mo Time interval 3y1mo 1y0mo 3y2mo 1y1mo in vertical facial growth during orthodontic treatment in 6-8 both extraction and nonextraction patients. Bazzucchi et al 9 compared vertical correction of open-bite malocclusions treated with the active vertical corrector or 4 first premolar extractions and fixed edgewise appliances with a matched sample of untreated openbite subjects from the Bolton collection. These authors also observed that most open-bite correction was achieved by lingual tipping of the maxillary and mandibular incisors. A second consideration for removing posterior teeth in patients with moderate to deep OB is to control vertical facial height during orthodontic treatment. A technique that focuses on vertical control of facial development ie, prevention of clockwise mandibular rotation during treatment is Tweed edgewise orthodontics. The Tweed orthodontic mechanical system is designed to control vertical facial growth during treatment. A major tenet of this treatment approach is that mechanically induced clockwise rotation of the mandible results in increased vertical facial growth, which is detrimental to facial appearance and chin projection. Although removal of permanent teeth is not required to apply these treatment mechanics, the extraction of premolars is commonly recommended for both openbite and deepbite patients. Although many case reports have supported the belief that the Tweed system controls vertical growth, there are no published casecontrolled clinical studies on this subject. 10-12 In this retrospective study, we used preexisting records to evaluate the effects of removing either 4 first premolars or 4 first molars on vertical dimensions and OB. When the first premolars were removed, specific orthodontic treatment mechanics designed to control clockwise mandibular rotation during treatment (Tweed mechanics) were used, whereas, when the first molars were removed, no special mechanical methods were used to control vertical facial growth. Fig 1. Landmarks used. MATERIAL AND METHODS Finished patients treated with all 4 first permanent molars extracted were solicited from the Cleveland Society of Orthodontists, the Electronic Study Club for Orthodontists, and clinicians known from the literature to treat patients with extraction of first molars. Seventy-two patients were identified, and, after review of their treatment records, 45 met the inclusion criteria: (1) pretreatment (T1) and posttreatment (T2) cephalograms of diagnostic quality, (2) extraction of all 4 first permanent molars, and (3) full maxillary and mandibular fixed orthodontic appliance treatment. These 45 subjects were then sorted by SN-mandibular plane (SN-MP) angle (high to low) and vertical OB (minimum to maximum). The final treatment sample consisted of the 30 patients with the highest SN-MP values and the lowest vertical OB measurements. These subjects were called the 6xT group. Ten orthodontists contributed subjects to this final sample. Thirty-one consecutively treated patients were selected from the archives of the Case Western Reserve University orthodontic clinic. These criteria were used for selection: (1) moderate to deep OB (75%-100%), (2) extraction of 4 first premolars, (3) Tweed philosophy of orthodontic mechanics, and (4) adequate T1 and

American Journal of Orthodontics and Dentofacial Orthopedics Volume 130, Number 2 Hans et al 185 Fig 2. Schematic diagram of the skeletal, dental, and angular variables used. T2 cephalometric radiographs. These subjects were called the 4xT group. Sixty-one untreated controls, matched for age and sex, were selected from the 4309 subjects in the Bolton collection. 13 After initial data analysis, 5 controls - se lected to match the 6xT sample were found to have had mandibular holding arches and, therefore, were treated. These subjects were removed from the first molar extraction control (6xC) sample, and the remaining 25 subjects were used for comparison with the 6xT group. The untreated Tweed sample (4xC) comprised 31 subjects (Tables I and II). All radiographs were traced on 0.003-in matte acetate with a 0.5-mm mechanical pencil. As previously described, a simplified cephalometric analysis was used; it included 6 linear and 2 angular measurements to describe vertical skeletal and dental changes over time. 14 This analysis compares serial radiographs with fiducial horizontal and vertical reference lines. At the T1 tracing, a horizontal line was drawn parallel to the Frankfort horizontal, and a perpendicular line was drawn to establish the vertical reference used for this analysis. The T2 tracing was superimposed on the T1 tracing by using cranial base landmarks, and both the horizontal and vertical fiducial lines were carried through to the T2 tracing. Six landmarks (anterior nasal spine [ANS], center of rotation of the maxillary and mandibular central incisors [CRU1 and CRL1], incisal edges of the maxillary and mandibular central incisors [IEU1 and IEL1], and menton [Me]) were identified on each cephalogram and projected onto the vertical reference line, keeping the landmark location parallel to the horizontal reference line (Fig 1). This procedure resulted in 6 linear variables. 1. Maxillary skeletal change (MXSK): the distance between the intersection of the vertical horizontal reference lines to ANS. 2. Bodily movement of the maxillary incisor (BUI): the distance between ANS and CRU1. 3. Tipping movement of the maxillary incisor (TUI): the distance between CRU1 and the IEU1. 4. Tipping movement of the mandibular incisor (TLI): difference between CRL1 and the IEL1. 5. Bodily movement of the mandibular incisor (BLI): distance between the CRL1 and Me. 6. Mandibular skeletal change (MNSK): the distance between ANS and Me projected onto the vertical reference line. The net changes in these 6 independent variables were used to compute the change in the dependent variable OB by using the following equation (Fig 2): Overbite MNSK BUI TUI BLI TUI where stands for the net change (tracing 2 minus tracing 1) for the 5 variables defined previously. Changes in the sixth variable anterior facial height were computed by using the following formula: Anterior facial height MXSK MNSK. This analysis is unique because it separates tipping from BUI and BLI. Linear measurements were made to the nearest 0.1 mm with an electronic digital caliper (Series 550, Mitutoyo America, Aurora, Ill). In addition to these 6 linear measures, 2 angular measures horizontal fiducial line to mandibular plane angle (MPA) and gonial angle were recorded. Angular measurements were made to the nearest 0.5 with an orthodontic protractor (Masel, Bristol, Pa). Dental and skeletal responses associated with vertical increases of incisor overlap were shown with a positive sign ie, movement toward the occlusal plane. For example, changes in lingual crown torque of the incisors that resulted in increases in TUI and TLI were given positive signs. Negative signs were given to movements associated with vertical decreases in incisor overlap ie, negative signs were given to dental and skeletal movements away from the occlusal plane. For example, increases in vertical mandibular growth that resulted in increased MNSK received negative signs. Statistical analysis All data were analyzed with the Statistical Package for the Social Sciences personal computer version (SPSS, Chicago, Ill). For OB, the 6 linear variables, and

186 Hans et al American Journal of Orthodontics and Dentofacial Orthopedics August 2006 Table III. Comparison between 6xC and 6xT Groups at T1 6xC 6xT MXSK 23.9 4.6 24.4 3.5.640 NS BUI 8.6 2.4 8.9 3.1.650 NS TUI 18.8 1.4 18.8 2.993 NS TLI 15.7 1.6 16.4 1.4.094 NS BLI 21 2.8 22.2 3.1.133 NS MNSK 60.1 4.7 64.1 5.6.006 * OB 4 1.6 2.2 1.7.000 Go 129.3 4.7 132.9 5.8.009 * MPA 35.3 3.7 36.9 8.363 NS NS, Not significant. *P.05; P.01. the 2 angular variables, means and standard deviations were calculated for both the treated groups (T1-T2) and the control groups (T1-T2). Mean changes were tested for statistical significance by using Student t tests. To evaluate investigator measurement error, 24 radiographs were randomly selected, and measurements for the 6 linear and the 2 angular variables were repeated. The second set of measurements was compared with the first set by using intraclass correlation coefficients, and tracing error was evaluated similarly. Twenty-four radiographs were retraced by an investigator (G.G.) and cephalometric analyses repeated on those tracings. The calculations for the 6 linear and 2 angular variables were compared for the T1 and T2 tracings with intraclass correlation coefficients. Table IV. Comparison of changes from T1 to T2 for 6xC and 6xT groups 6xC 6xT MXSK 2.0 1.8 2.5 2.6.438 NS BUI 1.1 1.8 1.4 1.6.607 NS TUI 0.1 1.1 0.5 1.4.243 NS TLI 0.1 0.9 0.2 1.2.858 NS BLI 2.3 2.3 2.3 2.3.942 NS MNSK 3.5 3.6 3.6 3.3.913 NS GO 0.7 2.2 0.5 3.6.800 NS MPA 0.1 1.8 0.5 2.2.484 NS OB 0.1 1.4 0.7 2.1.200 NS NS, Not significant. Fig 3. Schematic diagram illustrating net changes (treatment minus control) for patients treated with extraction of 4 first permanent molars. RESULTS 4xT having greater OB and more upright mandibular As previously noted, 30 subjects in the 6xT group incisors compared with the 4xC. During the experimental period, significant changes were recorded for were compared with 25 in the 6xC group. However, there were no observed differences between these both groups for MXSK, MNSK, BLI, Go, and MPA. groups for initial and final ages or time intervals. The 4xC group also showed a significant change in There were significant differences at T1 for MNSK, BUI. This same variable in the 4xT was not statistically different between T1 and T2. Statistically gonion (Go), and OB, with the 6xT group having longer faces, larger gonial angles, and less OB. significant differences between the changes in these During the observation period, both groups showed groups were observed for BUI ( 1.8 mm) and BLI significant changes in MXSK, BUI, BLI, and MNSK. ( 2.5 mm), resulting in a statistically significant However, there were no significant differences between the groups for any of these measurements. In variables were observed, none reached statistical decrease in OB. Although small changes in the other addition, both groups showed minimal changes in significance. The means and standard deviations for TUI and TLI during the 38-month period. There was all variables in the 4xT and 4xC groups are shown in a small increase in OB ( 0.7 mm) in the 6xT group. Tables V and VI, and the net changes are illustrated Means and standard deviations are shown in Tables schematically in Figure 4. III and IV, and the net changes are illustrated Measurement errors for both the 6xT, 6xC, 4xT, schematically in Figure 3. and 4xC samples were small, with all intraclass correlation coefficients between 97% and 99%. The tracing Thirty-one subjects comprised both the 4xT and 4xC samples; there were no significant differences error was considered acceptable, because the intraclass for age, sex, or time interval. Statistically significant correlation coefficients for the means of all variables changes were found at T1 for OB and TLI, with the were between 0.80 and 0.99.

American Journal of Orthodontics and Dentofacial Orthopedics Volume 130, Number 2 Hans et al 187 Table V. Comparison between 4xC and 4xT groups at T1 4xC 4xT MXSK 23.9 3.19 25.5 3.07.058 NS BUI 10.1 2.07 8.9 2.84.111 NS TUI 19.3 1.65 19.6 1.70 0.461 NS TLI 15.0 1.77 16.4 1.39.001 * BLI 23.5 1.59 23.4 2.19.803 NS MNSK 62.1 3.30 61.7 5.36.693 NS OB 5.8 1.28 6.8 1.05.001 * Go 122.4 5.92 122.5 5.68.898 NS MPA 31.4 4.22 32.5 5.60.441 NS *P.001. Table VI. Comparison of changes from T1 to T2 for 4xC and 4xT groups DISCUSSION 4xC 4xT MXSK 2.50 1.76 2.30 1.84.626 NS BUI 1.60 1.11 0.20 1.46.000 * TUI 0.05 0.29 0.32 1.28.111 NS TLI 0.08 0.48 0.14 0.99.259 NS BLI 1.90 1.55 0.72 1.81.000 * MNSK 3.50 2.14 3.80 2.39.581 NS Go 2.10 1.61 1.80 1.60.501 NS MPA 1.20 1.18 1.20 1.57.785 NS OB 0.30 0.89 4.20 1.15.000 * NS, Not significant. *P.001. A purpose of this molar extraction study was to test the wedge hypothesis. With the clinical popularity of this rationale for extraction in open-bite patients, it was surprising that removing the 4 largest teeth had minimal impact on OB. A possible explanation for this finding is that the sample was recruited solely on the basis of the extraction pattern chosen by the clinician. The 30 subjects with the least OB and highest SN-MP measurements were selected from the 45 subjects who met the inclusion criterion. Because the reason for removing the first molars was not an inclusion criterion, some first molars might have been removed because of caries rather than to control vertical facial growth. Thus, it is possible that the changes achieved by removing the first molars for vertical control and applying mechanics consistent with this treatment goal might cause greater reductions in vertical growth than those seen in this study. Clinicians should also keep in mind that this sample had increased mandibular vertical facial height and minimal OB, and a similar response might not be seen in patients with decreased lower vertical facial height and deep OB. Fig 4. Schematic diagram illustrating net changes (treatment minus control) for Tweed patients treated with extraction of 4 first bicuspids. In addition, the findings for TUI and TLI in the 6xT and 6xC groups were very different from those reported in a similar study of first premolar extraction in open-bite subjects. In the first premolar 9 extraction patients, lingual tipping of the TUI and TLI was the most important factor in increasing OB. However, the extraction of the first premolars was accompanied by a net increase in MNSK that partially neutralized the positive effects of uprighting the incisors. The most likely explanation for this increase in MNSK is that most orthodontic tooth movements tend to have an extrusive component. Previous cephalometric studies in nonextraction patients treated with cervical traction headgear, bionator, or straight-wire appliances all showed net increases in vertical facial growth. Therefore, a 11 clinically significant finding in this sample was that there was no net increase in MNSK in the 6xT group. This finding is especially important if one considers that this sample had increased lower vertical facial height and an open-bite tendency. Therefore, a possible approach for treating open-bite malocclusions could be to extract 4 first molars and 4 first premolars. This plan, although certainly aggressive given the strong preference for nonextraction treatment, would combine the positive effects observed for TUI and TLI in the previously published 4 premolar extraction open-bite sample while neutralizing the net increase in MNSK, as observed in our 6xT sample. The effects observed in the 4xT sample were consistent with published clinical case reports. Our results confirm that Tweed first premolar extraction mechanics in deep to moderate OB patients do not

188 Hans et al American Journal of Orthodontics and Dentofacial Orthopedics August 2006 increase MNSK or MPA more than in controls. In addition, we confirmed the long-held clinical belief that Tweed mechanics reduce vertical OB by bodily intrusion of the incisors and minimal tipping of the anterior teeth. Interestingly, there was no difference between the treatment and control groups for the skeletal measurements MXSK and MNSK; rather, the relative positions of the maxillary and mandibular incisors were affected by treatment. This means that without normal vertical growth ie, in an adult patient it might be impossible to achieve the same correction of OB. This is an instance when clinical case reports and a case-control study provide complementary evidence that clinicians can use to predict orthodontic treatment outcome. In the 4xT sample, we could not isolate the effect of the mechanical application system from the effect of premolar extractions. Based on previous studies of premolar extraction patients, it appears that the Tweed mechanical system is superior in controlling vertical facial growth during extraction treatment. It is not clear whether similar control can be achieved with Tweed mechanics without premolar extractions. CONCLUSIONS Extraction of 4 first molars or 4 first premolars with Tweed mechanics results in minimal changes in vertical facial height during orthodontic treatment. Although most orthodontic force applications tend to favor extrusion of molars, this tendency can be controlled by the extraction decision alone or, in deep-bite patients, by 13-14 premolar extractions and Tweed edgewise mechanics. We thank LaVerne Vogel and Bernard Tandler for their help in preparing this manuscript and the Bolton- Brush Growth Study for use of the records. REFERENCES 1. Paquette DE, Beattie JR, Johnston LE Jr. A long-term comparison of nonextraction and premolar extraction edgewise therapy in borderline Class II patients. Am J Orthod Dentofacial Orthop 1992;102:1-14. 2. Isaacson JR, Isaacson RJ, Speidel TM, Worms FW. Extreme variation in vertical facial growth and associated variation in skeletal and dental relations. Angle Orthod 1971;41:219-29. 3. Fränkel R, Fränkel C. Functional aspects of molar extraction in skeletal open bite. In: Graber LW, editor. Orthodontics: state of the art, essence of the science. St Louis: Mosby; 1986. p. 184-99. 4. Aras A. Vertical changes following orthodontic extraction treatment in skeletal open bite subjects. Eur J Orthod 2002;24:407-16. 5. Elham SJ, Aihaija A, McSheny PF, Richardson A. A cephalometric study of the effect of extraction of lower first permanent molars. J Clin Pediatr Dent 2000;24:195-8. 6. Staggers JA. Vertical changes following first premolar extractions. Am J Orthod Dentofacial Orthop 1994;105:19-24. 7. Staggers JA. A comparison of results of second molar and first premolar extraction treatment. Am J Orthod Dentofacial Orthop 1990;98:430-6. 8. Cusimano C, McLaughlin RP, Zernik JH. Effects of first bicuspid extractions on facial height in high-angle cases. J Clin Orthod 1993;27:594-8. 9. Bazzucchi A, Hans MG, Nelson S, Powers M, Parker S. Evidence of correction of bite malocclusion using active vertical corrector treatment. Semin Orthod 1999;5:110-20. 10. Lamarque S. Tweed-Merrifield sequential directional force nonpremolar extraction treatment: a case report. Semin Orthod 1996;2:268-72. 11. Boley JC, Mark JA, Sachdeva RC, Buschang PH. Long-term stability of Class I premolar extraction treatment. Am J Orthod Dentofacial Orthop 2003;124:277-87. 12. Klontz HA. Tweed-Merrifield sequential directional force treatment. Semin Orthod 1996;2:254-67. 13. Hans MG, Broadbent BH Jr, Nelson S. The Broadbent-Bolton Growth Study past, present, and future. Am J Orthod Dentofacial Orthop 1994;105:598-603. 14. Hans MG, Kishiyama C, Parker S, Wolf GR, Noachtar R. Cephalometric evaluation of two treatment strategies for deep overbite correction. Angle Orthod 1994;64:265-76.