Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
EFFECTS OF EXTRACTION VS NON-EXTRACTION ORTHODONTIC TREATMENT ON VERTICAL SKELETAL AND DENTAL MEASUREMENTS Vasileios Charalampakis, D.D.S. An Abstract Presented to the Graduate Faculty of Saint Louis University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Dentistry 2015 Abstract Introduction: One of the most important steps in orthodontic treatment planning is the extraction decision. The “wedge effect” hypothesizes that extracting posterior teeth allows the mandible to hinge closed, thereby allowing better vertical control. The objective of the present study was to compare the different outcomes between patients treated non-extraction and patients treated with extractions, in order to assess the validity of this concept. Subjects and Methods: Pre- and post-treatment cephalograms of 191 normal, healthy subjects (67 treated without extractions, 61 treated with four first premolar extractions and 63 treated with four second premolar extractions) were obtained from the archive of the Center for Advanced Dental Education orthodontic clinic of Saint Louis University. Class I, Class II and Class III subjects were similarly distributed across the three groups. Tracings were superimposed according to Björk’s method for structural superimposition and fiducial lines were drawn for the maxilla and mandible. A Cartesian system was used to measure vertical and horizontal changes of ANS, upper first molar and central incisor, Gonion, Menton, lower first molar and lower central incisor. Overbite, as well as SN-GoMe, SN-FOP and the angle between the maxillary and mandibular fiducial lines was also measured. Because the different treatment protocols resulted in significantly different treatment times, analysis of covariance was performed to test for different increments of change among the 3 groups while controlling for the effect of treatment time. Results: There was a statistically significant difference in horizontal displacement of upper first and lower first molars between the 3 groups. The lower incisors showed significantly different amounts of extrusion for each group. There were no statistically significant differences in amounts of change in any of the other variables. Conclusions: The present study failed to produce evidence of favorable vertical changes as a result of extractions. Other than lower incisor vertical displacement, there was no difference in magnitude of change in vertical measurements between the 3 groups. EFFECTS OF EXTRACTION VS NON-EXTRACTION ORTHODONTIC TREATMENT ON VERTICAL SKELETAL AND DENTAL MEASUREMENTS Vasileios Charalampakis, D.D.S. A Thesis Presented to the Graduate Faculty of Saint Louis University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Dentistry 2015 COMMITTEE IN CHARGE OF CANDIDACY: Professor Rolf G. Behrents, Chairperson and Advisor Professor Emeritus Lysle E. Johnston Jr. Professor Eustaquio A. Araujo Associate Clinical Professor Donald R. Oliver i “For every complex problem there is an answer that is clear, simple, and wrong.” —Henry Louis Mencken ii DEDICATION To my mother, the strongest person I have ever met and a beacon of patience and bravery. To my father, who taught me the value of humility, determination, hard work and perseverance. To my teachers in Saint Louis University, who honored me with the opportunity to learn from them. To my loving partner Panagiota, who supported me every step of the way. I am blessed to have you by my side. iii ACKNOWLEDGEMENTS The author acknowledges the following individuals and expresses his sincere gratitude for their contributions: Dr. Rolf G. Behrents, for his assistance in topic selection, result interpretation and corrections; Dr. Lysle E. Johnston Jr., for his assistance in tracing cephalograms, suggestions regarding the methods and statistical analysis and time spent on corrections; Dr. Eustaquio A. Araujo, for his suggestions, corrections and mentoring throughout the course of my graduate education; Dr. Donald R. Oliver, for his advice on this thesis and time spent on corrections; Dr. Heidi Israel, for her patience and assistance with the statistical analysis and result interpretation. Finally, I would like to thank the rest of the faculty for sharing their knowledge and experience, my classmates for riding along in this journey, and the staff at CADE for allowing us to focus on learning and making me look forward to going to school every morning. I am humbled and honored to have been part of such an incredible group of people. iv TABLE OF CONTENTS LIST OF TABLES......................................... vi LIST OF FIGURES....................................... vii CHAPTER 1: REVIEW OF THE LITERATURE..................... 1 Introduction .................................. 1 The Vertical Dimension ...................... 2 Growth Pattern .............................. 2 The Wedge Hypothesis ........................ 4 The Extraction Decision ..................... 5 Clinical Studies .............................. 6 Overview ..................................... 19 Statement of Thesis .......................... 23 CHAPTER 2: JOURNAL ARTICLE............................. 28 Abstract ..................................... 28 Introduction ................................. 30 Subjects and Methods ......................... 32 Sample ..................................... 32 Cephalometric Analysis ..................... 33 Error Study ................................ 37 Statistical Analysis ....................... 37 Results ...................................... 39 ANCOVA ..................................... 45 Discussion ................................... 47 Limitations ................................ 47 The Wedge Hypothesis ....................... 49 Results Compared to Literature ............. 50 Overview and Future Research ............... 54 Conclusions .................................. 56 Literature Cited ............................. 57 Appendix............................................... 61 Vita Auctoris.......................................... 67 v LIST OF TABLES Table 1.1 Studies analyzing impact of extractions on vertical measurements................ 17 Table 2.1 Distribution of patients by Angle classification in each extraction group. 33 Table 2.2 Means and standard deviations for treatment time in days....................... 39 Table 2.3 Post-hoc analysis for treatment time between groups.......................... 40 Table 2.4 Descriptive statistics pre-treatment.... 41 Table 2.5 Descriptive statistics post-treatment... 42 Table 2.6 Descriptive statistics for increments of change.................................. 43 Table 2.7 Descriptive statistics for increments of change adjusted for treatment time according to ANCOVA................... 44 Results compared to similar studies... 53 Table 2.8 vi LIST OF FIGURES Figure 2.1 Figure 2.2 Figure A.1 Figure A.2 Figure A.3 Measurements based on maxillary superimposition............................ 35 Measurements based on mandibular superimposition......................... 36 Upper first molar horizontal displacement.................................. 62 Lower first molar horizontal displacement.................................. 64 Lower incisor vertical displacement... 65 vii CHAPTER 1: REVIEW OF THE LITERATURE Introduction Since the fathers of orthodontics laid the foundation of the specialty, the first practitioners engaged in debates regarding solutions to common clinical problems, chief among which was addressing tooth space-arch length discrepancies. Angle was an avid supporter of nonextraction treatment,1 whereas Tweed is widely considered the pioneer of extraction treatment. The pendulum in the extraction versus non-extraction approach has swung back and forth numerous times since then, with ideas and trends that were once popular lost in time, only to be repackaged and presented anew. Therefore, it is important to review the scientific literature in order, not only to determine on which direction future research should focus, but also to guard against anachronistic arguments and superseded theories. Dealing with different types of malocclusion soon gave rise to the need for a classification system that would make diagnosis, communication and treatment planning more efficient and consistent. Angle’s nomenclature, first published in the Dental Cosmos in 18992 has proven to be a simple and powerful tool. As is often the price of simplicity, however, it suffers from certain 1 limitations.3 For example, Angle’s classification strictly refers to sagittal discrepancies, disregarding vertical and transverse issues. The Vertical Dimension In 1964, Sassouni and Nanda4 proposed an analysis that included eight skeletal open bite and eight skeletal deep bite combinations to describe facial proportions, introducing a system that took the vertical dimension into account. Sassouni5 further elaborated on this classification system in a paper published in 1969, describing four “syndromes” of characteristics, skeletal deep bite, skeletal open bite, skeletal Class II and skeletal Class III. Schudy6 in 1964 introduced the terms hypo- and hyperdivergent. Fields and coworkers7 argued there was a lack of documentation in the scientific literature regarding the morphological differences between long-, normal- and short-faced patients. In their research, they found that long-faced patients had greater anterior face height, mandibular plane angles and mandibulopalatal plane angles. Growth Pattern It is important to know what one may expect from normal growth in order to be able to assess the effects 2 of orthodontic and orthopedic interventions. In 1984, Cangialosi8 described the skeletal features of patients with anterior open bite. Comparing 60 open bite patients (30 permanent and 30 mixed dentition) to 60 untreated Class I normal subjects, he found significantly greater SN-GoGn and gonial angles in the open bite group as well as shorter posterior face height and greater anterior face height. He reported significant differences in size, but not proportions. Ratios and angles remained relatively constant between mixed and permanent dentition, which suggests vertical problems do not self-correct as an individual ages. His study is important because it provides useful information regarding changes that occur with growth. Surender Nanda9 examined a sample of 32 patients between 3 and 18 years of age to determine patterns of vertical development. Skeletal deep bite and open bite subjects were selected based on lower face height (ANSMe) to total face height (N-Me) ratio. He reported significant differences between the two groups in anterior total and anterior lower face height, whereas ramal height showed no difference. His findings contrast those of others4, 6 who reported significantly shorter ramal height for open bite patients. Nanda concluded that vertical developmental patterns are determined at a very 3 young age, before the eruption of the first permanent molars, and the differences between the two groups increase with time. The Wedge Hypothesis Schudy in 196810 argued that vertical growth affects all facial proportions and proposed treatment strategies he believed prudent in order to minimize deleterious effects of unfavorable growth and reduce the chance of relapse. He suggested that extractions may be considered for better vertical control, even in cases with sufficient space. Nasby and coworkers11 (1972) studied high and low angle patients. They believed that extractions allow the posterior teeth to drift mesially, and this drift in turn produces forward rotation of the mandible and a reduction in anterior facial height. They concluded that extractions are indicated for high angle patients but should be avoided for low angle individuals. Along the same lines, Joondeph and Riedel12 advocated second premolar serial extraction in order to control overbite, due to a favorable mesial drift pattern. Pearson13 discussed vertical control in hyperdivergent patients, advocating extraction of four first premolars and vertical-pull chincup to close open bites prior to placement of fixed appliances. He offered several explanations about the possible cause of the bite closure he observed, 4 among which was the mesial drift of the posterior teeth “out of the wedge” which in turn allowed the mandible to rotate forward and “hinge closed.” The Extraction Decision Clinicians may opt for different extraction patterns depending on each patient’s “needs.” Ketterhagen14 sought to investigate the decision making process involving the extraction of first versus second premolars. Feeling a more scientific approach was necessary, he compared two groups of patients that had all four first premolars or all four second premolars extracted. Several cephalometric measurements were examined (including mandibular plane angle) in an effort to identify what influenced the extraction decision. He found no difference between the two groups in terms of skeletal measurements, but was able to identify soft tissue differences. Vertical control was not the primary criterion when selecting an extraction pattern. Yamaguchi and Nanda’s15 work was not specifically focused on extraction patterns and vertical dimensions, but certain conclusions can be drawn by analyzing their data. They compared the effects of different types of force application on 73 extraction and 48 non-extraction patients. Interestingly, they reported larger lower anterior face height and smaller ramus height in the extrac5 tion group, suggesting that the vertical dimension was perhaps an important factor in the extraction decision. Controlling for mechanics and sex, they concluded that vertical control of maxillary and mandibular molars is necessary in order to prevent posterior rotation of the mandible. Clinical Studies Several studies have directly approached the subject of extractions and vertical dimensions. Others focus on different aspects of the extraction versus nonextraction dilemma, but pertinent information regarding vertical changes can be found in their tables or discussion as well. One of the first papers to examine the validity of the wedge concept was Garlington and Logan’s16 study (1990) based on a sample of 23 Class II, division 1 hyperdivergent patients (initial SN-GoGn greater than 38°) who underwent enucleation of mandibular second premolars followed by extraction of maxillary first premolars. Using Björk’s17 method of superimposition to measure preand post-treatment skeletal and dental changes, they compared their findings to those of Isaacson and co-workers18 and found a non-significant decrease of 0.8° in SN-GoGn and a statistically significant reduction in lower face height in the post-treatment group. They concluded that 6 treatment resulted in increased forward rotation of the mandible, with possible compensatory maxillary growth offered as an explanation for failing to detect a significant difference in total face height. Another study by Luecke and Johnston19 focused mainly on the alleged retropositioning of the condyle supposedly caused by extracting maxillary first premolars. It was published in 1992, when the debate regarding the relationship between orthodontics and temporomandibular disorders was at its prime. They used Björk’s methods of regional superimposition and measured growth and treatment changes parallel to the mean functional occlusal plane. To measure condylar displacement, a Cartesian coordinate system was used where averaged Frankfort Horizontal served as the x-axis and a perpendicular through SE point was the y-axis. Their sample showed forward displacement of the mandibular basal bone, and the FMA decreased by an average of 1°. Although not exclusively focused on the wedge effect, their findings seem to support the concept. In order to examine long-term changes after orthodontic treatment, Luppanapornlarp and Johnston20 recalled 62 Class II patients (33 extraction and 29 nonextraction) an average 15.3 years post-treatment. Again, Björk’s regional superimposition technique was used in 7 the “pitchfork analysis” to evaluate growth and treatment changes. Only y-axis change immediately post-treatment was different. There were no long-term differences in changes observed in any of the vertical measurements they studied. Anterior face height, defined as the distance between ANS-Menton, was the primary focus of Chua and coworkers,21 who measured ANS-Me changes on a sample of 174 Class I and Class II patients. They examined the effects of growth, duration of treatment, extraction pattern and treatment mechanics. They used data from the Michigan growth standards to normalize individual face height measurements, in an attempt to adjust for the different effects of growth. No superimposition was done; instead, pre- and post-treatment measurements were standardized and compared. A significant increase in ANSMenton was observed in the non-extraction group, whereas in the extraction group no change could be detected. Thus they noted that extractions alone may not be sufficient to reduce lower anterior face height. The authors failed to detect differences between the various treatment mechanics such as cervical headgear, intermaxillary elastics and tip-back bends. Another significant point was that in the Class II group, clinicians tended to extract more often in the patients with longer lower anterior 8 face height. They concluded that non-extraction treatment increases ANS-Me and results in a clockwise rotation of the mandible. Aras22 examined a sample of open bite patients, fifteen of which had first premolar extractions and seventeen had second premolar or first molar extractions. He attempted to minimize the effects of growth by using a hand-wrist radiographic method to select patients that were past their peak growth velocity. First premolar extractions did not result in any significant mandibular rotation; however, second premolar or first molar extractions produced a forward rotation of the mandible, therefore lending support to the wedge hypothesis. Molar extractions were also studied by Hans and coworkers23 (2006) who examined 30 patients who had undergone orthodontic treatment with first molar extractions, 31 patients with first premolar extractions, that were then matched with two groups of untreated subjects from the Bolton-Brush Growth Study based on age and gender. They found that both treated groups showed no increase in mandibular vertical height and concluded that both strategies were appropriate for vertical mandibular growth control. The minimal impact that first molar extractions had on overbite was labeled as “surprising.” Although their findings seemingly support the wedge hypothesis, it 9 should be noted that the lack of a non-extraction treated group does not allow comparisons to be drawn between the two modalities. Forward mandibular rotation and vertical control is thoroughly discussed in a recent article published in 2013 by Bayirli and coworkers.24 A group of 36 patients treated with four first premolar extractions and standard edgewise treatment was compared to a matched group of untreated subjects from the Bolton-Brush Growth Study Center. Structural superimposition in the cranial base, maxilla and mandible were used to analyze dental and skeletal contributions to the overall measured changes as measured by the so-called “pitchfork analysis” was performed. Anterior facial height, posterior facial height and overbite showed significantly different changes compared to the control group, but FMA was not significantly different. In the conclusions, the authors stress the importance of using structural superimposition to measure rotation, since it cannot be inferred from surface landmarks alone, a point taken into consideration in the design of the present study, Other authors, however, have reported similar changes in vertical measurements in both extraction and non-extraction groups, thereby contradicting the findings of the previously summarized studies. For instance, 10 Klapper and coworkers25 examined dolichofacial and brachyfacial groups that had undergone orthodontic treatment, either non-extraction or with four first premolar extractions. They were unable to detect differences in the facial axis change among the groups. Interestingly, a positive correlation was reported between upper molar displacement and facial axis change only in the nonextraction group, for both brachyfacial and dolichofacial subjects. Lin and coworkers26 approached the subject from a slightly different perspective. They compared a group of 28 dolichofacial patients and 29 mesofacial patients, matched for age; both groups were treated without extractions. This was done in order to examine the hypothesis that non-extraction treatment would result in greater vertical increase for the dolichofacial group. Cephalograms were taken pre- and post-treatment and 2 years after completion of orthodontic treatment. High-pull headgear was used for 12 dolichofacial and 6 mesofacial patients. No statistically significant changes were detected between the two groups, both of which showed similar increases in facial height, from which the authors concluded that facial height depends more on genetics and less on such environmental factors as orthodontics. 11 Hyperdivergent subjects (SN-GoGn angle of 36° or greater) were studied by Cusimano and coworkers,27 who investigated the effects of four first premolar extractions on patients averaging 11 years, 10 months of age at the beginning of treatment. A “modified mandibular plane” using the mandibular canal and the internal lower border of the symphysis was used to measure true matrix rotation. They were unable to detect a reduction in vertical dimensions and argued that, although mesial molar movement should cause a forward rotation of the mandible, this movement is nullified by extrusion of the posterior teeth. It was also concluded that the surface changes on the lower border of the mandible were not significant to the degree that GoGn would be rendered invalid. In 1992, Paquette and coworkers28 studied longterm changes in 33 extraction and 30 non-extraction patients, an average 14.5 years after the completion of treatment. They, too, used Björk’s structural superimposition methods to examine basal bone and tooth displacement and measure changes parallel to the mean functional occlusal plane. There was no difference between the two groups immediately post-treatment and the authors noted that there was no correlation between treatment strategy and overbite relapse. Interestingly, there were long-term differences in FMA, S-Go, ANS-Me and N-Me, contrary to 12 the findings of Luppanapornlarp and Johnston.20 However, any effects from mesial movement of the posterior teeth on vertical dimensions should be measurable immediately post-treatment. Therefore, the long-term difference between the two groups does not constitute solid evidence for the wedge concept. Staggers29 compared 38 Class I first premolar extraction and 45 Class I non-extraction cases at an average age of 14 years, 5 months. Pre- and post-treatment lateral cephalograms were traced, then measured using a digitization program (Dentofacial Planner version 5.3; Dentofacial Software Inc., Toronto, Canada). Student’s ttests were used to compare the changes of eight variables in the two groups, which showed no statistically significant differences. Mandibular plane angle increase, extrusion of upper and lower molars and anterior face height increase were observed in both groups. Staggers concluded that her findings do not support the notion of vertical dimension reduction after extraction orthodontic treatment. It was argued that perhaps extractions have a different effect on Class II patients, who require more retraction of the posterior teeth. This hypothesis was investigated by Kocadereli,30 who compared 40 Class I non-extraction to 40 Class I four first premolar extraction cases. Cephalograms were 13 traced, digitized, and 6 linear and 8 angular measurements were obtained. Paired t-tests were used to evaluate differences and the results showed no statistically significant differences between the two groups. In the discussion, it is hypothesized that, if anchorage is maintained and the extraction space is utilized to alleviate anterior crowding, vertical dimension loss cannot happen. Bishara and coworkers31 (1997) studied two groups of Class II division 1 patients; 46 non-extraction, 45 patients that underwent first premolar extractions, and 35 matched untreated subjects from the Iowa Growth Study. They examined lateral cephalograms and dental casts pretreatment, immediately post-treatment and two years posttreatment and used analysis of variance (ANOVA) and ttests to analyze the various effects. Although the main focus was on soft-tissue profile changes, vertical measurements increased in both treatment groups. These changes were also observed two years later. A study by Komolpis32 compared 32 patients treated with four first premolar extractions with 48 second premolar extraction subjects. A significant between-groups difference in linear measurements, such as ANS-Me, N-Me, Ar-Go was reported. It was suggested that the greater face height in the first premolar group at the end of treatment was perhaps due to the excessive vertical 14 dimension they had pre-treatment. The two groups showed no difference in bite closure as the mandibular plane angle remained constant. This study therefore did not support the wedge effect. Long-term changes in patients treated with or without extractions were examined by Stephens and coworkers.33 Forty patients, distributed evenly between the two groups were analyzed pre- and post-treatment by way of a coordinate system to measure cephalometric changes. Although this study was mainly focused on soft-tissue effects, it is noted that there were no significant differences between the two groups in terms of vertical and horizontal displacement of the cephalometric landmarks. Kim and coworkers34 tested the validity of the wedge hypothesis by comparing hyperdivergent (defined as SN-GoGn greater than 32°and FMA greater than 24°) Class I patients, 27 of which were treated with first premolar extractions and 27 were treated with second premolar extractions. The two groups were matched for crowding. No extraoral appliances, elastics or expanders were used. The maxilla was superimposed on ANS and palatal plane and the structural method was used for the mandible. More mesial movement of maxillary and mandibular molars was evident in the second premolar group, but there were no dif- 15 ferences in angular and proportional dimensions. They concluded that the wedge hypothesis seems invalid. In 2010, Kumari and Fida35 compared 41 nonextraction and 40 first premolar extraction patients by way of cephalograms and study casts. Interestingly, they found initial facial height was greater for the extraction group, hinting that vertical dimension could be an important criterion in the decision making process. Both groups showed similar increases in vertical parameters which were attributed to growth and the extrusive tendency of orthodontic treatment. The results of this study therefore did not support the notion of vertical control by extracting premolars. The following year, Gkantidis and coworkers36 compared two groups of Class II division 1 hyperdivergent patients (SN-GoGn greater than 32 degrees). A group of 29 patients was treated with four first premolar extractions and high-pull headgear and a second group consisting of 28 patients was treated non-extraction with no vertical control. They found no differences between the two groups and concluded that there are limitations to the effects of conventional orthodontics on skeletal pattern, imposed by factors beyond our control such as neuromuscular balance and function. 16 An overview of the reviewed studies can be found in Table 1.1. Table 1.1. Studies analyzing impact of extractions on vertical measurements. Author Year Groups Björk’s method Significant difference Garlington & 1990 23 U4 L5 1991 48 non-ext Yes Logan16 Yamaguchi & Nanda15 Luecke & No 73 extraction* 1992 42 U4 Yes 1993 29 non-ext Yes Johnston19 Luppanapornlarp & Johnston20 Chua et al.21 33 extraction* 1993 89 non-ext No 85 extraction* Aras22 2002 15 U4 L4 No 9 U5 L5 8 U6 L6 Hans et al.23 2006 56 untreated Yes 45 U4 L4 30 U6 L6 Bayirli et al.24 2013 36 untreated 36 U4 L4 17 Yes Table 1.1. Continued Author Year Groups Björk’s method No significant difference Klapper et al.25 1992 15 non-ext No 15 U4 L4 Paguette et al.28 1992 30 non-ext Yes 33 U4 L4 Cusimano et al.27 1993 37 U4 L4 Staggers29 1994 45 non-ext No Bishara et al.31 1997 35 untreated No Yes 46 non-ext 45 U4 L4 Komolpis32 1998 32 U4 L4 Yes 48 U5 L5 Kocadereli30 1999 40 non-ext No 40 U4 L4 Stephens et al.33 2005 20 non-ext No 20 extraction* Kim et al.34 2005 27 U4 L4 No 27 U5 L5 Kumari & Fida35 2010 41 non-ext No 40 U4 L4 Gkantidis et al.36 2011 28 non-ext 29 U4 L4 18 No Table 1.1. Continued Author Year Groups Björk’s method No significant difference Lin et al.26 2012 57 non-ext No non-ext = non-extraction, U4 = upper first premolars, L4 = lower first premolars, U5 = upper second premolars, L5 = lower second premolars U6 = upper first molars, L6 = lower first molars *extraction pattern not specified Overview After reviewing the literature, what becomes evident is a lack of consensus regarding the validity of the wedge hypothesis and several weaknesses in the various study designs and statistical methods. Most of the studies which used structural superimposition20, 23, 24, 27, 28, 32 reported significant differences between the extraction and non-extraction groups, while most of the authors who relied only on surface landmarks22, 25, 29, 30, 34, 35 concluded both groups were similar at the end of treatment. Some did not include a control group treated without extractions.16, 19, 22, 27, 32, 34 More importantly, none of the stud- ies reviewed took treatment time into account, a factor that, if significantly different between extraction and 19 non-extraction groups, raises the problem of differentiating between treatment effects and growth. Chua and coworkers21 examined a large sample of 174 patients and reported a significant difference between non-extraction and four first premolar extraction treatment, but did not utilize structural superimposition and their study therefore is susceptible to the methodological shortcomings due to use of surface landmarks. The same criticism applies to Staggers,29 who, contrary to the previous group, reported similar increases in facial height in both extraction and non-extraction patients. Bishara and coworkers31 detected some differences in vertical measurements, but again structural superimposition was not used and therefore caution must be employed when interpreting these results. Kocadereli30 found similar changes between the non-extraction and the extraction groups, but he too relied on surface landmarks. Aras’22 work differed in that he studied three groups based on extraction pattern. He compared first premolar, second premolar, and first molar extractions and their effect on open bite malocclusions. If the wedge hypothesis is true, then removing first molars should produce a greater effect compared to premolars. Indeed, he found that extracting first molars or second premolars produced a relative reduction in vertical dimensions, but 20 he too can be criticized for relying on surface landmarks and a relatively small sample size. Kim’s34 group found no difference between first and second premolar extractions and while they did use the structural method, they did not compare their groups with a non-extraction control group. No significant difference was reported by Cusimano and coworkers27 between pre- and post-treatment measurements, but there was no control group and the statistical methods used are unclear. One of the most important and well executed studies was the work of Hans’23 group, who found minimal differences between first molar and first premolar extractions. They had two untreated control groups matched for age and gender and used the structural method to superimpose tracings. Some interesting patterns emerge; only Bayirli and coworkers24 found significant differences between nonextraction treatment and four first premolar extractions. Kocadereli,30 Staggers,29 Kumari and Fida,35 Klapper and coworkers,25 and Paquette and coworkers28 reported no differences. Perhaps first premolars are too far forward to allow adequate molar protraction and produce a detectable wedge effect. Studies that included second premolar or first molar extractions produced mixed results. Kim and cowork21 ers34 and Komolpis32 found no difference between first and second premolar extractions. On the contrary, Aras22 and Hans and coworkers23 concluded second premolar and first molar extractions resulted in better vertical control. Although this review is extensive, it is by no means complete. For example, only studies published in English were included. However, the conflicting results and fundamental design flaws of some of the studies presented suffice to highlight the need for further research in order to expand the current body of evidence. It is worth pointing out once more that none of the reviewed studies accounted for potential discrepancies in treatment time between extraction and non-extraction treatment. Prospective, double-blind randomized control trials are the gold standard, but will never be available due to obvious ethical and practical limitations in this type of study. Therefore, the only reasonable alternative for high quality evidence on this subject would be a meta analysis of pooled data from several well designed retrospective and prospective longitudinal studies. 22 Statement of Thesis The present study is an effort to contribute to our knowledge on the effects of orthodontic treatment with and without extractions, focusing on vertical skeletal and dental measurements. Björk’s37, 38 method of super- imposition will be used in order to detect mandibular rotation without relying solely on surface landmarks. What sets this study apart from the ones reviewed is the inclusion of a control group treated without extractions, use of the structural method of superimposition, and the use of analysis of covariance (ANCOVA) in order to control for disparities in treatment time. Because the subjects studied are growing individuals, it is not easy to differentiate between changes brought on by growth and treatment effects. Moreover, it is reasonable to assume that as treatment time increases, so do the effects of treatment and growth. Should treatment time differ significantly between the extraction and non-extraction groups, the results may be skewed by this discrepancy. These issues were not addressed by any of the reviewed articles and a study designed accordingly may provide more reliable answers. 23 Literature Cited 1. Angle EH. Treatment of Malocclusion of the Teeth. 7th ed. Philadelphia: The S.S. White Dental Manufacturing Co.; 1907. 2. Angle EH. Cosmos. Classification of malocclusion. 1899;41:248-64. Dent 3. Rinchuse DJ, Rinchuse DJ. Ambiguities of Angle’s classification. Angle Orthod. 1989;59:295-8. 4. Sassouni V, Nanda S. Analysis of dentofacial vertical proportions. Am J Orthod. 1964;50:801-23. 5. Sassouni V. A classification of skeletal facial types. Am J Orthod. 1969;55:109-23. 6. Schudy FF. Vertical Growth Versus Anteroposterior Growth As Related To Function And Treatment. Angle Orthod. 1964;34:75-93. 7. Fields HW, Proffit WR, Nixon WL, Phillips C, Stanek E. Facial pattern differences in long-faced children and adults. Am J Orthod. 1984;85:217-23. 8. Cangialosi TJ. Skeletal morphologic features of anterior open bite. Am J Orthod. 1984;85:28-36. 9. Nanda SK. Patterns of vertical growth in the face. Am J Orthod Dentofac Orthop. 1988;93:103-16. 10. Schudy FF. The control of vertical overbite in clinical orthodontics. Angle Orthod. 1968;38:1939. 11. Nasby JA, Isaacson RJ, Worms FW, Speidel TM. Orthodontic extractions and the facial skeletal pattern. Angle Orthod. 1972;42:116-22. 12. Joondeph DR, Riedel RA. Second premolar serial extraction. Am J Orthod Dentofac Orthop. 1976;69:169-84. 13. Pearson LE. Vertical control in treatment of patients having backward-rotational growth tendencies. Angle Orthod. 1978;48:132-40. 24 14. Ketterhagen DH. First premolar or second premolar extractions: formula or clinical judgment? Angle Orthod. 1979;49:190-8. 15. Yamaguchi K, Nanda RS. The effects of extraction and nonextraction treatment on the mandibular position. Am J Orthod Dentofac Orthop. 1991;100:443-52. 16. Garlington M, Logan LR. Vertical changes in high mandibular plane cases following enucleation of second premolars. Angle Orthod. 1990;60:263-7. 17. Björk A. Variations in the growth pattern of the human mandible: longitudinal radiographic study by the implant method. J Dent Research. 1963;42(1)Pt 2:400-11. 18. 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. 19. Luecke PE, III, Johnston LE, Jr. The effect of maxillary first premolar extraction and incisor retraction on mandibular position: testing the central dogma of "functional orthodontics". Am J Orthod Dentofac Orthop. 1992;101:4-12. 20. Luppanapornlarp S, Johnston LE, Jr. The effects of premolar-extraction: a long-term comparison of outcomes in "clear-cut" extraction and nonextraction Class II patients. Angle Orthod. 1993;63:257-72. 21. Chua AL, Lim JY, Lubit EC. The effects of extraction versus nonextraction orthodontic treatment on the growth of the lower anterior face height. Am J Orthod Dentofac Orthop. 1993;104:361-8. 22. Aras A. Vertical changes following orthodontic extraction treatment in skeletal open bite subjects. Eur J Orthod. 2002;24:407-16. 23. Hans MG, Groisser G, Damon C, Amberman D, Nelson S, Palomo JM. Cephalometric changes in overbite and vertical facial height after removal of 4 first molars or first premolars. Am J Orthod Dentofac Orthop. 2006;130:183-8. 24. Bayirli B, Vaden JL, Johnston LE, Jr. Original article: Long-term mandibular skeletal and dental effects of standard edgewise treatment. Am J Orthod Dentofac Orthop. 2013;144:682-90. 25 25. Klapper L, Navarro SF, Bowman D, Pawlowski B. The influence of extraction and nonextraction orthodontic treatment on brachyfacial and dolichofacial growth patterns. Am J Orthod Dentofac Orthop. 1992;101:425-30. 26. Lin CH, Short LL, Banting DW. The effects of nonextraction orthodontic treatment on the vertical dimension: a comparison of a dolichofacial and a mesofacial group. Aus Orthod J. 2012;28:37-43. 27. Cusimano C, McLaughlin RP, Zernik JH. Effects of first bicuspid extractions on facial height in highangle cases. J Clin Orthod. 1993;27:594-8. 28. Paquette DE, Beattie JR, Johnston LE, Jr. A longterm comparison of nonextraction and premolar extraction edgewise therapy in "borderline" Class II patients. Am J Orthod Dentofac Orthop. 1992;102:114. 29. Staggers JA. Vertical changes following first premolar extractions. Am J Orthod Dentofac Orthop. 1994;105:19-24. 30. Kocadereli İ. The effect of first premolar extraction on vertical dimension. Am J Orthod Dentofac Orthop. 1999;116:41-5. 31. Bishara SE, Cummins DM, Zaher AR. Treatment and posttreatment changes in patients with Class II, division 1 malocclusion after extraction and nonextraction treatment. Am J Orthod Dentofac Orthop. 1997;111:18-27. 32. Komolpis R. Cephalometric comparison between first premolar and second premolar extraction. Ann Arbor, MI: University of Michigan; 1998. 33. Stephens CK, Boley JC, Behrents RG, Alexander RG, Buschang PH. Long-term profile changes in extraction and nonextraction patients. Am J Orthod Dentofac Orthop. 2005;128:450-7. 34. Kim TK, Kim JT, Mah J, Yang WS, Baek SH. First or second premolar extraction effects on facial vertical dimension. Angle Orthod. 2005;75:177-82. 35. Kumari M, Fida M. Vertical facial and dental arch dimensional changes in extraction vs. non-extraction orthodontic treatment. J Coll Physicians Surg Pak. 2010;20:17-21. 26 36. Gkantidis N, Halazonetis DJ, Alexandropoulos E, Haralabakis NB. Treatment strategies for patients with hyperdivergent Class II Division 1 malocclusion: is vertical dimension affected? Am J Orthod Dentofac Orthop. 2011;140:346-55. 37. Björk A, Skieller V. Normal and abnormal growth of the mandible. A synthesis of longitudinal cephalometric implant studies over a period of 25 years. Eur J Orthod. 1983;5:1-46. 38. Björk A, Skieller V. Postnatal growth and development of the maxillary complex. Ann Arbor, MI: The University of Michigan; 1976. 27 CHAPTER 2: JOURNAL ARTICLE Abstract Introduction: One of the most important steps in orthodontic treatment planning is the extraction decision. The “wedge effect” hypothesizes that extracting posterior teeth allows the mandible to hinge closed, thereby allowing better vertical control. The objective of the present study was to compare the different outcomes between patients treated non-extraction and patients treated with extractions, in order to assess the validity of this concept. Subjects and Methods: Pre- and post-treatment cephalograms of 191 normal, healthy subjects (67 treated without extractions, 61 treated with four first premolar extractions and 63 treated with four second premolar extractions) were obtained from the archive of the Center for Advanced Dental Education orthodontic clinic of Saint Louis University. Class I, Class II and Class III subjects were similarly distributed across the three groups. Tracings were superimposed according to Björk’s method for structural superimposition and fiducial lines were drawn for the maxilla and mandible. A Cartesian system was used to measure vertical and horizontal changes of ANS, upper first molar and central incisor, Gonion, Menton, lower first molar and lower cen- 28 tral incisor. Overbite, as well as SN-GoMe, SN-FOP and the angle between the maxillary and mandibular fiducial lines was also measured. Because the different treatment protocols resulted in significantly different treatment times, analysis of covariance was performed to test for different increments of change among the 3 groups while controlling for the effect of treatment time. Results: There was a statistically significant difference in horizontal displacement of upper first and lower first molars between the 3 groups. The lower incisors showed significantly different amounts of extrusion for each group. There were no statistically significant differences in amounts of change in any of the other variables. Conclusions: The present study failed to produce evidence of favorable vertical changes as a result of extractions. Other than lower incisor vertical displacement, there was no difference in magnitude of change in vertical measurements between the 3 groups. 29 Introduction The issue of vertical control has been a subject of discussion for several decades, since Sassouni and Nanda1 introduced their analysis. Schudy2 argued in 1968 that extractions may be considered for better vertical control, even in cases with sufficient space. Nasby and coworkers3 suggested that using the extraction space to move the posterior teeth forward allows the mandible to rotate closed therefore reducing anterior facial height. Pearson4 examined a group of hyperdivergent patients treated with first premolar extractions and attributed the bite closure he observed to the mesial drift of the posterior teeth “out of the wedge,” and so the wedge hypothesis was born. As a concept, it had a sound geometric basis and could be the key to some of the most challenging malocclusions. Logic, however, can sometimes be truth’s biggest enemy. Several investigators have examined the validity of this theory, but the review of the literature reveals a lack of consensus and conflicting results that seem to grow from technical differences. Some authors report significant differences between extraction and nonextraction orthodontic treatment,5-11 while others failed to detect significant differences.12-25 30 The gold standard in assessing changes brought on by growth and treatment is the fruit of Björk’s26-28 implant study, which identified structures that were stable enough to use for superimposition. Despite the fact that Björk’s study predates most of the reviewed articles, not all the authors opted to use the structural method for their own research, instead exposing their analysis to the limitations of relying on surface landmarks. If the wedge hypothesis is true, it may be difficult to validate by comparing first and second premolar extractions alone because their antero-posterior position is similar and may only produce minimal differences in bite closure. A sample including a first or second molar extraction group may be more useful in that regard, but given the scarcity of such cases it was not possible to obtain for this study. A non-extraction group is necessary in both scenarios in order to draw any meaningful conclusions regarding the benefits of extractions in terms of better vertical control. Finally, treatment time is an important factor that must be accounted for when trying to interpret any observed changes. It is reasonable to assume that as treatment time increases, so do the effects of treatment and growth. Should treatment time differ significantly between the extraction and non-extraction groups, results 31 may be difficult to compare and selecting the appropriate statistical analysis is of utmost importance. The present study is an effort to address some of the abovementioned shortcomings in methods and design in order to provide more reliable answers and contribute to the existing body of evidence. Subjects and Methods Sample For the present retrospective study, the records of 191 patients were obtained from the archive of the Center for Advanced Dental Education of Saint Louis University in Saint Louis, Missouri. The sample consisted of 67 patients treated without extractions (labeled “nonextraction” in tables and figures, mean age 13.18±2.51 years), 61 patients treated with four first premolar extractions (labeled “4s” in tables and figures, mean age 13.86±4.05 years) and 63 patients treated with four second premolar extractions (labeled “5s” in tables and figures, mean age 14.05±4.13 years). Inclusion criteria were availability of pre- and post-treatment lateral cephalograms of sufficient quality and extraction patterns that fell in one of the aforementioned categories. Patients that underwent orthognathic surgery and patients with craniofacial syndromes or missing teeth (other than third 32 molars) were excluded from the study. Patients with congenitally missing second premolars were included if the primary second molars were still present at the beginning of treatment. Table 2.1. Distribution of patients by Angle classification in each extraction group. Non-extraction 4s 5s Class I 47 40 41 Class II 19 21 20 Class III 1 1 1 Cephalometric Analysis Pre- and post-treatment lateral cephalograms were hand traced on 0.003” acetate by the principal investigator (V.C.) according to the guidance of a co-worker (L.E.J.) who subsequently checked a subset of series. In order to ensure consistency and reliability, tracings for each subject were done side by side. Bilateral structures were traced by bisecting the distance between the two outlines. Arbitrary maxillary and mandibular fiducial lines were drawn on one tracing, then transferred to the other using Bjork’s29 method of structural superimposition. Figure 2.1 shows the anatomical structures and landmarks traced as well as the Cartesian system used for horizontal and vertical measurements. 33 Maxillary superimposition was done first using the anterior surface of the zygomatic process28 and overbite pre-and post-treatment was measured perpendicular to the functional occlusal plane. Subsequently, a line was drawn averaging each series’ Sella-Nasion, from which 7 degrees were subtracted to draw a horizontal line which served as the x-axis of a Cartesian system. The following structures were projected on the y-axis and their vertical distance from the x-axis were measured and recorded: 1. Anterior Nasal Spine (ANS) 2. Upper First Molar (U6) 3. Upper Central Incisor (U1) 4. Gonion (Go) 5. Menton (Me) There was a single projection on the x-axis: 6. Upper First Molar (U6) 34 Angular variables measured are listed below: 7. Sella-Nasion to Gonion-Menton (SNGoMe) 8. Sella-Nasion to Functional Occlusal Plane 9. Structural Rotation, measured using the fiducial lines. Figure 2.1. Measurements based on maxillary superimposition. 35 The two tracings were then superimposed on the mandible using the inner cortex of the inferior border of the symphysis, the mandibular canal and the third molar germ if present.29 A line was drawn averaging each series’ Gonion-Menton, which served as the x-axis. This was done to measure dental changes within the mandible. As above, the projections of the following structures on the coordinate system were measured: 10. Lower First Molar (L6) 11. Lower Central Incisor (L1) Figure 2.2. Measurements based on mandibular superimposition. 36 Electronic digital calipers were used for linear measurements and angular measurements were done with a cephalometric protractor. Age was converted to decimal years by way of an online calculator.30 Treatment time was measured in days calculated by a different online calculator.31 Differences between pre-and post-treatment measurements were calculated using Microsoft Excel to subtract the T1 from the T2 value. A positive value for change designates a downward displacement for vertical variables and a forward displacement for horizontal variables. Error Study To assess reliability of measurements, subjects were assigned numbers from 1 to 191, which numbers were then used to generate a random set of 20 patients based on www.randomizer.org.32 The cephalograms of those patients were then retraced and remeasured to test for intra-observer reliability, which showed a high level of consistency, as determined by Cronbach’s alpha. The lowest value was .945, for lower incisor vertical displacement. Statistical Analysis Descriptive statistics were calculated for each group pre-and post-treatment. Analysis of variance was 37 performed to test for significant differences in SN-GoMe between the 3 groups pre-treatment and to compare the mean treatment times. Because treatment time was significantly different between the non-extraction and the extraction groups, analysis of covariance was selected as the most appropriate statistical test. The T2-T1 differences were calculated, and ANCOVA was performed to compare the 3 groups with treatment time as the covariate. Because certain assumptions of ANCOVA were violated, such as linearity of measurements and normality of distribution, the results were confirmed with ANOVA and KruskalWallis. Tukey’s alpha and Mann-Whitney U test was performed to identify where the differences lay between the 3 groups. All analyses were done using SPSS (version 22.0, SPSS Inc., Chicago, IL). 38 Results Descriptive statistics for treatment time are included in the following table. Table 2.2. Means and standard deviations for treatment time in days. Non-extraction (n=67) Mean SD 745.1 329.6 4s (n=61) Mean SD 939 302.2 5s (n=63) Mean SD 1015.4 299.3 Analysis of variance revealed statistically significant differences in treatment time between the 3 groups (F(2,188)=13.096, p<.001). Post-hoc analysis was performed to identify where the differences lay among the 3 groups. The results are shown in table 2.2. The nonextraction group displayed significantly reduced treatment time compared to the extraction groups. The differences in mean treatment time between the first and second premolar groups were not statistically significant. 39 Table 2.3. Post-hoc analysis for treatment time between groups. Contrast p Non-extraction 4s .002 Non-extraction 5s <.001 4s 5s .360 Descriptive statistics for pre- and post- treatment cephalometric values are summarized in tables 2.3 and 2.4. Descriptive statistics for change are summarized in table 2.5. The only variables showing significantly different extent of change among the 3 groups are upper molar horizontal displacement, lower molar horizontal displacement and lower incisor vertical displacement. Table 2.6 shows adjusted means for change after controlling for treatment time by way of ANCOVA. 40 Table 2.4. Descriptive statistics pre-treatment. Non-extraction (n=67) Mean SD OB(mm) 3.4 1.9 4s (n=61) Mean SD 3.0 2.1 5s (n=63) Mean SD 3.2 1.7 ANS Y-axis (vertical distance in mm) 44.9 3.4 44.9 3.9 44.8 3.1 U6 65.1 4.6 65.2 4.7 65.2 4.5 U1 74.7 5.1 75.1 5.2 75.1 4.4 Gonion 74.2 5.8 73.9 5.9 75.5 7.3 Menton 109.9 6.7 111.8 7.6 111.4 8.6 L6 26.5 3.1 26.8 3.2 26.6 3.0 L1 38.9 3.9 40.7 3.5 40.1 3.9 U6 X-axis (horizontal distance in mm) 37.5 5.5 38.4 5.2 39.4 4.1 L6 39.8 5.1 41.5 4.9 Angular measurements 40.4 4.4 SN-FOP 19.2 4.3 19.0 4.2 18.4 4.3 SNGoMe 36.2 5.5 38.4 4.7 36.9 5.3 FOP = Functional Occlusal Plane 41 Table 2.5. Descriptive statistics post-treatment. Non-extraction (n=67) Mean SD OB(mm) 1.9 0.7 4s (n=61) Mean SD 1.6 0.7 5s (n=63) Mean SD 1.87 0.8 ANS Y-axis (vertical distance in mm) 46.3 3.6 46.3 3.9 46.5 3.4 U6 67.8 4.5 68.1 4.8 68.5 4.5 U1 76.7 5.1 77.2 5.3 76.9 6.3 Gonion 78.2 5.9 78.1 6.8 79.5 7.3 Menton 115.1 7.2 116.8 8.0 116.5 9.3 L6 28.9 2.9 29.8 3.1 29.5 2.9 L1 40.7 4.0 41.9 3.6 40.7 3.6 U6 X-axis (horizontal distance in mm) 39.1 5.4 41.2 5.3 42.8 4.2 L6 40.7 4.9 43.7 4.7 Angular measurements 43.5 4.3 SN-FOP 18.8 4.1 19.1 3.7 18.6 4.0 SNGoMe 36.7 5.5 38.4 4.8 37.1 5.4 FOP = Functional Occlusal Plane 42 Table 2.6. Descriptive statistics for increments of change. Non-extraction (n=67) Mean SD OB(mm) ANS 4s (n=61) Mean SD 5s (n=63) Mean SD -1.46 1.78 -1.40 2.03 -1.32 Y-axis (vertical changes in mm) 1.44 1.05 1.40 1.04 1.71 p 1.63 ns 1.18 ns U6 2.69 1.65 2.87 1.69 3.30 2.21 ns U1 1.99 1.73 2.15 1.91 1.94 4.07 ns Gonion 3.91 2.47 4.11 2.56 3.98 2.76 ns Menton 5.18 2.91 5.04 3.11 5.14 3.24 ns L6 2.41 1.24 2.98 1.62 2.92 1.53 ns L1 2.10 * U6 1.74 1.18 1.24 1.34 0.62 X-axis (horizontal changes in mm) 1.56 1.38 2.77 1.36 3.40 1.85 ** L6 0.92 SN-FOP 3.03 1.41 ** -0.36 0.65 2.17 1.30 Angular changes 3.10 0.07 3.61 0.22 3.12 ns SN-GoMe 0.54 2.98 0.01 1.66 0.19 1.86 ns StRot -0.22 2.06 -0.53 2.03 0.06 2.70 ns ns, p>.05; *p<.05; **p<.001 FOP = Functional Occlusal Plane; StRot = Structural Rotation, measured using fiducial lines. 43 Table 2.7. Descriptive statistics for increments of change adjusted for treatment time according to ANCOVA. OB(mm) Non-extraction (n=67) Mean SE 4s (n=61) Mean SE -1.51 -1.38 0.23 p -1.28 0.23 ns ANS Y-axis (vertical changes in mm) 1.64 0.13 1.34 0.13 1.55 0.13 ns U6 3.03 0.22 2.77 0.22 3.04 0.22 ns U1 2.16 0.35 2.11 0.35 1.81 0.35 ns Gonion 4.27 0.31 4.01 0.32 3.70 0.32 ns Menton 5.78 0.36 4.88 0.36 4.67 0.36 ns L6 2.63 0.17 2.91 0.17 2.74 0.18 ns L1 0.19 * U6 1.94 0.19 1.19 0.19 0.46 X-axis (horizontal changes in mm) 1.69 0.19 2.74 0.19 3.30 0.19 ** L6 0.95 0.14 0.14 3.02 0.15 ** SN-FOP -0.44 Angular changes 0.41 0.09 0.42 0.28 0.42 ns SN-GoMe 0.57 0.29 0.00 0.29 0.16 0.29 ns StRot -0.13 0.29 -0.55 0.29 0.00 0.29 ns 2.17 0.23 5s (n=63) Mean SE ns, p>.05; *p<.05; **p<.001 FOP = Functional Occlusal Plane; StRot = Structural Rotation, measured using fiducial lines. 44 ANCOVA The increased treatment time for the extraction groups must be considered when analyzing the effects of treatment and growth. In order to eliminate the differences that may be caused by this discrepancy, ANCOVA was performed to control for treatment time. There was a significant effect of different extraction patterns on the horizontal displacement of the upper first molars after controlling for the effect of treatment time, F(2,187)=16.208, p<.001. Post-hoc analysis revealed that there was significantly more horizontal displacement in the four first premolar extraction group (2.77 ± 1.36mm, p=.001) and the second premolar extraction group (3.4 ± 1.85mm, p<.001) compared to the nonextraction group (1.56 ± 1.38mm). There were no statistically significant differences between the first and second premolar extraction groups. There was also a significant effect of different extraction patterns on the horizontal displacement of the lower first molars after controlling for the effect of treatment time, F(2,187)=45.837, p<.001. Post-hoc analysis revealed significantly greater horizontal displacement in the four first premolar group (2.17 ± 1.3mm, p<.001) and the four second premolar group (3.03 ± 1.41mm, p<.001) compared to the non-extraction group (0.92 ± 0.65mm). 45 A significant effect was also observed for lower incisor vertical displacement, F(2,187)=13.308, p<.001. Post-hoc tests showed statistically significantly greater extrusion in the non-extraction group (1.74 ± 1.18mm, p<.001) compared to the four second premolar extraction group (0.62 ± 2.1mm). No other variables showed statistically significant differences when controlling for treatment time. Certain assumptions of ANCOVA, however, were violated. The dependent variables were not normally distributed for all groups. Homogeneity of variances was also not equal for all variables according to the Levene statistic, and treatment time was not linearly related to the dependent variables for each group according to the scatterplots. To confirm the results of the ANCOVA, ANOVA was performed as well as Kruskall-Wallis one-way analysis of variance. Both tests and their post-hoc analyses (Tukey’s post-hoc analysis and Mann-Whitney U test) confirmed the differences detected by ANCOVA. The findings are summarized in the appendix. 46 Discussion Limitations In order to be able to better interpret the results of the present study, it is important to acknowledge certain limitations. As with any retrospective study, it is difficult to guard against selection bias, given that individuals are not randomly assigned to each group. Depending on their initial vertical measurements, they may be more likely to be treated with different approaches, creating a biased measure of association. However, there were no statistically significant differences between group means for initial SNGoMe as shown by one-way ANOVA (F(2,188)=2.844, p=.061). There were also no statistically significant differences between group means for initial overbite (F(2,188)=.513, p=.60). Exposure identification bias is not a concern, since misclassification was avoided easily by referring to patient charts and extraction orders. This study is also relatively immune from recall bias, as there is no reason to assume a patient treated with extractions has a higher or lower probability of completing orthodontic treatment versus one that was treated non-extraction. A long-term recall of patients treated with different extraction protocols may be less accurate in de- 47 termining the validity of the wedge hypothesis, due to recall but also temporal bias. The present study measured subjects immediately post-treatment and therefore is advantageous in that regard. Interviewer bias is not easy to guard against, since a single investigator classified the subjects, traced their radiographs and measured the tracings. In an effort to minimize the risk, however, there were no identifiers on the tracings regarding each patient’s extraction pattern. It should be noted that the subjects in the present study were treated by numerous clinicians over a span of 3 decades. A variety of appliances and treatment modalities were used; however, it was not the purpose of this investigation to evaluate the effects of each protocol. In fact, controlling for different mechanics and use of directional forces can be very challenging. For example, it would require a reliable method for assessing compliance with use of headgear or intermaxillary elastics. Another factor not taken into account was the amount of crowding in each case. If anchorage is maintained and the extraction space is used to relieve anterior crowding or reduce protrusion, one would expect less mesial movement of molars. However, the results of the 48 present study for horizontal displacement of upper and lower molars are similar to those reported by other authors.20,21 It was determined that the most efficient design for the present investigation would be to group subjects only by extraction pattern, disregarding factors such as Angle classification, amount of crowding, and treatment mechanics. Instead, treatment time was selected as the most important differentiating factor. Extraction groups showed significantly longer treatment times compared to the non-extraction group, which could produce greater changes due to growth thus making the results difficult to interpret. The Wedge Hypothesis As expected, molars in the 3 groups showed different increments of mesial displacement. If the wedge hypothesis is true, this should result in different amounts of change in vertical dental and skeletal measurements. However, the only vertical measurement significantly different between the 3 groups was the amount of lower incisor extrusion. Mandibular rotation was assessed by measuring the SN-GoMe angle before and after treatment and also by using fiducial lines. This was done to examine basal bone rotation as well as possible compensatory surface changes that might serve to mask it by keeping 49 SN-GoMe relatively unchanged. Both showed similar increments of change. Results Compared to Literature The results of the present study are in agreement with those reported by some authors14-25 but contradict others that reported significant vertical changes.5-11 Table 2.8 summarizes findings from similar studies. Molar displacement in the first premolar extraction group is similar to the findings of Paquette and coworkers,13 Kim and coworkers,20 and Komolpis.25 On the contrary, Aras8 reported higher values, as one might expect due to the fact that one of the inclusion criteria for his study was use of most of the extraction space for forward movement of molars. Cusimano and coworkers,15 too, showed more molar displacement, but no error study was conducted and their sample was relatively small. Klapper and coworkers23 also found a significantly higher value, which could be due to their methodology. They used the distal contact point of the upper first molar (possibly more sensitive to the tipping that could occur during space closure) to a vertical drawn from the pterygomaxillary fissure. It is also worth pointing out that no error study was reported. 50 For molar displacement in the second premolar extraction group, the findings of the present study are similar to those of Aras,8 and Kim and coworkers.20 Komolpis25 reported a slightly higher value for the lower molars, which could be due to the fact that sliding mechanics with elastic chains were used for space closure. An interesting observation worth analyzing is the finding that upper molars displaced forward slightly more than the lower molars. One might expect identical values for both, given that most of the subjects studied were Class I. The reason behind this discrepancy may be growth; as the mandible outgrows the maxilla in the present sample consisting mainly of growing patients, the upper molars move forward more within the maxillary alveolar bone than the lowers in order to maintain intercuspation. Another important point to consider is the methodology used, which examined molar displacement relative to basal bone. If changes were measured along the functional occlusal plane the values would probably have been different; however, it is not within the scope of the present investigation to analyze the effects of growth in an antero-posterior direction. In terms of vertical changes, anterior face height increased by almost identical increments in all three groups. For the non-extraction group, Staggers,16 51 Luppanapornlarp and Johnston,14 Kumari and Fida21 and Kocadereli17 reported similar values. The slightly larger increase detected by Paquette and coworkers13 could be due to sample selection, which only included Class II, division 1 patients. If Class II elastics were used, molar extrusion could explain this discrepancy. In the four first premolar group, Aras,8 Staggers,16 Paquette and coworkers,13 and Kocadereli17 show results similar to those of the present study. Kumari and Fida21 reported significantly less increase in anterior face height, which could be explained due to the fact that their extraction group had a greater pre-treatment value. In the four second premolar group, Aras8 and Komolpis25 showed similar increases in anterior face height that agree with the findings of this investigation. Mandibular plane angle showed similar magnitude of change in all reviewed studies and the results of the present investigation are in agreement with the literature. 52 Table 2.8. Results compared to similar studies. Non-extraction 4s 5s Menton (mm) 5.18 U6H (mm) 1.56 L6H (mm) 0.92 MPA (°) -0.22 Menton (mm) 5.04 U6H (mm) 2.77 L6H (mm) 2.17 MPA (°) 0.01 Menton (mm) 5.14 U6H (mm) 3.40 L6H (mm) 3.03 MPA (°) 0.19 Aras8 . . . . . . . . 5.40 3.66 4.14 -0.20 4.00 3.89 3.72 -1.06 Komolpis25 . . . . . . . . 7.02 3.11 2.10 0.24 4.80 3.92 3.66 0.05 Kim et al.20 . . . . . . . . . . 2.72 2.14 0.56 . . 3.84 3.62 0.52 Cusimano et al.15 . . . . . . . . . . 5.00 4.10 0.40 . . . . . . . . Paquette et al.13 7.00 1.00 0.10 -2.00 5.40 2.50 3.30 0.30 . . . . . . . . Luppanaporlarp & Johnston14 5.00 0.80 . . 0.70 . . . . . . . . . . . . . . . . Kocadereli17 5.98 . . . . 0.27 4.67 . . . . 0.10 . . . . . . . . Kumari & Fida21 5.50 . . . . 0.60 1.50 . . . . -0.20 . . . . . . . . Klapper et al.23 . . 1.27 . . . . . . 4.43 . . . . . . . . . . . . Staggers16 5.08 . . . . 0.14 5.38 . . . . 0.11 . . . . . . . . . . . . . . . . 7.20 . . . . . . . . . . . . . . Author(s) Present study Bayirli et al. 10 U6H = Upper 6 horizontal displacement, L6H = Lower 6 horizontal displacement, MPA = Mandibular plane angle 53 Overview and Future Research The present study examined a sample larger than any of its predecessors mentioned in the literature review. Even so, it failed to produce evidence that would indicate a forward rotation of the mandible and better vertical control for the extraction groups. However, this does not mean orthodontists are powerless when it comes to managing problems in the vertical dimension. Instead, it should perhaps serve as an indication that forward mandibular rotation as a result of moving posterior teeth forward into the extraction spaces may not be a realistic expectation. Using directional extra-oral forces33 or temporary anchorage devices to intrude or minimize extrusion of posterior teeth34 may be a less aggressive and more effective approach. Schudy35 argued that, in deep-bite patients, levelling should be done by extruding molars due to the inherent instability in extrusive or intrusive movement of upper and lower incisors35 but presented no evidence to support this statement. This argument has been disproven by other authors.36, 37 Managing incisors in the vertical direction by using extra-oral forces, temporary anchorage devices, or intermaxillary elastics are reasonable approaches and should not be discounted, although more research is needed to determine which modality provides 54 greatest long-term stability. Surgical solutions, such as maxillary impactions or sub-apical osteotomies may also be indicated for certain patients. In summary, the dentofacial complex apparently does not behave like a hinge. Addressing vertical dimension problems should take all factors into account, including facial soft tissue and skeletal proportions, growth potential, eruption patterns, habits, lip length, gingival display, Angle classification and crowding. There is a significant body of evidence against the wedge hypothesis and it is in the best interest of our profession and our patients to acknowledge it rather than anachronistic and unproven concepts. 55 Conclusions The present investigation failed to detect differences in amounts of vertical change between extraction and non-extraction orthodontic treatment. As expected, first molars showed greater mesial displacement in the four second premolar extraction group, followed by the four first premolar extraction group and the non-extraction group. However, the alleged counterclockwise rotation of the mandible and subsequent bite closure was not evident. Treatment decisions should be based on all available diagnostic parameters and the alleged improved vertical control does not suffice as justification for extractions in absence of other discrepancies, such as crowding or dental and soft tissue protrusion. 56 Literature Cited 1. Sassouni V, Nanda S. Analysis of dentofacial vertical proportions. Am J Orthod. 1964;50:801-23. 2. Schudy FF. The control of vertical overbite in clinical orthodontics. Angle Orthod. 1968;38:1939. 3. Nasby JA, Isaacson RJ, Worms FW, Speidel TM. Orthodontic extractions and the facial skeletal pattern. Angle Orthod. 1972;42:116-22. 4. Pearson LE. Vertical control in treatment of patients having backward-rotational growth tendencies. Angle Orthod. 1978;48:132-40. 5. Garlington M, Logan LR. Vertical changes in high mandibular plane cases following enucleation of second premolars. Angle Orthod. 1990;60:263-7. 6. Luecke PE, III, Johnston LE, Jr. The effect of maxillary first premolar extraction and incisor retraction on mandibular position: testing the central dogma of "functional orthodontics". Am J Orthod Dentofac Orthop. 1992;101:4-12. 7. Chua AL, versus growth Orthod Lim JY, Lubit EC. The effects of extraction nonextraction orthodontic treatment on the of the lower anterior face height. Am J Dentofac Orthop. 1993;104:361-8. 8. Aras A. Vertical changes following orthodontic extraction treatment in skeletal open bite subjects. Eur J Orthod. 2002;24:407-16. 9. Hans MG, Groisser G, Damon C, Amberman D, Nelson S, Palomo JM. Cephalometric changes in overbite and vertical facial height after removal of 4 first molars or first premolars. Am J Orthod Dentofac Orthop. 2006;130:183-8. 10. Bayirli B, Vaden JL, Johnston LE, Jr. Original article: Long-term mandibular skeletal and dental effects of standard edgewise treatment. Am J Orthod Dentofac Orthop. 2013;144:682-90. 11. Yamaguchi K, Nanda RS. The effects of extraction and nonextraction treatment on the mandibular position. Am J Orthod Dentofac Orthop. 1991;100:443-52. 57 12. Gianelly AA, Anderson CK, Boffa J. Longitudinal evaluation of condylar position in extraction and nonextraction treatment. Am J Orthod Dentofac Orthop. 1991;100:416-20. 13. Paquette DE, Beattie JR, Johnston LE, Jr. A longterm comparison of nonextraction and premolar extraction edgewise therapy in "borderline" Class II patients. Am J Orthod Dentofac Orthop. 1992;102:114. 14. Luppanapornlarp S, Johnston LE, Jr. The effects of premolar-extraction: a long-term comparison of outcomes in "clear-cut" extraction and nonextraction Class II patients. Angle Orthod. 1993;63:257-72. 15. Cusimano C, McLaughlin RP, Zernik JH. Effects of first bicuspid extractions on facial height in highangle cases. J Clin Orthod. 1993;27:594-8. 16. Staggers JA. Vertical changes following first premolar extractions. Am J Orthod Dentofac Orthop. 1994;105:19-24. 17. Kocadereli İ. The effect of first premolar extraction on vertical dimension. Am J Orthod Dentofac Orthop. 1999;116:41-5. 18. Bishara SE, Cummins DM, Zaher AR. Treatment and posttreatment changes in patients with Class II, division 1 malocclusion after extraction and nonextraction treatment. Am J Orthod Dentofac Orthop. 1997;111:18-27. 19. Stephens CK, Boley JC, Behrents RG, Alexander RG, Buschang PH. Long-term profile changes in extraction and nonextraction patients. Am J Orthod Dentofac Orthop. 2005;128:450-7. 20. Kim TK, Kim JT, Mah J, Yang WS, Baek SH. First or second premolar extraction effects on facial vertical dimension. Angle Orthod. 2005;75:177-82. 21. Kumari M, Fida M. Vertical facial and dental arch dimensional changes in extraction vs. non-extraction orthodontic treatment. J Coll Physicians Surg Pak. 2010;20:17-21. 22. Lin CH, Short LL, Banting DW. The effects of nonextraction orthodontic treatment on the vertical dimension: a comparison of a dolichofacial and a mesofacial group. Aus Orthod J. 2012;28:37-43. 58 23. Klapper L, Navarro SF, Bowman D, Pawlowski B. The influence of extraction and nonextraction orthodontic treatment on brachyfacial and dolichofacial growth patterns. Am J Orthod Dentofac Orthop. 1992;101:425-30. 24. Gkantidis N, Halazonetis DJ, Alexandropoulos E, Haralabakis NB. Treatment strategies for patients with hyperdivergent Class II Division 1 malocclusion: is vertical dimension affected? Am J Orthod Dentofac Orthop. 2011;140:346-55. 25. Komolpis R. Cephalometric comparison between first premolar and second premolar extraction. Ann Arbor, MI: University of Michigan; 1998. 26. Björk A. Prediction of mandibular growth rotation. Am J Orthod. 1969;55:585-99. 27. Björk A, Skieller V. Normal and abnormal growth of the mandible. A synthesis of longitudinal cephalometric implant studies over a period of 25 years. Eur J Orthod. 1983;5:1-46. 28. Björk A, Skieller V. Postnatal growth and development of the maxillary complex. Ann Arbor, MI: The University of Michigan; 1976. 29. Björk A. Variations in the growth pattern of the human mandible: longitudinal radiographic study by the implant method. J Dent Research. 1963;42(1)Pt 2:400-11. 30. Calculate age in decimal years [September 5, 2014]. Available from: http://www.ucsdbglab.org/tools/age.asp. 31. Calculate duration between two dates [September 4, 2014]. Available from: http://www.timeanddate.com/date/duration.html. 32. Instantly generate random numbers [September 10, 2014]. Available from: http://www.randomizer.org/. 33. MacGilpin DH, Araujo EA, Behrents RG, Rowan KB. Spatial changes in the relationship of the mandible and maxilla with different extraction patterns and techniques. Angle Orthod. 2011;81:584-91. 59 34. Sherwood KH, Burch JG, Thompson WJ. Closing anterior open bites by intruding molars with titanium miniplate anchorage. Am J Orthod Dentofac Orthop. 2002;122:593-600. 35. Schudy FF. Sound biologic concepts in orthodontics. Am J Orthod. 1973;63:376-97. 36. Lo FM, Shapiro PA. Effect of presurgical incisor extrusion on stability of anterior open bite malocclusion treated with orthognathic surgery. Int J Adult Orthod Orthog Surg. 1998;13:23-34. 37. Varlik KS, Alpakan OO, Turkoz C. Deepbite correction with incisor intrusion in adults: a long-term cephalometric study. Am J Orthod Dentofacial Orthop. 2013;144:414-9. 60 Appendix ANOVA One-way ANOVA confirmed that there was a significant difference in horizontal displacement of upper first molars between groups, F(2,188)=23.576, p<.001. A Tukey’s post-hoc test revealed that there was significantly more horizontal displacement of upper first molars in the four first premolar extraction group (2.77 ± 1.36mm, p<.001) and the second premolar extraction group (3.4 ± 1.85mm, p<.001) compared to the non-extraction group (1.56 ± 1.38mm). It was also confirmed that there was a significant difference in horizontal displacement of lower molars between the 3 groups, F(2,188)=54.514, p<.001. However, Tukey’s post-hoc analysis showed significantly more horizontal displacement of lower first molars between all groups. The lower molars in the four second premolar extraction group showed the greatest forward movement (3.03 ± 1.41mm, p<.001), the four first premolar group showed less change (2.17 ± 1.3mm, p<.001) and the non-extraction group showed minimal displacement (.92 ± .65mm). The difference between the four first premolar and four second premolar extraction groups was statistically significant at p<.001. 61 A statistically significant difference was also confirmed in the amount of lower incisor extrusion, F(2,188)=8.125, p<.001. Tukey’s post-hoc analysis showed more lower incisor extrusion in the non-extraction group (1.74 ± 1.18mm, p<.001) compared to the four second premolar extraction group (.62 ± 2.1mm). No other variables showed statistically significant differences in amount of change between the 3 groups. Non-parametric Tests Figure A.1. Upper first molar horizontal displacement. A Kruskal-Wallis H test showed that there was a statistically significant difference in upper molar hori- 62 zontal displacement between the 3 groups, χ2=42.211, p<.001, with a mean displacement of 1.56mm for the nonextraction group, 2.77mm for the four first premolar extraction group and 3.4mm for the four second premolar extraction group. A Mann-Whitney U test was used to determine where the among-groups differences lay. Distributions were not similar, as assessed by visual inspection. Upper first molar horizontal displacement in the four first premolar extraction group (mean rank=82.10) was significantly greater than in the non-extraction group (mean rank=48.48), U=970, z=-5.122, p<.001. The four second premolar extraction group (mean rank=85.17) showed significantly greater upper first molar horizontal displacement compared to the non-extraction group, U=871.5, z=5.772, p<.001. The four first premolar group (mean rank=56.45) and the four second premolar group (mean rank=68.36) were not significantly different, U=1552.5, z=-1.844, p=.065. It was also confirmed that there was a statistically significant difference in lower molar horizontal displacement among the 3 groups, χ2=75.871, p<.001, with a mean displacement of .92mm for the non-extraction group, 2.17mm for the four first premolar extraction group and 3.03mm for the four second premolar extraction group. 63 Figure A.2. Lower first molar horizontal displacement. Again, a Mann-Whitney U test was used to determine where the differences lay between the groups. Distributions were not similar, as assessed by visual inspection. Lower first molar horizontal displacement in the four first premolar group (mean rank=83.73) was significantly greater than in the non-extraction group (mean rank=46.99), U=870.5, z=-5.596, p<.001. The four second premolar extraction group (mean rank=94.02) showed significantly greater lower molar horizontal displacement than the non-extraction group (mean rank=38.68), U=313.5, z=-8.372, p<.001. The four second premolar group (mean rank=72.99) showed significantly greater horizontal dis64 placement than the four first premolar extraction group (mean rank=51.66), U=1260.5, z=-3.304, p=.001. The lower incisor vertical displacement was also significantly different between the 3 groups, χ2=8.265, p=.016, with a mean displacement of 1.74mm for the nonextraction group, 1.24mm for the four first premolar group and .62mm for the four second premolar extraction group. No other variables showed statistically significant differences. Figure A.3. Lower incisor vertical displacement. Once more, a Mann-Whitney U test was run to identify the differences between the 3 groups. Distributions were not similar, as assessed by visual inspection. Lower 65 incisor vertical displacement in the non-extraction group (mean rank=70.49) and the four first premolar extraction group (mean rank=57.93) were not significantly different, U=1642.5, z=-1.913, p=.056. The non-extraction group (mean rank=74.39) showed statistically significantly greater lower incisor vertical displacement than the four second premolar group (mean rank=56.05), U=1515, z=2.774, p=.006. The four first premolar group (mean rank=65.53) and the four second premolar group (mean rank=59.56) were not significantly different, U=1736.5, z=-.925, p=.355. A possible explanation for the outliers in the four first premolar and four second premolar extraction groups may be that patients with deep curves of Spee are more susceptible to extraction treatment, keeping in mind that levelling requires space. When levelling, lower incisors may be intruded or tipped forward, which would result in a negative value for lower incisor vertical displacement. 66 Vita Auctoris Vasileios Charalampakis was born on July 28, 1986 in Athens, Greece. He received his D.D.S. degree from the National and Kapodistrian University of Athens in 2011. In June 2012 he moved to the United States of America and began his orthodontic residency program at the Center for Advanced Dental Education at Saint Louis University, in Saint Louis, Missouri. He is presently a candidate for the degree of Master of Science in Dentistry. 67