UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Changes in the occlusal curves secondary to serial sextractions compared to late premolar extractions… Feldman, Esther 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_september_feldman_esther.pdf [ 1.4MB ]
Metadata
JSON: 24-1.0167464.json
JSON-LD: 24-1.0167464-ld.json
RDF/XML (Pretty): 24-1.0167464-rdf.xml
RDF/JSON: 24-1.0167464-rdf.json
Turtle: 24-1.0167464-turtle.txt
N-Triples: 24-1.0167464-rdf-ntriples.txt
Original Record: 24-1.0167464-source.json
Full Text
24-1.0167464-fulltext.txt
Citation
24-1.0167464.ris

Full Text

CHANGES IN THE OCCLUSAL CURVES SECONDARY TO SERIAL EXTRACTIONS COMPARED TO LATE PREMOLAR EXTRACTIONS AND CONTROLS  by Esther Feldman M.Sc., The University of British Columbia, 2014 D.M.D., McGill University, 2011 B.Sc., McGill University 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE  AND POSTDOCTORAL STUDIES (Craniofacial Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  May 2014 © Esther Feldman, 2014  ii Abstract Objectives: To determine the relationships between tooth tipping and occlusal curvature patterns in serial extraction (SE) cases versus late premolar extraction (LPE) cases and controls.   Methods: Mandibular dental casts and cephalometric radiographs were collected from 90 subjects (30 Class I controls, 30 SE cases, 30 LPE cases) at 3 time points: baseline (T0) for controls and SE, after natural drift/pre orthodontics (T1) for controls, SE, and LPE, and after comprehensive orthodontic treatment (T2) for the SE and LPE groups. The casts were scanned and rendered as virtual models. 12 cusp tips (4 incisors and 2 first molars) were digitized and the occlusal curves were measured using Rhinoceros by fitting a sphere to the landmarks (least-squares method).  Radiographs were digitized and the long axes of the central incisor, canine, and first molar were landmarked and related to palatal plane to determine the direction and amount of tipping.    Results: The SE group differed significantly from the other two groups at T1, with smaller radii in 2/3 outcomes. From T0-T1, the SE group showed a tendency for the incisor and canine to tip distally, while the molar tipped mesially.  From T1-T2, the SE group had smaller radii versus LPE and demonstrated mesial tipping of the incisor and canine angulations, with the molar tipping distally.  Conclusions: SE cases tend to have steeper occlusal curves (smaller radii), post tooth drift and after orthodontic treatment as compared to the other groups. In general, orthodontic treatment, post SE, involved incisor and canine proclination, as well as molar uprighting. SE is unlikely to save more than 4-6 months of active treatment due to the time needed to correct the increased occlusal curvature that results from this procedure.  iii Preface This research was passed by the UBC Research Ethics Board at the University of British Columbia, ethics certificate number H12-02568.  iv Table of Contents  Abstract .................................................................................................................................... ii Table of Contents ................................................................................................................... iv List of Figures ........................................................................................................................ vii List of Abbreviations ............................................................................................................. ix Acknowledgements ................................................................................................................. x Dedication ............................................................................................................................... xi Chapter  1: Introduction ........................................................................................................ 1 1.1 Normal Development ................................................................................................ 3 1.2 Responses to Serial Extraction .................................................................................. 9 1.2.1 The Effects of Extractions on the Soft Tissue Profile ...................................... 15 1.2.2 Serial Extractions and Occlusal Treatment Goals ........................................... 17 1.3 Curvature of the Occlusal Surfaces ......................................................................... 18 1.3.1.1 The Curve of Spee ...................................................................................... 23 1.3.1.1.1 Definition .............................................................................................. 23 1.3.1.1.2 Functional and Clinical Significance .................................................. 25 1.3.1.1.3 Methods of Measurement: Strengths and Limitations ..................... 29 1.3.1.2 The Curve of Wilson ................................................................................... 30 1.3.1.2.1 Definition .............................................................................................. 30 1.3.1.2.2 Functional and Clinical Significance .................................................... 30 1.3.1.2.3 Methods of Measurement: Strengths and Limitations ..................... 31 1.3.1.3 Curves of Spee and Wilson in Serial Extraction ...................................... 32  v Statement of the Problem ..................................................................................................... 35 1.4 Objectives ............................................................................................................... 36 1.4.1 Null Hypotheses .................................................................................................. 37 1.4.2 Study Hypotheses................................................................................................ 37 Chapter  2: Methodology...................................................................................................... 38 2.1 Materials and Methods ............................................................................................ 38 2.1.1 Subjects ............................................................................................................... 38 2.1.2 Data Collection and Measurements ................................................................. 39 2.1.2.1 Cast Measurements ..................................................................................... 40 2.1.2.2 Cephalometric Measurements ..................................................................... 42 2.1.3 Statistical Analysis .............................................................................................. 43 2.2 Reliability testing .................................................................................................... 44 2.3 Laser Scanner .......................................................................................................... 45 Chapter  3: Results................................................................................................................ 47 3.1 Testing of matching of study groups ...................................................................... 47 3.2 Preparing for Data Analysis – assessing data for normality ................................... 48 3.3 Analyses of Cast Data ............................................................................................. 48 3.4 Analyses of  Cephalometric Data ........................................................................... 63 Chapter  4: Discussion .......................................................................................................... 74 Chapter  5: Conclusion ......................................................................................................... 95 Bibliography .......................................................................................................................... 96  vi List of Tables Table 3.1    Matching of study groups .................................................................................... 48 Table 3.2    Cast Radii at TO (baseline) comparisons between Serial Extractions and Control groups  ..................................................................................................................................... 51 Table 3.3    Comparing at T1 (after intervention in SE): controls, SE and LPE groups (ANOVA with Bonferroni post hoc adjustment)....................................................................52 Table 3.4    Multiple comparisons of cast radii at T1 between controls, SE, and LPE...........53  Table 3.5    Cast Radii at T2 (post orthodontic treatment) - comparisons between SE and LPE groups .....................................................................................................................................54 Table 3.6    Predictors of variance in radii at T1 Multiple Linear Regression (MLR) models.....................................................................................................................................56 Table 3.7    Predictors of variance in radii at T2 –MLR models............................................58 Table 3.8    Cephalometric tip measurements at TO (baseline)  - comparisons of mean angles between SE and control groups .............................................................................................65 Table 3.9    Multiple comparisons of cephalometric tip measurements at T1 between control, SE, and LPE groups ..............................................................................................................67 Table 3.10   Cephalometric tipping measurements at T2 (post orthodontic treatment)  - comparisons between SE and LPE groups ...........................................................................68 Table 3.11   Predictors of variance in tooth tip at T1– Multiple Linear Regression (MLR) models...................................................................................................................................71 Table 3.12   Predictors of variance in tooth tip at T2 –MLR models...................................73   vii List of Figures Figure 2.1    Rhinoceros’ digitized representations of Monson’s sphere, curve of Wilson, and curve of Spee.........................................................................................................................41 Figure 2.2    Measurement of tipping of incisor, canine and molar represented by changes in angulation relative to palatal plane (ANS-PNS) at T0, T1, and T2......................................43 Figure 3.1    Control group radii from T0-T1.......................................................................49 Figure 3.2    SE group radii from T0-T1-T2.........................................................................49 Figure 3.3    LPE group radii from T1-T2.............................................................................50 Figure 3.4    Comparison of radii in controls and SE at T0...................................................51 Figure 3.5    Comparison of radii in controls, SE, and LPE at T1.........................................52 Figure 3.6    Comparison of radii in SE and LPE at T2.........................................................57 Figure 3.7    First molars’ ML and DL cusps’ distance from the reference plane (MB and DB first molar cusp tips) from T0-T1 in the control group..........................................................59 Figure 3.8    First molars’ ML and DL cusps’ distance from the reference plane from T0-T1-T2 in the SE group.................................................................................................................59 Figure 3.9    First molars’ ML and DL cusps’ distance from the reference plane from T1-T2 in the LPE group....................................................................................................................60 Figure 3.10  Comparison of first molars’ ML and DL cusps’ distance from the reference plane at T0 in controls and SE...............................................................................................60 Figure 3.11  Comparison of first molars’ ML and DL cusps’ distance from the reference plane at T1 in controls, SE, and LPE.....................................................................................61 Figure 3.12  Comparison of the first molars’ ML and DL cusps’ distance from the reference plane at T2 in SE and LPE....................................................................................................62  viii Figure 3.13  Cephalometric tip in the control group from T0-T1........................................64 Figure 3.14  Cephalometric tip in the SE group from T0-T1-T2.........................................64 Figure 3.15  Cephalometric tip in the LPE group from T1-T2............................................64 Figure 3.16  Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in controls and SE at T0.................................................................................65 Figure 3.17  Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in controls, SE, and LPE at T1.......................................................................68 Figure 3.18  Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in SE and LPE at T2.......................................................................................69        ix List of Abbreviations  SE: serial extractions LPE: late premolar extractions TSALD: Tooth size-arch length discrepancy    x Acknowledgements  I offer my enduring gratitude to the faculty, staff and my fellow students at UBC, who have inspired me to continue my work in this field. I owe particular thanks to Dr. D.B. Kennedy, whose clinical experience and wisdom guided me through the research process.    I’d like to thank Dr. E.H. Yen, who pushed me to go beyond my means, Dr. A. Hannam for his patience and gentle approach, and Dr. J. Aleksejuniene, who helped me maintain focus. A special thank you to Alex Vo at OHSU for his help with data collection.  Last, but not least, thank you to my rock, Jordan Kravitz.  You stayed positive for me during the hard times, always believed in me, and were never too tired to give me a much-needed pep talk.    xi Dedication I dedicate this thesis to my parents, Howard and Debbie, who have supported me, encouraged me, and stood by me throughout my many years of education. I love you.       1 Chapter  1: Introduction  The objective of orthodontic treatment is to align and level the teeth to establish a stable, esthetic and functional occlusion (Kjellgren, 1947; Hotz, 1970; Graber, 1971; Andrews, 1972;  Boley, 2002; Proffit, Fields, & Sarver, 2012).  Malocclusions that present with tooth size-arch length deficiencies (TSALD), can be treated by either extraction or non-extraction strategies (Boley, 2002; Gianelly, 1994; Graber, 1971; Proffit et al., 2012).  For mixed dentition patients, a non-extraction plan would utilize the leeway space, preserving the arch length to achieve proper alignment (Gianelly, 1994).   Another alternative would be expansion therapy, however this method can move teeth into abnormal stress positions, increasing muscular system pressures (Graber, 1971).  When leeway space management and/or expansion therapy would be inadequate to resolve the crowding, serial extraction (SE) has been employed to resolve severe TSALD in the mixed dentition (Boley, 2002; Gianelly, 1994; Graber, 1971; Kjellgren, 1947; Proffit et al., 2012).  When the same malocclusion presents in the permanent dentition, severe crowding likely dictates late premolar extractions (LPE) (Gianelly, 1994). When resorting to extractions, it is essential that they be systematically planned, to limit disadvantages and to facilitate the best possible intended effect (Kjellgren, 1947).  The earliest documented attempts at SE were made in the mid 18th century (Ringenberg, 1967).  Casual references to SE were made up until the mid 19th century, and then lost momentum and fell under the radar (Ringenberg, 1967).  SE treatment was re-popularized by pioneers of orthodontics such as Tweed, Dewel and Kjellgren in  2 the 1950’s and 60’s (Kjellgren, 1947; Dewel, 1954; Ringenberg, 1967; Brandt & Tweed, 1967).  The consensus regarding the advantages of SE include: (1) permits physiologic tooth movement, (2) reduces active treatment time, and (3) reduces the retention period (Ringenberg, 1967; Brandt & Tweed, 1967; Gianelly, 1994; Boley, 2002).  Other mentioned advantages include a reduced load on anchorage units and better maintenance of the alveolar bone and periodontal tissues (Kjellgren, 1947; Brandt & Tweed, 1967; Ringenberg, 1967).    The disadvantages of SE include: (1) deeper overbite, (2) lingual tipping of incisors, (3) the potential of scar tissue build up in the extraction space, (4) diastema and spacing, and (4) alteration of tongue function (Brandt & Tweed, 1967; Ringenberg, 1967).  Craniofacial changes concerning amount of growth, direction of growth, and changes in facial convexity demonstrated no significant differences between SE and control groups (Kjellgren, 1947; Ringenberg, 1967).  The SE disadvantage of an increased overbite implies a deeper curve of Spee, which is defined as the anatomic curve established by the occlusal alignment of the teeth, as projected onto the median plane (Prosthodontics, 1994).  Simons and Joondeph observed that malocclusions with an exaggerated curve of Spee are frequently observed in conjunction with deep overbites (Simons & Joondeph, 1973).   This severe curve of Spee may be further enhanced by tooth drift in SE cases.    Another occlusal curve, the curve of Wilson, results from the lingual inclination of the lower posterior teeth, making the lingual cusps lower than the buccal cusps on the mandibular arch; the buccal cusps are higher than the lingual cusps on the maxillary arch because of the outward inclination of the upper posterior teeth (Dawson, 1988). It  3 can be visualized clinically by a lingually rolled lower arch, which occurs in many deep bite cases (McLaughlin & Bennett, 2003).    The importance of these conceptual occlusal curves are reflected by their presence in the criteria to score posttreatment dental casts and panoramic radiographs using the American Board of Orthodontics’ Objective Grading System (OGS), which was renamed the Cast-Radiograph Evaluation (CRE) in 2000 (Casko et al., 1998).  Specifically, the alignment and buccolingual criteria of the CRE address the aforementioned curves (Casko et al., 1998).  The curve of Spee is considered in terms of anterior and posterior alignment in the sagittal plane, as well as marginal ridge height discrepancies (Casko et al., 1998).  Since flattening of these occlusal curves is an end goal of quality orthodontic treatment, it would be helpful to better understand the way in which different treatment options can affect these curves (Andrews, 1972).  Ringenberg (1967) and O’Shaughnessy (2011) compared SE and LPE patients and found reductions in active treatment time of 6 months and between 4-6 months compared to LPE cases, respectively (Ringenberg, 1967; O'Shaughnessy, 2011).  Ringenberg found no difference in treatment outcome between SE and LPE patients, except a later age at treatment finish in the LPE group (Ringenberg, 1967).   After searching the literature, no evidence as to how SE affects occlusal curves has been researched in detail.   Perhaps, changes in occlusal curves will help explain why SE cases only show a minimal 4-6 months reduction of active treatment time. 1.1 Normal Development Attempts to develop or understand preventative methods in the field of orthodontics must be based on an adequate foundation of normal growth and development of the  4 dentitions and the many modifying factors (Baume, 1950).  Human occlusion and arch changes are a dynamic process, as seen by the developmental trends and changes observed during child growth (Moorrees, Lebret, & Frohlich, 1969).  Growth and development studies have described the natural dental changes from the eruption of the first primary incisors through to adulthood (Moorrees et al., 1969; Sinclair & Little, 1983).   Longitudinal samples of untreated indivuduals have provided invaluable research data that enable orthodontists to understand changes in the dentition and jaws with increasing age (Sinclair & Little, 1983).  Growth changes in human jawbones have been reported coincident with the emergence of the permanent teeth (Moorrees et al., 1969).  As such, the data is grouped by stage of dental development or dental age instead of chronological age (Moorrees et al., 1969).    The occlusion is first established in the primary dentition, and little change occurs in the interdental space between the deciduous incisors (Moorrees et al., 1969).  The spaces between the deciduous molars close when the first permanent molars erupt (age six) and the primate spaces in the mandibular arch (between the deciduous mandibular canines and first molar) close partially or completely as well (Moorrees et al., 1969; Sinclair & Little, 1983).  The intercanine width is established in both males and females on average between 6-10 years, after eruption of the permanent lateral incisors (Moorrees et al., 1969; Sinclair & Little, 1983).  According to Moorrees, arch breadth does not change from 4-6 years, but increases markedly (3mm) with the emergence of the permanent incisors (Moorrees et al., 1969).  During the second phase of development (age 6-12), additional space becomes available – the leeway space, as the deciduous molars occupy a greater mesiodistal crown diameter than their permanent  5 successors (Moorrees et al., 1969).  This space is supposed to accommodate the alignment of the permanent canine, which is greater in size compared to its predecessor (Moorrees et al., 1969).    The second phase terminates with the eruption of the second premolars (age 10-12).  Arch length is measured as the length of the tangent line drawn from the mesial of the proximal surface of the lower first permanent molar to the midpoint of the lower central incisors (Moorrees et al., 1969).    The trends in the mandible include two separate decreases in arch length up until age eighteen.  First, there is a slight decrease occurring prior to the emergence of the permanent first molars due to closure of primate spaces spaces between the deciduous canines and molars (Moorrees et al., 1969).  This is followed by a second decrease in arch length post exfoliation of the deciduous molars, particularly the second primary deciduous molar as a result of mesial drifting of the first permanent molars.  Therefore, mandibular arch length is shorter at 18 years than at 4 years (Moorrees et al., 1969). The changes in arch length and width directly affect arch circumference or perimeter (measured from the distal surface of the second premolar to its antimere) (Moorrees et al., 1969).  Between the ages of 5 and 18 years, the mandibular arch circumference exhibits a mean decrease of 3.4 mm in males and 4.5 mm in females (Moorrees et al., 1969).    Considerable individual variations in the changes in arch circumference exist (Moorrees et al., 1969).  These individual variations are explained by differences in the amount of interdental space in the deciduous dentition, the changes in arch breadth and arch length, the ratio of mesiodistal crown diameters of deciduous teeth and their  6 permanent successors, and the sequence of shedding and emergence of the permanent posterior teeth(Moorrees et al., 1969). The aforementioned variables constitute the determinants of tooth alignment during dental development (Moorrees et al., 1969).    Moorrees’ review studied growth changes and emphasized the importance of permanent tooth alignment because of the individual variations that exist in the relationship of mesiodistal crown diameters of permanent teeth, their predecessors, the migration of teeth in the arches, and changes in arch dimension (Moorrees et al., 1969).   Another critical consideration is the individual differences in the timing and sequence of permanent tooth emergence (Moorrees et al., 1969).   These remain potential sources of favorable and unfavorable trends in dental occlusion development for the individual child (Moorrees et al., 1969).     The Iowa Facial Growth Study began in 1946 and consists of approximately 3600 lateral cephalometric radiographs, anteroposterior cephalometric radiographs, frontal and profile facial photographs, dental radiographs, models, demographic data, and anthropometric measurements taken annually on 125 subjects from the ages of five to eleven years (Meredith, 1959).  Bishara used this sample to investigate changes in tooth size arch length discrepancies and consider the associated dentofacial variables that play a role (Bishara, Jakobsen, Treder, & Stasl, 1989).    The results showed a decrease in available arch length in both sexes, and both the maxillary and mandibular arches with increasing age (Bishara et al., 1989).  This decrease has a multifactorial etiology, including changes in facial height, overbite, incisor inclination, arch dimensions and the mesiodistal diameter of various teeth (Bishara et al., 1989; Moorrees et al., 1969).    Arya et al (1973) studied the permanent molar occlusion longitudinally as the  7 teeth transition from primary to permanent dentition (Arya, Savara, & Thomas, 1973).   The study included casts at 3 different time points: 1) the complete deciduous dentition prior to first permanent molar eruption, 2) when the first permanent molars make initial occlusal contact, and 3) once all of the permanent teeth have erupted.  Patterns that existed in the primary dentition terminal plane of occlusion were used to predict where the permanent molars will likely occlude initially, and finally, with all of the remaining teeth erupted (Arya et al., 1973). Within the observed sample, they determined that while a direct association of considerate degree exists between the initial occlusion of the first permanent molar and the deciduous terminal plane relationship, the former is not completely dependent on the latter (Arya et al., 1973).  The classification of occlusal development in the primary dentition is based on the anteroposterior distance-terminal plane difference between the distal surfaces of the opposing maxillary and mandibular primary second molars (Arya et al., 1973).  It includes three distinct categories: a mesial step, a flush terminal plane, and a distal step occlusion (Arya et al., 1973).   Almost half of the first permanent molars in the study erupted in a flush terminal plane initial occlusion, with the remainder equally split between mesial (normal) and distal step initial occlusion (Arya et al., 1973).   Of those casts with an initial cusp-to-cusp flush relationship, 70% proceeded to settle into a Class I molar relationship, and 30% became Class II (Arya et al., 1973).   The distal step (Class II) and mesial step (Class I) occlusion samples remained constant throughout the period of permanent tooth eruption (Arya et al., 1973).     Sinclair’s study on the maturation of untreated normal occlusions demonstrated that arch length and intercanine width decreases over time with a substantial average  8 loss of 4.83 mm and 0.75mm respectively, noted by early adulthood in both genders (Sinclair & Little, 1983).  Intercanine width constriction in males was gradual over the entire period, while females experienced the majority of changes during the permanent dentition stage (Sinclair & Little, 1983).  Intermolar width changes are gender specific, with a small increase seen in males up until eighteen years and a significant loss in width seen in females from early teenage years up to seventeen years old (Sinclair & Little, 1983).  Overall, intermolar width is a generally very stable measure (Sinclair & Little, 1983).  Minimal changes amongst the sexes were seen in both overbite and overjet measurements, with general trends of decreasing values with age seen with respect to both (Sinclair & Little, 1983).  Incisor irregularity increased during the permanent dentition interval and females showed a greater degree of crowding versus the males (Sinclair & Little, 1983).  Patterns revealed by this study included a deterioration in ideal alignment from eight to seventeen years old, a progressive worsening in crowding with age, and a tendency for uncrowded patients to start crowding as they age (Sinclair & Little, 1983).    To evaluate the impact of facial growth patterns on crowding, a long-term posttreatment study evaluating the relationship between divergent facial growth patterns and lower anterior crowding, revealed that the type of facial pattern, high or low angle, was not associated with increased risk of postretention relapse of mandibular incisors at the time of appliance removal (Fudalej & Årtun, 2007; Goldberg, Behrents, Oliver, & Buschang, 1994).  Goldberg studied seventy-five white non-hypodivergent extraction patients to better understand the impact of facial divergence on posttreatment mandibular crowding (Goldberg et al., 1994).  The results were gender  9 specific, with the female subjects with the greatest posttreatment increases in facial divergence exhibiting the greatest posttreatment crowding (Goldberg et al., 1994).   Additionally, the females with the most posttreatment lower incisor eruption and greatest increases in anterior facial growth showed the greatest increases in crowding (Goldberg et al., 1994).  Goldberg concluded that posttreatment increases in crowding were also related to posttreatment decreases in arch widths in both male and female patients (Goldberg et al., 1994).  The above growth and development studies, covering everything from eruption patterns, the transitions from primary to mixed to permanent dentition, changes in intercanine width, arch breadth, arch circumference, as well as facial growth, are the multifactorial facets that contribute to the development of the occlusal curves.  When normal development has failed to occur, and deficient structures are apparent at an early age, the treatment alternatives become limited.  The orthodontist cannot recover jaw growth deficiencies and therefore can either opt to extract and reposition the remaining teeth over the available supporting bone, or to attempt to expand and distalize the molars (Dewel, 1954). 1.2 Responses to Serial Extraction In 1954, Dewel stated that intercepting malocclusion with a preliminary program of serial extraction had a legitimate place in the orthodontic profession. This was only accurate provided that the indications to extract were justified and the postextraction shifting was biomechanically controlled (Dewel, 1954).  Crowding in the mixed dentition is one of the most common problems in orthodontics (Boley, 2002).   More often than not, crowding results from a TSALD  10 (Boley, 2002).  Once the complete set of permanent teeth have erupted, the two main available treatment options for crowding are either arch expansion or reducing tooth number by extraction to provide adequate space for alignment (Boley, 2002).  One orthodontic goal is to make a timely diagnosis and take the necessary measures to prevent the future development of serious malocclusions. “Serial extraction” (SE), first described by Kjellgren in 1947, is also known as “guidance of eruption”, which was the term coined by Hotz in 1970.  It is an interceptive process for correcting discrepancies between tooth size and arch size (R. P. Hotz, 1970; Kjellgren, 1947).  In many severe crowding (>7mm) cases, a decision is made during the early mixed dentition.  On a case by case basis, when expansion is deemed inappropriate to treat the crowding, the resulting decision often includes extractions in the permanent teeth (Proffit, Fields, & Sarver, 2012).  Serial extraction is the planned, sequential, removal of primary and permanent teeth and involves extraction of the deciduous canines, followed by the first deciduous molars and finally the first premolars (Proffit et al., 2012).  This results in the creation of space in the anterior segments in severely crowded cases, at the expense of the lateral segment (Wagner & Berg, 2000).  This increase in space may allow the permanent incisors and canines to erupt into position in the attached gingiva (Wagner & Berg, 2000).  While tooth extraction is irreversible, it can have both social and economic benefits such as spontaneous alignment of the incisors and reducing the need for appliances, the treatment time, the cost of treatment, discomfort for the patients and potential iatrogenic sequelae (Boley, 2002; O'Shaughnessy, Koroluk, Phillips, & Kennedy, 2011; Papandreas, Buschang, Alexander, & Kennedy, 1993).  The theory is that after SE, guidance of eruption of the permanent teeth into the physiologic position  11 over the basal bone leads to proper formation of the periodontal fibres.  This may result in a more stable orthodontic outcome, with a lower risk of relapse (Graber, 1971).  However, the majority of patients that are treated with SE will eventually require comprehensive fixed orthodontics to resolve tipping, residual spacing, and imperfect alignment (Boley, 2002; Graber, 1971; Kennedy, Joondeph, Osterberg, & Little, 1983; Moorrees, Fanning, & Gron, 1963; O'Shaughnessy et al., 2011; Wilson, Little, Joondeph, & Doppel, 1999).   SE is a biologically sound, proven technique and should not be considered as compromised treatment (Graber, 1971).  While progress into full treatment frequently occurs, a lag period between the extractions and fixed appliances is recommended (Creekmore, 1982).  The purpose of this is to reap all the potential benefits such as improvements in occlusal relationships, alveolar bone support, and some spontaneous alignment as a result of physiological dental drift, termed “driftodontics” post extraction (Alexander, 2001; Creekmore, 1982).   A major advantage of this is providing the canines with ample room to erupt into the attached gingival, and not labially into alveolar mucosa (Kjellgren, 1947).  Tooth migration, by means of mesial or distal shifting, is a physiologic phenomenon (Moorrees et al., 1969).  Kau et al in 2004 performed a randomized controlled trial looking at the effects of primary canine extractions on the dental arch parameters in 8-9 year old children over a two-year period using dental casts, but no cephalometric radiographs were taken (Kau, Miotti, & Harzer, 2004).  Incisor crowding reduced over time in both the SE and the control groups, while arch perimeter decreased to a greater degree in the SE cohort (Kau et al., 2004). The improvement in the irregularity index measured 4.43 mm in the SE group, while it was 2.44 mm in the untreated controls (Kau et al., 2004).  Arch  12 length decreased 2.95 mm in the SE group and only 1.51 mm in the control group (Kau et al., 2004).  Their results suggest that the reason for this loss in arch perimeter is due to the lower molars migrating forward, as the incisor inclination remained constant (Kau et al., 2004). They also addressed how this reduction in incisor crowding was only temporary, as in actuality, the crowding is being transferred posteriorly (Kau et al., 2004).  It is important to consider the financial implications from a public health standpoint, given that only a one in four chance of improvement in incisor irregularity was seen in Kau’s extraction cases.   Also important to note is that these results should be accepted with caution, as no radiographic records were available to support their model-based findings.  Another limitation is the lack of standardization, as no dental age (which does not correlate to chronological age) or time frame for follow up was defined.  Yoshihara et al in 2000 addressed the spontaneous changes in the dentition post SE, observing the drift of their sample at three different time points: before primary canine extraction, average age 8.74 years (T1), after permanent premolar extraction, average age 11.91 years (T2), and at the end of the observation period, approximately six years later, average age 14.73 years (T3) (Yoshihara, Matsumoto, Suzuki, Sato, & Oguchi, 2000).  They were particularly interested in incisor and molar tipping movements, as well as improvement in alignment.  They concluded that overall, spontaneous changes in the dentition after SE might be useful for therapeutic purposes of the correction of or reduction of crowding in most Class I cases (Yoshihara et al., 2000).   Careful follow up is a necessary underlying theme for every SE case, as individual variability exists (Yoshihara et al., 2000).  Their results showed an overall  13 decrease in Little’s irregularity index as serial extractions proceeded; this crowding decreased more during the observation period (Yoshihara et al., 2000).  The majority of the crowding was corrected after primary canine extractions, followed by spontaneous improvement with increasing age after first premolar extraction (Yoshihara et al., 2000).  The patterns observed included greater lower incisor cusp and apex movement (tipping) after primary canine extraction and greater molar cusp and apex movement post first premolar extraction (Yoshihara et al., 2000).  However, the authors noted that molar movement differed amongst the three Angle classifications, with a net mesial movement being most prominent at -5.56 mm in Class III and to a much lesser extent, at  -1.62 mm amongst Class II cases (Yoshihara et al., 2000).  In addition a significant correlation was found between the amount of arch length discrepancy (space required subtracted from the space available) and the change in the irregularity index.  Increased crowding yielded the greatest alignment correction.  A negative correlation was observed between canine tipping with arch length discrepancy and the annual change in the irregularity index (Yoshihara et al., 2000).   Quantification of this “ driftodontics” was examined by Papandreas et al in 1993 who compared first premolar extraction cases in the mixed dentition versus late extraction in the permanent dentition. They quantified the amount of physiologic drift of the untreated mandibular dentition following extractions in both the mixed and permanent dentitions (Papandreas et al., 1993).  The late premolar extraction (LPE) group showed greater proclination and irregularity of the incisors before extractions versus the early premolar extraction (EPE) group.  The rate of lower incisor retroclination was four times greater in the LPE compared to the EPE group and was  14 attributed to the distal movements of the incisal edge.  The LPE group showed a significantly greater increase in overbite and more pronounced decrease in incisor irregularity, while exhibiting less of a decrease in intermolar width (Papandreas et al., 1993).   This was mainly thought to be due to the lower incisors being more proclined in the LPE group (Papandreas et al., 1993).  The authors recommended future studies using a more homogeneous sample to analyze changes (Papandreas et al., 1993).    O’Shaughnessy et al studied treatment efficiency by comparing treatment times and outcomes for SE and LPE cases in a private practice setting (O'Shaughnessy et al., 2011).   They found that pretreatment (T1) PAR scores were greater in the LPE cohort, while no significant difference was observed posttreatment, at T2, showing that both treatment modalities can clinically achieve similar occlusal results (O'Shaughnessy et al., 2011).  SE had significantly reduced mean active treatment time of four to six months; however this group had a significantly greater overall treatment time, because pre-active recalls were counted, as well as the total number of visits (O'Shaughnessy et al., 2011).  There was no significant difference between early and late extraction groups observed for total chair time in minutes.  Their final conclusion stated that SE reduces active treatment time for severe crowding cases, but an increased commitment in overall time and effort precedes the active-treatment stage (O'Shaughnessy et al., 2011). Wagner and Berg reached a similar conclusion in their study, where the group that underwent SE had a significantly shorter period of time in fixed appliances, but it came at the expense of a significantly higher number of overall appointments and a longer duration of treatment/observation time (Wagner & Berg, 2000).   They discussed other factors that could have a significant impact on treatment duration.  These include:  15 cooperation and motivation of child and parents, efficiency of the mechanics used, and operator skill (Wagner & Berg, 2000).  Both the Wagner & Berg study and the O’Shaughnessy study support that the end results in SE and LPE groups, were statistically similar as determined by the PAR index (Wagner & Berg, 2000; O'Shaughnessy et al., 2011).   1.2.1 The Effects of Extractions on the Soft Tissue Profile Concerns have been raised that extractions have a negative soft tissue profile effect (Stephens, Boley, Behrents, Alexander, & Buschang, 2005; Wilson et al., 1999).  Wilson et al evaluated these soft tissue profile changes within and between groups of patients treated with SE, both with and without subsequent orthodontic treatment and patients treated with late extractions of 4 first premolars (Wilson et al., 1999).  Changes in lower incisor position were not reflected in the lower lip or other soft tissues.  No differences were found in soft tissue profile changes amongst the observed groups (Wilson et al., 1999).  Although differences were observed between the groups over the treatment period, during the post-retention/extraction periods, the changes observed were similar for all, most likely as a consequence of natural growth.  This resulted in overall similar soft tissue facial profile changes in both sexes long-term (Wilson et al., 1999).   Stephens et al determined that if extraction and nonextraction patients are treated to the same incisor position and lip line, the treatment modality should not affect any long-term changes in the soft tissue profile (Stephens et al., 2005). Stephens’ extraction and nonextraction groups also turned out to be no different at the long-term  16 follow-up 15 years later, indicating that the changes were similar for both groups (Stephens et al., 2005).  Both groups demonstrated significant changes over time; their lips became significantly more retruded in relation to the E- and S-lines, and their facial convexity decreased considerably over the long term (Stephens et al., 2005).  These changes follow patterns expected for untreated subjects, and are due to greater relative growth in the nose and chin areas (Stephens et al., 2005).  They are a result of age changes rather than treatment selection, provided that the treatment goal of final incisor position was met (Stephens et al., 2005).   Bishara and Jakobsen, who randomly presented extraction and nonextraction profile silhouettes of pretreatment and posttreatment patients to laypeople, demonstrated that the general public does not prefer the profiles of one group over another (Bishara & Jakobsen, 1997).  Information about preferences is important because a major goal of orthodontic treatment is to attain and maintain a good esthetic result, which ultimately must be based on the general public’s opinions (Bishara & Jakobsen, 1997).  In a similar study, Bowman and Johnston conducted a study with two panels: one containing 58 laypersons and one containing 42 dentists (Bowman & Johnston, 2000). The panels were randomly presented with pre- and posttreatment profiles of 70 extraction and 50 nonextraction Class I and II Caucasian patients and asked to evaluate them (Bowman & Johnston, 2000). The initial evaluation of pretreatment samples were similar, while the posttreatment extraction patients’ faces were evaluated, on average, as being 1.8 mm ‘‘flatter’’ as compared to the nonextraction subjects, and were preferred by both panels.  (Bowman & Johnston, 2000). The overall consensus was that nonextraction treatment had little effect on the profile, whereas the  17 perceived effect of extraction treatment was a function of initial soft tissue protrusion in profile; the greater the initial protrusion, the greater the benefit of the extraction treatment. Both panels viewed four premolar extraction treatment as beneficial when the lips were more protrusive than 2 to 3 mm behind Ricketts’ E-plane.  The conclusion was that the esthetic effect of treatment on the facial profile was directly correlated with the type of treatment, the initial protrusion of the profile, and the observer’s background.  Bowman’s study supports the notion that extraction treatment commonly produces positive esthetic outcomes for patients when the objective is to reduce lip procumbency, but this is not a universal rule and careful diagnosis with an individualized, evidence-based, treatment plan should always be applied (Bowman & Johnston, 2000).   Kim and Gianelly looked at the effects of extractions on smile width esthetics. Their study refuted the idea that extraction treatment results in narrower dental arches which, in turn, would lead to an unesthetic, less-full smile with negative spaces or larger buccal corrdiors (Kim & Gianelly, 2003). Fifty laypersons were asked to assess and randomly rate sixty smile photographs (thirty were extraction and thirty were nonextraction) (Kim & Gianelly, 2003). The results indicated that arch width was not reduced as a function of extractions, and that smile esthetics are comparable in both groups of patients (Kim & Gianelly, 2003).   1.2.2 Serial Extractions and Occlusal Treatment Goals  Tweed, one of the forefather’s of modern orthodontics, became discouraged with Angle’s non-extraction orthodontic approach when he examined the records of relapsed  18 treated cases.   Common findings included protrusive appearances and generally unstable dentitions (Brandt & Tweed, 1967).  Tweed realized that those patients with pleasing facial balance and harmony had mandibular incisors that were upright over basal bone, and then retreated his failures by extracting first premolars (Brandt & Tweed, 1967).  He concluded that carefully planned extractions allowed him to improve facial esthetics and that finishing with the mandibular incisors over the basal bone frequently lead to posttreatment facial balance and stability (Brandt & Tweed, 1967; Vaden, Dale, & Klontz, 2000).  Tweed popularized the once taboo concept of extractions for orthodontic correction and made it widely acceptable (Vaden et al., 2000).  In addition, he developed a fundamentally sound and consistent guide to orthodontics using serial extractions of deciduous and permanent teeth (Vaden et al., 2000).   1.3 Curvature of the Occlusal Surfaces   Virtually everyone has some sort of “malocclusion,” therefore it’s misleading to call it normal or proper occlusion when what is really being referred to is idealized, perfect occlusion (Harris & Corruccini, 2008). A more appropriately neutral term would be “occlusal variation” to refer to the positional variations of tooth relationships that are ubiquitous in the absence of orthodontic treatment.  Occlusion, like all other aspects of human anatomy-physiology, is the result of man's evolution, influenced by the environment and genetic factors (Harris & Corruccini, 2008).  Ideal, flat occlusion provides several benefits over maloccluded teeth (Harris & Corruccini, 2008).  First and foremost, in an esthetically driven western society, ideal occlusion has a superficial benefit.  Studies have shown that individuals with good occlusion are rated as more  19 attractive, more intelligent, and more desirable employees and spouses versus people with malocclusions (Shaw et al., 1979; Shaw, 1981; Shaw et al., 1985; Birkeland et al., 2000; Cerny, 2005; Traebert and Peres, 2007).  A normal, flat occlusion also promotes more tooth-to-tooth contacts, which enables better masticatory performance (Owens et al., 2002).  The occlusal curve deepens as teeth slip their anatomic contacts occlusogingivally  (Harris & Corruccini, 2008).  A physiological determinant of tooth position in the jawbones is the process of tooth migration (Begg, 1954).  Throughout life, as part of the physiologic process of continual tooth eruption, teeth continue to migrate in two directions: in the horizontal dimension (mesial migration) and in the vertical dimension (continual eruption) (Begg, 1954).   Another important factor in the development and maintenance of correct occlusal relationships is the changing anatomy of the teeth, which begin to change soon after eruption because of wear, both occlusally and interpoximally (Begg, 1954).  In all of the aforementioned studies, there is no reference to any changes observed in the occluding surface curvatures, namely the curve of Spee and the curve of Wilson, post SE.  Irrespective of the orthodontic technique used, an important objective in the diagnosis, treatment, retention, and stability of orthodontic therapy is the attainment of a level occlusal plane (Andrews, 1972; Preston, Maggard, Lampasso, & Chalabi, 2008). For clinicians who adhere to Tweed’s principles, the use of continuous arch wires with a reverse curve of Spee incorporated into them are utilized to flatten the occlusal plane (Carcara, Preston, & Jureyda, 2001).  Accordingly, the leveling occurs for the most part by extruding the lower premolars in conjunction with a minimal intrusion of the mandibular incisors (Carcara et al., 2001; Simons & Joondeph, 1973).  In most instances,  20 this leveling tends to reduce the deep anterior overbite associated with a severe curve of Spee (Carcara et al., 2001; Simons & Joondeph, 1973). In order to achieve the objectives of treatment, flattening of the occlusal plane and the elimination of all interferences during function are critical end points (Casko et al., 1998).   The benefits of a flat occlusal plane with natural dentition include obtaining canine rise and anterior guidance, discouraging posterior contact in lateral and protrusive function.  Andrews described the six keys of occlusion in 1972; he advocated leveling the curve of Spee, a conceptual occlusal curvature as viewed in the sagittal plane, which resulted in the best intercuspation and in his opinion, an optimal occlusion (Andrews, 1972). Andrews studied the occlusal planes in 120 nonorthodontically treated and allegedly “normal” occlusions varied from being generally flat to having a mild curve of Spee (Andrews, 1972).  This observation led him to believe that the presence of a curve of Spee could potentially be associated with postorthodontic treatment relapse (Andrews, 1972).  Andrews concluded, "even though not all of the orthodontic normals had flat planes of occlusion, I believe that a flat plane should be a treatment goal as a form of over-treatment” (Andrews, 1972).  Next, Andrews compared the best 1,150 ABO treated cases with these best in nature/untreated cases to provide significant insight on how orthodontics could be improved (Andrews, 1972). From this comparison, the six keys of occlusion were drawn (Andrews, 1972).  In addition, he postulated that a deep curve of Spee increases the difficulty in achieving a Class I canine relationship though it may also result in occlusal interferences that will manifest during mandibular function (Dawson, 1988). There have been reports on how to level the curve of Spee but relatively little is known about the long-term stability of this leveling and how the choice of arch leveling  21 technique may affect its subsequent relapse (Bernstein, Preston, & Lampasso, 2007).  Presently, although limited scientific evidence exists that a shallow curve of Spee is an integral part of a sound occlusion, leveling a deep curve of Spee has been generally accepted as an important contributor to orthodontic treatment success (Andrews, 1972; Bernstein et al., 2007; Hellsing, 1990; Osborn, 1993; S. Braun, Hnat, & Johnson, 1996; Dawson, 1988; Germane, Staggers, Rubenstein, & Revere, 1992; Koyama, 1979; Shannon & Nanda, 2004). Osborn, who studied mammal skulls, demonstrated a highly significant correlation between the forward inclination of the masseter muscle and the mesial tip of the molar teeth in the sagittal plane, conforming to the posterior end of the curve of Spee (Osborn, 1993).  His paper stated that the tilt of the curve of Spee increases the crush/shear ratio of the force produced upon masticating between the molars (Osborn, 1993).  The curve of Wilson, the human occlusal surface curvature as viewed in the coronal plane, is another important consideration addressed in the literature (Ferrario, Sforza, & Miani, 2003).  It can be visualized by a lingually rolled lower arch, which occurs in most palatal expansion cases and many deep bite cases (McLaughlin & Bennett, 2003).   Buccal uprighting in the lower arch to correct the severity in the curve of Wilson is indicated for stability in these instances (McLaughlin & Bennett, 2003). Flattening of the curve of Wilson helps reduce the frequency of non-working side and working side tooth interferences, provided there is adequate anterior disclusion (Dawson, 1988).  22 The 3-dimensional arrangement of dental cusps and incisal edges in the natural dentition has been described in the literature as spherical, as each tooth’s occlusal surface touches a segment making up a sphere. The Sphere of Monson, a non-scientific observation – a combination of Spee and Wilson curves in 3 dimensions, depicts the spherical arrangement of dental cusps and incisal edges in natural human dentitions, with the occlusal surfaces of all teeth touching a segment of the spherical surface (Ferrario et al., 2003).  Monson believed that in order to achieve well balanced geometric proportions to the face along with optimum function, the jaws can be related to a sphere whose radius is approximately four inches, with the center being equidistant from the occlusal surface of the teeth and the center of each condyle (Nam, Park, Lee, Ahn, & Lee, 2013). These concepts are identified in the criteria for case management forms and to score posttreatment dental casts and panoramic radiographs using the American Board of Orthodontics’ Cast-Radiograph Evaluation (CRE) (Casko et al., 1998, ABO Website, 2014).  Specifically, the alignment and buccolingual criteria of the CRE address the aforementioned curves (Casko et al., 1998).  The curve of Spee is considered in terms of anterior and posterior alignment in the sagittal plane, as well as marginal ridge height discrepancies (Casko et al., 1998).  Anterior alignment is verified by looking at incisal edges and labial-incisal surfaces (Casko et al., 1998).  Proper posterior mandibular alignment is assessed by measuring the buccal cusps of the premolars and molars, which are easily identifiable (Casko et al., 1998).  Marginal ridge heights of adjacent posterior teeth are supposed to fall within a maximum of 0.50 mm of the same level (Casko et al., 1998).  Buccolingual angulations of the posterior teeth are assessed with a  23 special step gauge to measure the height difference between buccal and lingual cusps in the posterior segments (Casko et al., 1998; Owens et al., 2006).   1.3.1.1 The Curve of Spee 1.3.1.1.1 Definition  In the late 19th century, Ferdinand Graf von Spee examined skulls with abraded teeth and characterized the natural phenomenon of the human occlusal curvature as viewed in the sagittal plane (Spee, Biedenbach, Hotz, & Hitchcock, 1980).  This curve changes over time, as the arch transitions from the primary dentition through to the permanent teeth (Marshall et al., 2008).  It is defined in the glossary of prosthodontic terms as the anatomic curve established by the occlusal alignment of the teeth, as projected onto the median plane (Prosthodontics, 1994).  The literature suggests that the development of this curve results from a combination of factors including the eruption of teeth, growth of the orofacial structures, and the concomitant development of the neuromuscular system (Marshall et al., 2008). Malocclusions with an exaggerated curve of Spee are frequently observed in dental conjunction with deep overbites (Simons & Joondeph, 1973).  As such, it has been postulated that an excessive curve of Spee alters the muscle imbalance, ultimately leading to improper functional occlusion (Kumar & Tamizharasi, 2012).  It has been proposed that an imbalance between the anterior and the posterior components of occlusal force can cause the lower incisors to overerupt, the premolars to infraerupt, and the lower molars to be mesially inclined, producing an accentuated curve (Kumar & Tamizharasi, 2012).    24 In the deciduous dentition, the curve of Spee is minimal, but changes in the early mixed dentition with the corresponding eruption of the mandibular permanent first molars and central incisors (Marshall et al., 2008).  According to Marshall’s longitudinal study using a sample from the Iowa Facial Growth Study, the curve of Spee depth increases significantly to a maximum depth of 1.32 mm relative to a perpendicular drawn from the most central point on the most erupted central incisor to the distobuccal cusp of the most posterior molar (Marshall et al., 2008). Findings indicate that the greatest increase in the curve of Spee occurs with the eruption of the mandibular second molars (Marshall et al., 2008).  This is one reason why the second molars need to be incorporated in orthodontic appliances (Casko et al., 1998; Marshall et al., 2008).  This value remains essentially unchanged until eruption of the second molars, where the depth increases to a mean maximum depth of 2.17 mm (Marshall et al., 2008).  In adolescence, the curve of Spee decreases slightly to a mean of 1,98 mm, and remains relatively constant in adulthood at a mean of 2.02 mm (Marshall et al., 2008). Marshall’s summary of findings include: 1. The occlusal plane in the deciduous dentition is relatively flat. 2. The largest increase in the maximum depth of the curve of Spee occurs during, and results specifically from, the differential eruption of the mandibular permanent first molars and incisors relative to the deciduous second molars. 3. The curve of Spee maintains this depth until the mandibular permanent second molars erupt above the occlusal plane, when it again deepens. 4. During the adolescent dentition stage, the curve decreases slightly and then remains relatively stable into early adulthood.  25 5. There are no significant differences in maximum depth of the curve of Spee between either the right and left sides of the mandibular arch or the sexes (Marshall et al., 2008). A recent cross-sectional study in Pakistan recorded a range of depths spread over different malocclusions ranging from 0.75mm-6mm, with a mean of 2.84 mm (Ahmed, Nazir, Erum, & Ahsan, 2011).  In Class I malocclusions, they found a mean depth of the curve of Spee to be 2.4mm in these cases (Ahmed et al., 2011).  1.3.1.1.2 Functional and Clinical Significance  Although several theories have been proposed to elucidate the presence of a curve of Spee in the natural dentition, its role during typical mandibular function has been doubted (Sicher, 1949; Dawson, 1988; Osborn, 1993; Carcara, 2001; Preston et al, 2008).  The functional significance of the curvature has been suggested to be biomechanical in nature (Marshall et al., 2008), although there is no physiologic evidence or definitive studies to elucidate this idea.  It has been suggested that an imbalance between the anterior and the posterior components of occlusal force can cause the lower incisors to overerupt, the premolars to infraerupt, and the lower molars to tip mesially (Gresham, 1957).  Its function has been described as being expressed during food processing by increasing the crush-shear ratio between the posterior teeth and the efficiency of occlusal forces during mastication (Marshall et al., 2008).  Spee himself suggested that the curve was the most efficient model enabling the teeth to remain in contact during the forward and backward motion of the mandible during chewing (Spee et al., 1980).  This was believed to be its primary functional implication, as it determined posterior tooth contact in protrusion or not, and its  26 relationship to anterior tooth guidance and condylar inclination.  The curve of Spee has also been described as having a direct effect on the normal functional protrusive movement of the mandible (Adaskevicius & Svalkauskiene, 2011).  The gnathology viewpoint states that the main purpose of the curve of Spee was to bring the occlusal surfaces of the upper and lower molars into a more ideal position for enabling optimal mastication (Schumacher, 1985).  The molar surfaces are considered to be included in the main force vector of the masticatory muscles, and therefore, the opposing molar placement into proper inclination and position to each other facilitates adduction, protraction, retraction and laterotrusion (Schumacher, 1985).  Leveling of the curve of Spee represents a routine procedure in orthodontics; the more pronounced the curve is, the greater the space requirements to flatten the dentition (Farella, Michelotti, Van Eijden, & Martina, 2002).  A deep curve of Spee results in a more confined space for the upper teeth, creating progressive spillage of the maxillary teeth mesially and distally (Andrews, 1972).  It has been related to incisor overbite, lower arch circumference, lower incisor proclination, and to a minor extent, craniofacial morphology (Farella et al., 2002; Simons & Joondeph, 1973).  Baydas et al studied the effect of the depth of the curve of Spee on overbite in a sample of 137 patients (Baydas, Yavuz, Atasarai, Ceylan, & Metin Dagsuyu, 2004).  He divided the sample for a basis of comparison into 3 groups; normal, flat and deep curve of Spee (Baydas et al., 2004).  The results demonstrated statistically significant correlations between the depth of the curve of Spee, overbite, and overjet (Baydas et al., 2004).  In El-Dawlatly et al’s analysis of the underlying components in deep overbite malocclusions, an exaggerated curve of Spee showed the highest contribution of all the observed dental components (78%) (El- 27 Dawlatly, Fayed, & Mostafa, 2012).   The amounts of overbite and overjet have a significant influence on the variation of the curve of Spee in the mandibular arch, whereby the greater the values are, the more pronounced the curve will be (Farella et al., 2002).  Orthodontic correction of the overbite often involves leveling the curve of Spee by anterior intrusion, posterior extrusion, or a combination thereof (Shannon & Nanda, 2004).  With regards to craniofacial growth and development, Andrews stated that there was a natural tendency for this curve to deepen with time because the lower jaws’ growth downward and forward sometimes was faster and continued longer than that of the upper jaw (Andrews, 1972).  This is suggested to cause the lower anterior teeth, confined by the upper anterior teeth and lips, to be forced back and up, resulting in crowding, a deeper overbite and a deeper curve of Spee (Baydas et al., 2004).  The value of the depth of the curve of Spee should be considered and quantified in space management procedures in order to prevent incisor flaring and consequently assuring esthetics, stability of the treatment results, and function (Ahmed et al., 2011).  Andrews also advocated for leveling the curve of Spee since he noted that the occlusal planes of nonorthodontically treated normal occlusions tended to be level, and he therefore associated the curve of Spee with postorthodontic treatment relapse (Andrews, 1972). Andrews noted that the most ideal intercuspation occurred when the occlusal plane was relatively flat (Andrews, 1972). While not every normal occlusion has a flat occlusal plane, and even though this only provides weak anectodal evidence for the need to level the curve of Spee, it has remained a mainstay of orthodontic treatment goals (Bernstein et al., 2007; Koyama, 1979).  A comparative analysis of the curve of Spee before and after orthodontic treatment found that in patients with pronounced overbites, a more severe  28 curve tends to appear after the retention period (Koyama, 1979).  When the curve of Spee in untreated subjects was compared with post-orthodontic patients after retention, the treated cases were found to have either reverse curves of Spee or straight and leveled occlusal planes (Koyama, 1979).  On the other hand, the control group’s curve of Spee was described as smooth and slight (Koyama, 1979).  Another school of thought suggests that straightening out the curve of Spee through the use of the straight edgewise orthodontic wire technique can damage the functional organization of teeth and muscles, causing a complex dysfunction of the masticatory system and the craniomandibular system (Schumacher, 1985). Functionally, the curve of Spee permits total posterior disclusion upon mandibular protrusion, given proper anterior tooth guidance (Lynch & McConnell, 2002).   However, in order to establish proper incisor relationships and posterior disclusion in excursive movements, the curve must be relatively mild (Shannon & Nanda, 2004).  With increasing age, there is a significant change observed in the curve of Spee, resulting in a decrease in posterior disclusion during mandibular protrusion (Ahmed et al., 2011).  Hence, as patients grow older, clinicians should be aware that the occlusal adjustments with age have gradually altered the curve of Spee of youth toward a more favorable individual occlusal curvature (Kumar & Tamizharasi, 2012).  Thus, if the curve of Spee is not maintained in these dentitions during full mouth rehabilitation, it may lead to interferences along the mandibular movements which will jeopardize the health of the masticatory system (Kumar & Tamizharasi, 2012).  Analysis of the curve of Spee can assist dentists in determining the sagittal arrangement of the teeth (Ahmed et al., 2011).  The curve of Spee can also be used as a reference point for prosthodontic  29 restorations to achieve stability in complete denture cases and implant-supported rehabilitations (Ahmed et al., 2011).  1.3.1.1.3 Methods of Measurement: Strengths and Limitations Clinically, the curve of Spee can be roughly evaluated on clinical exam by comparing the distal marginal ridges of the most posterior teeth in the arch to the incisal edges of the central incisors (Germane et al., 1992).  This qualitative value of the curve of Spee is useful as a severity gauge upon visual examination, but is not precise.   Various versions of curve of Spee measurements exist in the literature (Germane et al., 1992; S. Braun et al., 1996; Baydas et al., 2004; El-Dawlatly, Fayed, & Mostafa, 2012).  A more accurate method of measuring the curve of Spee is done using study models in the lab, as in the Baydas’ study (Baydas et al., 2004).  Baydas measured the depth of curve of Spee as the perpendicular distance between the deepest cusp tip and a flat plane that was laid on top of the mandibular dental cast, touching the incisal edges of the central incisors and the distal cusp tips of the most posterior teeth in the lower arch (Baydas et al., 2004).  The mean value from right and left measurements was used as the total depth of curve of Spee (Baydas et al., 2004). El-Dawlatly defined his measurement of the curve of Spee by the line formed between the deepest point on the mandibular buccal segment and a horizontal line formed between the most overerupted mandibular incisor and the most overerupted molar (El-Dawlatly, Fayed, & Mostafa, 2012).  Another method is to measure each cusp tip to the occlusal plane, from the distobuccal cusp tip of the second molar proceeding around the arch, recording each successive cusp tip coordinates of the first molars, premolars, and canines (S. Braun et al., 1996)  The center point of each incisal edge was recorded for each of the four incisors (S. Braun et al.,  30 1996).  Past literature has shown the absence of sexual dimorphism in the curve of Spee (Ahmed et al., 2011; M. L. Braun & Schmidt, 1956; Farella et al., 2002). 1.3.1.2 The Curve of Wilson 1.3.1.2.1 Definition The curve of Wilson is the mediolateral curve that contacts the buccal and lingual cusp tips of each side of the arch (Dawson, 1988).  This transverse occlusal curve exists normally for each pair of right and left-side teeth and is concave upwards in the maxillary arch and convex downwards in the mandibular arch (Kaifu, Kasai, Townsend, & Richards, 2003).  It results from the inward inclination of the lower posterior teeth, making the lingual cusps lower than the buccal cups on the mandibular arch; the buccal cusps are higher than the lingual cusps on the maxillary arch because of the outward inclination of the upper posterior teeth (Dawson, 1988).  The curve of Wilson considers the tooth arrangement in the bucco-lingual plane, which influences lateral excursive movements (Lynch & McConnell, 2002). 1.3.1.2.2 Functional and Clinical Significance Clinically, it is the curve viewed in the frontal plane at the level of the molars (Ferrario et al., 2003).   When viewed from a frontal aspect, the mandibular molars have a slight lingual inclination and the buccal cusps of these teeth are higher than the lingual (Lynch & McConnell, 2002).  This design serves a masticatory function, as it permits easy access to the occlusal table.  As the tongue lays the food on the occlusal surfaces, it is stopped from going past the chewing position by the taller buccal cusps (Dawson, 1988).  The curve of Wilson configuration facilitates lateral  31 excursions free from posterior interferences (Lynch & McConnell, 2002).  With occlusal function, the curve of Wilson tends to change from an upwardly concave curve to a flat or even helicoidal anteroposterior curve as a result of natural tooth wear (Ferrario et al., 2003).   During orthodontic treatment, evaluating the original cuspid position and the curve of Wilson in the lower arch is important in determining the correct lower archform. By the finishing stage of treatment, the lower archform should be accurately established in the rectangular arch wire (Casko et al., 1998; McLaughlin & Bennett, 2003; Owens et al., 2006). 1.3.1.2.3 Methods of Measurement: Strengths and Limitations In the ABO CRE, the buccolingual inclination of the mandibular teeth is assessed by using a flat surface that is extended between the occlusal surfaces of the right and left posterior teeth (Casko et al., 1998).  In this position, the straight edge should ideally contact the buccal cusps of contralateral lower molars (Casko et al., 1998).  An acceptable measurement is to have the lingual cusps within 1 mm of the surface of the straight edge (Casko et al., 1998).   Thus, the ABO de facto implies that leveling the curve of Wilson is an important orthodontic outcome measure.    32  In summary, malocclusion is defined as any disharmonious variation from the accepted or theoretical normal arrangements of the teeth (Grainger, 1967). But, in nature some degree of variation among individuals of a species is always present (Grainger, 1967).  The functional goals of orthodontic treatment with regard to occlusal dynamics include a canine rise pattern of function with anterior guidance and no cuspal interferences to wear down (a leveled curve of Wilson) (Casko et al., 1998).   1.3.1.3 Curves of Spee and Wilson in Serial Extraction  Occlusal curvatures in the human dentition are a naturally occurring phenomenon (Ahmed, 2011).  The most likely explanation for the development of these curves most likely results from a combination of factors including eruption of teeth, growth of the craniofacial structures, and development of the neuromuscular system (Marshall, 2008).  Although leveling these curves is ubiquitous in orthodontic practice, little research has been dedicated to the examination of the relationship between the depth of the curves of Spee and Wilson to early interceptive treatments, such as SE.  The value of the occlusal curves (Spee specifically), could be important considerations in space management procedures in order to control lower incisor position and thereby ensure esthetics, treatment stability and function (Ahmed, 2011).  The two major occlusal morphological descriptors, namely the curve of Spee and the curve of Wilson have yet to be evaluated in terms of how they are affected secondary to SE.  SE is most typically a treatment chosen in Class I mixed dentition cases with severe TSALD to establish a stable orthodontic outcome in harmony with the surrounding tissues (Kjellgren, 1947; Hotz, 1970; Graber, 1971; Boley, 2002; Proffit, Fields, & Sarver, 2012).  SE, as a treatment will not help with abnormal jaw relationships, i.e. Class II and III  33 discrepancies (Graber, 1971).    Several difficulties occur post SE including problems with axial inclinations, space closure, deeper overbite, more rotations, and torque issues (Graber, 1971).  All of these can contribute to altered occlusal curvatures, and potentially increase the comprehensive phase of treatment in SE patients. Do these curves worsen and become more pronounced as the teeth drift into the extraction spaces? Graber described the occlusal curvature in the mandibular arch as a concave arc, and that the long axes of the premolars and molars diverge (Graber, 1971).   In terms of the effects of SE, Graber observed that it encouraged the crowns to tip together, which further emphasized the “ditch” (Graber, 1971).  This ditch, which occurs between the mandibular canine and second premolar (after extraction of the first premolar), leads to a large interradicular distance, which necessitates correction by means of fixed appliances (Graber, 1971).  How does this affect future treatment with fixed appliances? Does it lengthen treatment and/or make the mechanics more difficult?  The success of SE treatment is based on resolving TSALD and allowing permanent canines and premolars to erupt into keratinized tissue appropriately over the basal bone.  It is also based on treating early to avoid more difficult future treatment.  Keeping the notion of guidance of eruption in mind, do these permanent teeth subsequently erupt at the expense of deepening occlusal curves?  Could this possible sequela negate the treatment time saving advantage advocated by SE practitioners (leveling takes up significant active treatment time)?  As such, the long-term effects of SE on the curves of Spee and Wilson should be considered by orthodontists looking to implement this interceptive treatment.  In terms of cost versus benefit, for both the orthodontist and the patient, these concepts along with other arch  34 changes should be quantified to deepen our understanding of how and where and when the teeth are moving.   35 Statement of the Problem  1. Under optimal conditions, SE’s are a tool used to manage arch length and tooth alignment irregularities in the mixed dentition, while LPE’s are used when a patient presents with moderate-severe crowding or bimaxillary protrusion in the permanent dentition. 2. The occlusal curvatures both have functional and anatomical implications with tooth and arch forms, and presently, there is no evidence as to how the different occlusal curvatures quantitatively change as a result of SE.  3. SE requires children to have multiple rounds of extractions, justifiable if it shortens active treatment. Presently, SE only saves 4-6 months. This minimal gain could be explained by the extra time needed to correct the particular occlusal curvature it produces.  36 1.4 Objectives  The purpose of this study was to address the following question: If we treat by means of serial extractions, and in doing so, knowingly place a child through multiple rounds of extractions, why does it not seem to save more than four-six months of active treatment time (O'Shaughnessy et al., 2011)?  Can it be that this interceptive process results in altered mandibular occlusal curves and tipping patterns, which require more time to correct orthodontically?  In addition, the study’s aim was to: 1. Serially measure changes in the occlusal curves of untreated mandibular dentitions following SE and compare these to Class I control subjects and LPE subjects. 2. Serially measure changes in tip of the mandibular teeth following SE and compare them to Class I control subjects and LPE subjects.  37 1.4.1 Null Hypotheses  1. There are no differences in the measured mandibular occlusal curvatures between Class I serial extraction cases versus untreated Class I controls at T0 and T1, and between late premolar extraction cases at T1 and T2.  2. There are no differences in the specific tooth (incisor, canine, first molar) movements between Class I serial extraction cases versus untreated Class I controls at T0 and T1, and between late premolar extraction cases at T1 and T2. 1.4.2 Study Hypotheses 1. Significant difference to be observed in the SE group at T1 (post drift for the SE group) versus the untreated controls and LPE groups (not yet extracted) 2. No difference between controls and LPE at T1 3. No difference between SE and LPE at T2 (after comprehensive orthodontic treatment) 4. The incisors to be more upright (obtuse angle) at T1 in the SE group 5. The angle of the incisors and canines to become more proclined (acute angles) at T2 in the SE group 6. Proclined incisors (acute angle) at T1 for the LPE group, which are uprighted (obtuse angle) at T2 7. The angle of the incisors, canines, and molars to stay relatively consistent in all three time points for the control group 38                                 Chapter  2: Methodology  2.1 Materials and Methods This retrospective case control study was reviewed and approved by the Research Ethics Board at the University of British Columbia (H12-02568).  The sample selected consisted of a total of 90 subjects.  Treatment records for 30 SE and 30 LPE patients treated between 2007 and 2012 in a specialty pediatric-orthodontic practice were screened using general inclusion criteria to identify potential serial extraction cases: Angle Class I, extraction of four permanent teeth, one in each quadrant and the availability of pre and post treatment records.  The control group of 30 Class I cases was made available via the longitudinal sample from Oregon Health and Sciences University, and courtesy of the AAO Legacy Fund. The inclusion criteria included untreated, Angle Class I cases with complete records.  Records for the control sample were taken every 6 months.     2.1.1 Subjects Each sample group contained 15 males and 15 females.  The 30 subjects included in the SE group had premolars extracted before the permanent canines erupted or enucleated if the canines erupted first, and a minimum of one year physiological drift occurred following extractions.  The two orthodontists in the practice jointly treatment planned all the cases from 2007-2012 and used consistent treatment criteria to plan extractions for all patients.  A mixed dentition analysis was not routinely performed although the majority of cases that received SE had 8 mm of  39 crowding or more per arch based on visual examination of the mandibular casts.   The sample selection process included going through binders of case summaries identifying the Angle classification, any anomalies/missing teeth, and the method of treatment.  The following exclusion criteria were imposed: no anteroposterior discrepancies, no decreased vertical facial heights, no anomalies (missing teeth, clefting, etc.), and no excessively deep bites.  The orthodontists prescribed SE if the subjects had significant crowding and were referred to the practice while in the mixed dentition (O'Shaughnessy et al., 2011).  The 30 subjects comprising the LPE group were selected from the same private practice.  This treatment protocol was prescribed when subjects presented later on in development, in the permanent dentition with severe crowding.  Extractions occurred after the permanent dentition was partially or fully erupted and fixed appliances were placed no longer than 3 months after extraction (no drift).   The 30 subjects making up the control group were matched to the 30 SE subjects at T0, initial records, according to gender and dental age, as determined by the Demirjian method (Demirjian & Goldstein, 1976; Demirjian, Goldstein, & Tanner, 1973). Subjects in the two treatment groups were only matched for gender, and not matched for severity of initial malocclusion as the developmental timing did not coincide. 2.1.2 Data Collection and Measurements There were two sets of data: mandibular dental casts and cephalometric radiographs.      40   2.1.2.1 Cast Measurements In this study, the mandibular dental casts were digitized using the Ortho Insight 3D™ scanner, and analyzed from 90 subjects (30 with SE, 30 with LPE, 30 controls).  The mandibular dental casts and their corresponding lateral cephalometric radiographs were digitized and measured in random order. For the SE group, data was analyzed at 3 time points: T0 - Initial records, before the extraction of the canines and first molars as well as the first permanent premolars (for which a control case was dentally matched to), T1 – post-drift, at the end of the observation period, prior to comprehensive treatment, T2 - after the completion of orthodontic treatment (final records).   The LPE group casts were analyzed at 2 time points: T1 - Initial records (permanent dentition, with the second molars partially or fully erupted), prior to 4 premolar extractions, T2 - after the completion of orthodontic treatment (final records). Finally, the control group casts were analyzed at 2 time points: T0 - for which an SE case was dentally matched to, and T1 – 2-3 years later to mirror the SE group during the period of drift.  In total, 210 casts were digitized and measured. The three-dimensional analysis was obtained utilizing Rhinoceros NURBS modeling for Windows version 4.0 (Robert McNeel & Associates, 2010).  The digital casts were first converted from the scanner into stereolithography files and imported into Rhinoceros.  The following points were identified: the vertices of the first molar cusps                                                   MotionView 3D LLC, 2730 Kanasita Dr, Hixson, TN 37343  Robert McNeel & Associates, 3670 Woodland Park Ave N,
Seattle, WA 98103   41 (DB, MB, DL, ML), the incisal edges of 31, 32, 41, and 42.  Three different least-squares spheres of best fit were produced and their radii used as the outcome measure.  The radii of the spheres were used to correspond to the depth of the occlusal curvatures in 3-dimensions; therefore the steeper the curvature, the smaller the radius.  The measured outcomes were from the following point schemes:  1. All of the digitized points (DB, MB, DL, ML of the lower first molars, and the incisal edges of the four lower incisors) to represent Monson’s sphere. 2. Only the DB, MB, DL, ML of the lower first molars to represent the curve of Wilson (mediolateral curve, viewed in the frontal plane). 3. The DB + MB cusps of the lower first molars and the four lower incisors to represent the curve of Spee (anteroposterior curve, viewed in the sagittal plane).        Figure 2.1. Rhinoceros’ digitized representations of Monson’s sphere, curve of Wilson, and curve of Spee A sphere of best fit is typically determined by means of an optimization function which minimizes the sum of ((x-xc)^2 + (y-yc)^2 + (z-zc)^2 – r^2)^2, where x,y,z are the point coordinates, xc,yc,zc are the sphere's center, and r is the sphere’s radius.  A measurement was added to quantify the amount of molar tip or roll.  A  42 reference plane through the mesiobuccal (MB) and distobuccal (DB) cusp tips of both right and left first molars was drawn and then from this plane we measured the linear distances of the mesiolingual (ML) cusps and the distolingual (DL) cusps to the plane.    2.1.2.2 Cephalometric Measurements The cephalometric radiographs of the subjects were digitized using Rhinoceros and a custom analysis as per Yoshihara (Yoshihara et al., 2000).  The palatal plane served as a reference plane from which tooth tip/inclination could be measured. Connecting the anterior nasal spine (ANS) and posterior nasal spine (PNS) formed the palatal plane.  Using palatal plane as the reference, the following 3 angles were calculated to determine tooth tip (Figure 1):  1. Palatal plane to the long axis of the mandibular incisor 2. Palatal plane to the long axis of the mandibular canine 3. Palatal plane to the axis formed by the line connecting the mesiobuccal cusp and mesial root apex of the first molar (Yoshihara et al., 2000).    43   Figure 2.2. Measurement of tipping of incisor, canine and molar represented by changes in angulation relative to palatal plane (ANS-PNS) at T0, T1, and T2  2.1.3 Statistical Analysis Initially, a power calculation was done for detecting differences in the curve of Spee, using the averages from the Marshall article (Marshall, 2008).  Assuming equal variances of 0.77 mm as per Marshall (Marshall, 2008), and a mean difference of 0.6, an 80% power and an =0.05  26 patients would be needed in each group (n1=n2=26).  The post-hoc power calculation for our first objective, assuming the effect size in the sample is equal to the effect size in the population, yielded a power of 62% for overall radii, (required sample size 45), 91% for radii of lower molars (required sample size of 21) and 10% for the radii of lower incisors and buccal cusps of the molars (required sample size of 471).  For the second objective, from the cephalometric analysis, the post-hoc  44 power for incisor angulations was 23% (required sample size of 153), 27% for the canine (required sample size of 125) and 68% for the molar angulations (required sample size of 39).  The SPSS statistical software program version 21.0 was used for all statistical analyses.  Bivariate analyses were used to analyze the following: 1. test the quality of matching between the controls and SE at T0, 2. determine the intra-examiner reliability of the measurements (intra-class correlation) and 3. compare the means of the different measurements for a) cast data b) cephalometric data, between the SE patients and their matched controls at T0, between SE, their matched controls and the LPE group at T1, and finally between SE and LPE at T2 (independent sample t-tests or ANOVA with Bonferroni Post Hoc adjustment). Multiple regression analyses were done to evaluate the joint effects of multiple factors. For all statistical tests, the significance level was set at 5% (p <0.05).  2.2 Reliability testing  To evaluate intra-examiner reliability, 10 mandibular casts and 10 cephalometric radiographs were randomly selected from the overall sample for duplicate measurements.  The intra-class correlation coefficient for double recordings of radii in the 10 casts ranged from 0.711 to 0.834 indicating the reliability of these measurements was satisfactory.  For the cephalometric angular measurements, the intra-class correlation coefficient ranged from 0.808 to 0.956 and this also indicates satisfactory intra-examiner reliability.  45 The potential machine error, as a result of using two Ortho Insight 3D™ laser scanners (one at Oregon Health Sciences University and one at a private practice in Vancouver, BC) was considered by measuring the same dental cast in both scanners and then choosing unambiguous landmarks on the virtual casts and comparing linear distance measurements made between them.  As reported by the Motion View representative, both machines used in this study are identical and according to the manufacturer, the digitizer has a resolution of 40 microns between points in the point cloud. Consequently, measurement bias due to machine error was unlikely.  In addition, measuring the casts in a random order so that the same patient’s pre-treatment and treatment completion casts were not measured consecutively to reduce the potential for operator bias. Other negligible errors could be derived from the approximation algorithms.    2.3 Laser Scanner Ortho Insight 3D, Motion View Software LLC 3D laser scanner is ABO compliant. This machine is able to scan full arch impressions, plaster models, and/or bite registrations to make virtual plaster 3D models in the computer.  The several advantages include: (1) efficiency of having patient records instantly accessible on a computer screen, (2) decreased requirements for storage space, (3) accuracy, efficiency, and ease of measurement of tooth and arch sizes and dental crowding, (4) accurate and simple diagnostic set-ups (5) the ability to communicate these virtual images to other specialists, (6) objective, rather than subjective model grading analysis for ABO certification(Stevens et al., 2006).  Validation studies comparing standard plaster casts  46 to their digitized counterparts have revealed that scanned models are not a compromised choice and are diagnostic (Kusnoto & Evans, 2002; M. F. Leifert, Leifert, Efstratiadis, & Cangialosi, 2009; Redlich et al., 2008; Stevens et al., 2006).  The dimensional accuracy of laser scanned digital models has been shown to be within 0.05 mm (Kuroda, Motohashi, Tominaga, & Iwata, 1996; Kusnoto & Evans, 2002; Motohashi & Kuroda, 1999; Sohmura, Kojima, Wakabayashi, & Takahashi, 2000).   47  Chapter  3: Results 3.1 Testing of matching of study groups The accuracy of matching between the control and SE groups at T0 was assessed using a Chi Square test for gender proportions (Table 1). Dental age matching was evaluated using two tests, a correlation analysis along with an independent t-test. The gender matching among the groups was satisfactory as no statistically significant gender-based differences were observed between the groups (Chi Square, P = 1.000).  In terms of dental ages, there was a high degree of correlation among the control and SE groups at T0 (Pearson Correlation Coefficient = .981, P<0.001, 95% CI: 0.73; 0.77).  This satisfactory matching was also further verified by the results of an independent t-test, i.e. no significant difference in the mean ages and their variance was observed between the 2 groups (P=0.959), indicating that the age-dependent matching was successful.                 48 Table 3.1. Matching of Study Groups  Groups Gender Total Males       N (%) Females N (%) N (%) Control 15 (50%) 15 (50%) 30 (100%) Serial Extractions 15 (50%) 15 (50%) 30 (100%) Late Premolar Extractions 15 (50%) 15 (50%) 30 (100%) Chi Square Test, P=1.000  Dental Age Mean ±SD Control 10.14±1.50 Serial Extractions 10.16±1.44 Independent Sample t-test P=0.959  Intra-class Correlation Coefficient 0.981; P=0.001  3.2 Preparing for Data Analysis – assessing data for normality  In order to choose appropriate statistical tests (i.e. to know if the data requires parametric or nonparametric tests), the outcomes distribution at T0, T1 and T2 were assessed for normality. The results of the skewness and kurtosis measures fell below ±1.96, and therefore, parametric statistics were chosen for all subsequent analyses.    3.3 Analyses of Cast Data  The three spherical variables were measured from the 3-D cast data for the control group at T0 and T1 using Rhinocerus (Figure 3.1).  The paired sample t-test for the control group cast data showed a non-statistically significant increase with overall radii (P = 0.107), a statistically significant increase with respect to radii of the first  49 lower molars (P = 0.022), and a statistically non-significant decrease in radii of the lower buccal cusps, first molars and incisors (P = 0.225).  Figure 3.1.  Control group radii from T0-T1  The paired sample t-test for the SE group from T0 to T1 (Figure 3.2) a statistically significant decrease with overall radii (P = 0.001), a statistically significant increase with respect to radii of the first lower molars (P = 0.031), and a statistically significant decrease in radii of the lower buccal cusps, first molars and incisors (P = 0.044).   From T1 to T2, the SE group had a statistically significant increase with overall radii (P = 0.001), a statistically significant increase with respect to radii of the first lower molars (P = 0.001), and a statistically non-significant decrease in radii of the lower buccal cusps, first molars and incisors (P = 0.001).   Figure 3.2.  SE group radii from T0-T1-T2   50 The LPE group’s paired sample t-tests (Figure 3.3) demonstrated a non-statistically significant increase with overall radii (P = 0.308), a statistically significant increase with respect to radii of the first lower molars (P = 0.035), and a statistically significant decrease in radii of the lower buccal cusps, first molars and incisors (P = 0.001).  Figure 3.3. LPE group radii from T1-T2 At T0, the three radii outcomes were measured for the control group and the SE group (Table 3.2).  This was to test for any potential baseline differences between the groups prior to the intervention, serial extractions, in our experimental group.  The differences between the SE and control groups at T0 were statistically significant for two out of three measurements, namely T0radiusoverall  (included all 12 landmarks) and T0radiuslower6s (P<0.05).  Therefore, this baseline difference at T0 was considered in the multivariate regression analyses.  The third measure, T0radiusDBMBincisors, although not statistically significant, produced one outlier in the control group (Figure 3.4).   51 Table 3.2. Cast Radii at TO (baseline) comparisons between Serial Extractions and Control groups  Outcome:  CAST RADII T0 Controls Serial Extractions  P values (95% CI) N Mean ± SD N Mean ± SD T0radiusoverall 30 51.00±3.25 30 53.00±3.42 0.024 (0.27; 3.72) T0radiuslower6s 30 57.78±4.22 30 53.79±4.80 0.001 (1.66; 6.33) T0radiusDBMBincisors 30 30.19±3.85 30 31.00±4.95 0.484 (-3.10; 1.48)    Figure 3.4. Comparison of radii in controls and SE at T0  Table 3.3 describes the comparison of the mean radii among the three groups, which had data available to be measured at T1.  The time point described by T1 occurs post extraction and tooth drift in the SE group, with no intervention yet in the LPE group; thus we would expect similar measurements in the SE and LPE groups at T1 to the control group. There were statistically significant differences in mean radii among the three groups for all cast data measures at T1.  The data analysis showed that the mean radii values obtained for the control group was similar to the LPE group for all outcomes, except for T1radiusDBMBincisors.  For this particular measure, the SE group was more similar to the control than to the LPE group (Figure 3.5).   52  Table 3.3. Comparing at T1 (after intervention in SE): controls, SE and LPE groups (ANOVA with Bonferroni post hoc adjustment) Outcome:  CAST RADII T1 Controls Serial Extractions Late Premolar Extractions  P values  N Mean ± SD N Mean ± SD N Mean ± SD T1radiusoverall 30 54.10±3.76 30 48.41±3.44 30 55.14±5.02 0.001 T1radiuslower6s 30 59.34±4.43 30 52.20±4.44 30 59.79±6.35 0.001 T1radiusDBMBincisor 30 29.02±3.75 30 29.12±4.19 30 32.22±6.95 0.028   Figure 3.5. Comparison of radii in controls, SE, and LPE at T1  The results of the ANOVA with Post Hoc Bonferroni adjustment are presented in Table 4. This test, which compares every group to each other, showed that the LPE and controls had similar means in outcome measurements.  All of the P values for comparisons at T1 for the LPE and controls were non-significant (Table 3.4).  At this time point (T1), we measured the initial records for the LPE group (pre-extractions), which served as another control at T1.  The SE group differed significantly from the other two groups with respect to T1radiusoverall and T1radiuslower6s, while T1radiusDBMBincisors did not differ across all three groups.  53  Table 3.4. Multiple comparisons of cast radii at T1 between controls, SE, and LPE  Outcome: CAST RADII AT T1 ANOVA with Bonferroni post hoc adjustment  P values (95% CI) Group Comparisons Controls versus Serial Extractions Controls versus  Late Premolar Extractions  Serial Extractions Late Premolar Extractions T1radiusoverall 0.001 (3.09; 8.30) 1.000 (-3.64; 1.57) 0.001 (-9.33; -4.13) T1radiuslower6s 0.001 (3.89; 10.39) 1.000 (-3.70; 2.80) 0.001 (-10.84; -4.35) T1radiusDBMBincisors 1.000 (-3.35; 3.16) 0.056 (-6.44; 0.06) 0.067 (-6.35; 0.15)  Time point T2 considered the LPE group as the control group, enabling us to better visualize the finished result after comprehensive orthodontics and to be able to compare if any differences existed at this stage.  Significant differences between the SE and LPE groups were observed (Table 3.5) in the same outcomes as at T1, T2radiusoverall and T2radiuslower6s. Following the same trend again T2radiusDBMBincisors was similar for both groups (P=0.500).   54 Table 3.5. Cast Radii at T2 (post orthodontic treatment) - comparisons between SE and LPE groups  Outcome: CAST RADII T2 Serial Extractions Late Premolar Extractions P values (95% CI) N Mean ± SD N Mean ± SD T2radiusoverall 30 52.31±3.48 30 55.85±3.52 0.001 (-5.34; -1.73) T2radiuslower6s 30 57.94±4.17 30 61.88±4.00 0.001 (-6.06; -1.83) T2radiusDBMBincisor 30 25.45±3.09 30 26.00±3.19 0.500 (-2.17; 1.07)  Multiple linear regression (MLR) models were used to assess the joint effect of multiple predictors associated with the differences observed.  The purpose was to evaluate the effect of serial extractions while adjusting for other factors.  At T1 (Table 3.6) we ran three MLR models, one for each outcome (overall radii, radii of 1st lower molars at T1, and radii of lower B cusps of 1st molars and incisors at T1) and adjusted for gender and the radii measures at T0.  In all three MLR models, the radii measures at T0 were associated with the radii measurements at T1.  For the model with the overall radii at T1 as the dependent variable, the overall model was significant (P=0.001) and the predictors explained 55.1% of the variance in the overall radii at T1.  The group dependence (=-0.490, P=0.001) and the baseline T0 (=0.461, P=0.001) were strongest predictors of the T1 outcome.  For the model with radii of 1st lower molars at T1 as an outcome, the overall model was statistically significant (P=0.001) and this model explained 64.7% of the variance in T1 radii of 1st lower molars.  Both, the group dependence  (=-0.399, P=0.001) and the baseline T0 measure (=0.570, P=0.001)  55 were the strongest predictors in this MLR model.   The model with radii of lower B cusps of 1st molars and incisors at T1 was not statistically significant (P=0.154).     56  Table 3.6. Predictors of variance in radii at T1 – Multiple Linear Regression (MLR) models Predictors Standardized Coefficients Significance Overall Radii at T1 Controls versus Serial Extractions -0.490 0.001 Gender 0.070 0.442 Overall radii at T0 0.461 0.001 Model summary: Adjusted R square = 0.551; P= 0.001 Radii of 1st lower molars at T1 Controls versus Serial Extractions -0.399 0.001 Gender 0.037 0.640 Radii of 1st lower molars at T0 0.570 0.001 Model summary: Adjusted R square =0.647; P= 0.001 Radii of lower B cusps of 1st molars and incisors at T1 Controls versus Serial Extractions -0.016 0.900 Gender -0.040 0.759 Radii of lower B cusps of 1st molars and incisors at T0 0.291 0.028 Model summary: Adjusted R square =0.040; P= 0.154  More details about outcome at T2 can be found in Figure 3.6, the comparison between the SE and LPE group yielded one outlier in each of the three radii outcomes.  The LPE group had the same subject as an outlier (one individual case) in overall radii and radii of first lower molars, while the SE group had one outlier for the final measure, T2radiusDBMBincisors.  Statistical significance was observed in the first two radii  57 measurements, with the LPE’s exhibiting consistently larger radii in both T2overallradius and T2radiuslower6s.   There was no statistical difference for T2radiusDBMBincisors between the two groups.   Figure 3.6. Comparison of radii in SE and LPE at T2   The regression analysis at T2 for the SE versus the LPE group yielded three statistically significant MLR models (Table 3.7). The MLR for T2overallradius (P=0.001) showed that 39.5% (Adjusted R square) of the observed variance could be explained by T1overallradius (=0.451, P=0.001), followed by gender (=-0.222, P=0.040), and then the group dependence effect (=0.176, P=0.185).  With regards to T2radiuslower6s (outcomes), T1radiuslower6s (=0.348, P=0.013), and gender (=-0.266, P=0.019) were the strongest predictors and this MLR model explained 36.6% of the variance in T2radiuslower6s.  The group dependence had some effect but this was not statistically significant (=0.240, P=0.069).  The final MLR model for T2radiusDBMBincisors, revealed that 11.3% of the variance observed in the radii of lower B cusps of 1st molars and incisors could be explained by T1radiusDBMBincisors (=0.398, P=0.003).  58  Table 3.7. Predictors of variance in radii at T2 –MLR models Predictors Standardized Regression Coefficients Significance Overall Radii at T2 SE versus LPE 0.176 0.185 Gender -0.222 0.040 Overall radii at T1 0.451 0.001 Model summary: Adjusted R square = 0.395; P= 0.001 Radii of 1st lower molars at T2 SE versus LPE 0.240 0.069 Gender -0.266 0.019 Radii of 1st lower molars at T1 0.348 0.013 Model summary: Adjusted R square =0.366; P= 0.001 Radii of lower B cusps of 1st molars and incisors at T2 SE versus LPE -0.017 0.898 Gender 0.029 0.817 Radii of lower B cusps of 1st molars and incisors at T1 0.398 0.003 Model summary: Adjusted R square =0.113; P= 0.021     59  The results for molar tip for the control group showed statistically significant decreases in both ML (P=0.016) and DL  (P=0.045) cusp tip linear distances from the reference plane formed by the MB and DB cusp tips of both right and left first molars.    Figure 3.7. First molars’ ML and DL cusps’ distance from the reference plane (MB and DB first molar cusp tips) from T0-T1 in the control group  The molar tip measurements in the SE group showed non-statistically  significant increases in lingual tip of the ML (P=0.127) and DL (P=0.823) cusps from T0-T1.  However, statistically significant decreases in the ML (P=0.001) and DL (P=0.001) cusps were observed from T1-T2.  Figure 3.8. First molars’ ML and DL cusps’ distance from the reference plane from T0-T1-T2 in the SE group   60  In the LPE group, the molar tip from T1-T2 had statistically significant decreases in both ML (P=0.001) and DL (P=0.030) cusp lingual tips.  Figure 3.9. First molars’ ML and DL cusps’ distance from the reference plane from T1-T2 in the LPE group  The observations at T0 for molar tip between the control and SE groups revealed statistically significant differences for both the ML (P=0.003) and DL (P=0.001) cusp tips.  The SE group had significantly larger linear distances from the reference plane as compared to the control group.  Figure 3.10. Comparison of first molars’ ML and DL cusps’ distance from the reference plane at T0 in controls and SE  The ML cusp tip values at T1 among the controls, SE, and LPE groups were statistically significant between the controls and SE (P=0.001), and  the SE and LPE groups  61 (P=0.001).  However, a non-statistically significant difference in ML values was observed between the controls and LPE (P=0.632).  The SE and LPE groups each had one outlier.  The DL cusp tip values at T1 revealed the same pattern, statistical significance was observed between the controls and SE (P=0.001), and  the SE and LPE groups (P=0.001).  Similar to the ML outcome, a non-statistically significant difference in DL values was observed between the controls and LPE (P=0.241).  Three outliers were present in the SE group for the DL cusp tip outcome and one in the LPE group.    Figure 3.11. Comparison of first molars’ ML and DL cusps’ distance from the reference plane at T1 in controls, SE, and LPE   After orthodontic treatment, at T2, the ML and DL molar cusp tip outcomes had no statistically significant differences (P=0.056 and P=0.117, respectively).   62  Figure 3.12. Comparison of the first molars’ ML and DL cusps’ distance from the reference plane at T2 in SE and LPE At T1 we ran two MLR models, one for each outcome (the mean linear distance of the lower first molars’ ML cusp tips to the reference plane and the mean linear distance of the lower first molars’ DL cusp tips to the reference plane) and adjusted for gender and the linear measurements at T0.  In both MLR models, the baseline measures at T0 were associated with the linear distance measurements at T1.  For the model for T1ML as the dependent variable, the overall model was significant (P=0.001) and the predictors explained 82.4% of the variance in the outcomes for the ML cusp linear distances at T1.  The baseline T0 outcomes (=0.670, P=0.001), and the group dependence (=0.289, P=0.001), were the strongest predictors of the T1 outcome.  For the model with the DL cusp tips at T1 as an outcome, the overall model was statistically significant (P=0.001) and this model explained 83.1% of the variance in T1 linear distances of the DL cusp tips to the reference plane formed by all eight of the first molars cusp tips.  Both, the baseline T0 measure (=0.690, P=0.001) and the group  63 dependence (=0.218, P=0.016) were the strongest predictors in this MLR model.   The gender effect in both models was not statistically significant.  The regression analysis at T2 for the SE versus the LPE group yielded one statistically significant MLR model (P=0.001). The MLR for linear distance from the ML cusp tips of the first molars showed that 50.3% (Adjusted R square) of the observed variance could be explained by the T1ML outcome (=0.309, P=0.023), followed by the gender (=0.243, P=0.044).  With regards to T2DL outcomes, the MLR was not statistically significant (P = 0.090).    3.4 Analyses of  Cephalometric Data Looking across the time points for each group, the following results can be extrapolated: the control group (Figure 3.13) incisor average angle decreased (tipped) minimally from T0-T1 (P = 0.869), the canine uprighted by +2.55 degrees (P = 0.001), and the molar uprighted slightly, by 0.37 degrees (P = 0.497). The SE group (Figure 3.14), from T0-T1, showed a tendency for uprighting of the incisor and canine (+4.29 and + 6.70 degrees, respectively), which was statistically significant for both (P = 0.001).  The molar angle was reduced by -1.53 degrees, which was not statistically significant (P = 0.126).  However, from T1-T2, the SE group demonstrated the opposite trend; average decreases of 0.07 degrees in incisor angulation (P= 0.953) and a statistically significant decrease of 4.38 degrees in canine angulations (P = 0.002), with a statistically significant increase in average molar angulation (+3.07 degrees, P = 0.002).  The LPE group (Figure 3.15) exhibited tip increases for every outcome measure from T1-T2.  The average incisor angulation increased by +6.19 degrees (the most  64 significant outcome difference, P = 0.001), the average canine angulation increased by +2.22 degrees (P = 0.047), and the average molar angulation increase was recorded at 2.18 degrees (P = 0.008).     Figure 3.13.  Cephalometric tip in the control group from T0-T1  Figure 3.14.  Cephalometric tip in the SE group from T0-T1-T2  Figure 3.15.  Cephalometric tip in the LPE group from T1-T2  In comparing the means of these angulations between the study groups, at T0 (Table 3.8 and Figure 3.16), the controls and SE groups had non-statistically significant differences in the mean incisor and canine tip, the significance P=0.220 and 0.175  65 respectively.  On the other hand, the molar inclination was significantly different at baseline, P=0.015.     Table 3.8. Cephalometric tip measurements at TO (baseline)  - comparisons of mean angles between SE and control groups  Outcome: CEPHALOMETRIC TIP AT T0 Controls Serial Extractions P values (95% CI) N Mean ± SD N Mean ± SD Incisor T0 30 63.58±2.72 30 62.24±5.26 0.220 (-0.84; 3.52) Canine T0 30 69.20±3.34 30 67.32±6.68 0.175 (-0.87; 4.63) Molar T0 30 74.42±3.22 30 71.58±5.27 0.015 (-0.57; 5.11)   Figure 3.16. Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in controls and SE at T0  The bivariate analysis at T1 presented in Table 3.9 considered all three groups.  The SE and control groups had no statistically significant difference for the incisor angulation at this stage (P=0.063), while significant differences (P=0.001) were observed between the controls and LPE, as well as between the SE and LPE groups (Figure 17).   The same pattern was observed for the canine average angulation, P=0.280, indicating no significant difference among the control and the SE groups, while  66 significant difference was present for the LPE as compared to the other two groups (P=0.001).  The mean molar angulation at T1 revealed a different pattern; the SE and LPE groups exhibited no significant difference, while a significant difference was observed between the SE and LPE groups and the control group (P=0.001).   67 Table 3.9. Multiple comparisons of cephalometric tip measurements at T1 between control, SE, and LPE groups  Outcome: CEPHALOMETRIC TIP AT T1 (degrees)  ANOVA with Bonferroni Post Hoc Adjustment  P values (95% CI)  Group Comparisons Controls versus SE Controls versus LPE  SE versus  LPE IncisorT1  (Mean±SD) 0.063 (-6.26; 0.12) 0.001 (3.48; 9.86) 0.001 (6.55; 12.93) Controls 63.46 ± 4.17 N/A N/A N/A SE 66.53±5.38 N/A N/A N/A LPE 56.79±5.51 N/A N/A N/A Canine T1  (Mean±SD) 0.280 (-5.55; 0.99) 0.001 (4.24; 10.79) 0.001 (6.52; 13.07) Controls 71.75±3.40 N/A N/A N/A SE 74.02±5.70 N/A N/A N/A LPE 64.23±6.08 N/A N/A N/A Molar T1  (Mean±SD) 0.001 (2.03; 7.44) 0.001 (2.93; 8.33) 1.00 (-1.81; 3.60) Controls 74.79±3.80 N/A N/A N/A SE 70.05±5.21 N/A N/A N/A LPE 69.16±3.68 N/A N/A N/A   68  Figure 3.17. Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in controls, SE, and LPE at T1  At T2, all SE group’s tooth angulations were larger versus the LPE group (Table 3.10, Figure 3.18).  Statistically significant differences were observed for the averages of the incisor and canine angulations (P=0.026 and 0.035 respectively) while the average molar angulation was non-significant between the two groups (P=0.085).    Table 3.10. Cephalometric tipping measurements at T2 (post orthodontic treatment)  - comparisons between SE and LPE groups  Outcome: CEPHALOMETRIC TIPPING T2 Serial Extractions Late Premolar Extractions   P values (95% CI) N Mean ± SD (angles) N Mean ± SD (angles) Incisor T2 30 66.46±6.31 30 62.92±5.65 0.026 (0.45; 6.64) Canine T2 30 69.64±7.03 30 66.55±3.29 0.035 (0.22; 5.94) Molar T2 30 73.12±4.71 30 71.34±2.95 0.085 (-0.25; 3.82)   69  Figure 3.18. Comparison of angulations of the lower incisor, canine and first molar relative to ANS-PNS in SE and LPE at T2  The Multiple Linear Regression (MLR) models were used to assess the joint effect of multiple predictors associated with the differences in the study outcomes.  Regarding the variation across the incisor angulation at T1 (Table 3.11), the MLR model was statistically significant, P=0.001, with 31.4% of the variation in the incisor angulation at T1 explained jointly by multiple predictors such as the incisor angulation at T0 (=0.409, P=0.001), followed by the group dependence effect (=0.383, P=0.001), and a smaller contributory gender effect (=0.270, P=0.016).  The MLR model for variation in canine angulations was also statistically significant (P=0.008), and the variation in this outcome measure was explained by the effects of the same multiple predictors as in the MLR model for the incisor angulation at the T1.  Similarly as in the previous MLR model, the canine angulation at T0 was the strongest predictor of canine position at T1 (=0.366, P=0.004), followed by group dependence (=0.303, P=0.016) while gender effects were negligible and not statistically significant (=-0.060, P=0.622).  The MLR model for the molar angulation yielded a statistically significant overall model (P=0.001), with 41.5% of the variance explained by multiple predictors.   Consistent with the angulation of the other two teeth, the molar position at T0 was the  70 strongest predictor (=0.504, P=0.001), followed by the group dependence (=-0.310, P=0.005) and small non-significant gender effects (=-0.022, P=0.832).   71 Table 3.11. Predictors of variance in tooth tip at T1– Multiple Linear Regression (MLR) models  Predictors Standardized Coefficients Significance Incisor tip at T1  Group 0.383                   0.001 Gender 0.270 0.001 Incisor at T0 0.409 0.016 Model summary: Adjusted R square = 0.314; P= 0.001 Canine tip at T1  Group 0.303 0.016 Gender -0.060 0.622 Canine at T0 0.366 0.004 Model summary: Adjusted R square =0.147; P= 0.008 Molar tip at T1  Group -0.310 0.005  Gender -0.022 0.832  Molar T0 0.504 0.001  Model summary: Adjusted R square =0.415; P= 0.001  72 The overall MLR models for T2 (Table 3.12) were all statistically significant.  Specifically for the incisor angulation at T2, 12.3% of the variance was explained by the angulation measured at T1 (=0.401, P=0.020), while the group dependence and gender effect were non-significant (=-0.018, P=0.913 and =-0.072, P=0.565 respectively).  Considering the differences observed, 12.6% of the variance in canine angulation at T2 was explained jointly by multiple predictors.   The strongest contributor (predictor), similar to the case of the incisor angulation, was the canine position at T1 (=0.400, P=0.015), while the other two predictors were non-significant gender (=0.034, P=0.778) and group dependence (=-0.016, P=0.920).     Finally, the MLR for the molar angulation at T2 revealed that 19% of the variance was explained by the molar position at T1 (=0.413, P=0.001) and both gender and group dependence effects were non-significant.  73 Table 3.12. Predictors of variance in tooth tip at T2 –MLR models  Controls versus Serial Extractions Predictors Standardized Coefficients Significance Incisor tip at T2  Group -0.018 0.913  Gender -0.072 0.565  Incisor tip at T1 0.401 0.020 Model summary: Adjusted R square = 0.123; P= 0.016 Canine tip at T2  Group -0.016 0.920  Gender 0.034 0.778  Canine tip at t1  0.400 0.015 Model summary: Adjusted R square =0.126; P= 0.015 Molar tip at T2  Group -0.184 0.125  Gender -0.155 0.192  Molar tip at T1 0.413 0.001 Model summary: Adjusted R square =0.190; P= 0.002        74 Chapter  4: Discussion This study sought to address the clinical question why do serial extractions (SE), as an interceptive treatment option, not seem to save more than four-six months of active treatment time. Our aim was to shed light on quantifiable changes that result from SE, specifically with regards to the changes in occlusal curvature depth and the direction of tooth tipping.  The results of the cast data revealed that at T1, the SE group as a whole had the smallest overall radius (mimicking Monson’s sphere) and smallest radius of lower first molars (mimicking the curve of Wilson), meaning that teeth were most likely tipped lingually, creating steeper occlusal curves.  The radius of the lower incisors and buccal cusps of the molars (mimicking the curve of Spee) was not significant among the groups.  The cephalometric data, although limited by virtue of being two- dimensional, confirms that the incisors and canines post drift (T1) had larger angulations (more obtuse, i.e. the incisor and canine tipped back secondary to premolar extraction) relative to palatal plane.  This may indicate that these teeth have a tendency to upright as time elapses from T0-T1.   We chose to use the spherical radii as outcomes to determine different patterns that exist in the development of occlusal curvatures.  The logic was that this novel method, combined with conventional cephalometric radiographs could help identify and isolate the different teeth and their tipping pathways that result in either increased or decreased occlusal curvature depth.  Nam et al used a similar methodology of measuring the occlusal curves in 3D using spheres, however his landmarks differed as did his methods (Nam et al, 2013).  Nam was able to reduce measurement error by using a custom-made software program that provided a mechanism to specify the  75 direction along which to measure the height of each vertex from a region of interest and then automatically select the highest vertex and identify it as the cusp tip (Nam et al, 2013).  However, inherent error exists in the examiner’s delineation of the axes (Nam et al, 2013).  In the present study, inter-examiner measurement error for the spherical outcomes was highly reliable with a reproducibility correlation coefficient of 0.711-0.834.  The advantage of the three spheres we measured was that they were chosen to mimic the occlusal curves individually (Spee and Wilson), as well as combined (Monson’s sphere).  This method enabled us to isolate the changes associated with SE treatment to the specific curves.  Our results show that one measure, the spheres fitted to the buccal cusps and incisors only (attempting specifically to mimic the curve of Spee), proved consistently statistically insignificant between the groups. This leads us to question the interpretation of this specific outcome because a sphere of best-fit to a 3D curve represented by the buccal cusps and incisors would not only be affected by anteroposterior arch curvature, but also by the arch curvature in the horizontal plane. This effect of fitting a sphere as best as possible could lead to an overestimation of how curved the arch actually is (without any internal constraints, the sphere could sink down to match the curve delineated by the selected landmarks on the cusps).   A check on actual landmark variance was done by manually checking the distance from each of the 12 landmarks (comprising the measure for Monson’s sphere) to the centre of the fitted sphere on a handful of casts.  The standard deviations of the 12 landmark distances to the fitted sphere centre were recorded and considered as reasonable indicators of landmark variance from the fitted radius.  The standard deviations observed were all within 1 mm.   76 Only mandibular casts were measured, as the mandibular arch is the final diagnostic guide, with particular emphasis on the harmonious relation of the mandibular incisors to the alveolar bone (Dewel, 1954).  Serial extraction treatment addresses this objective by reconciling the difference between an amount of tooth material and a persistent deficiency in basal bone (Dewel, 1954).   It is still important to consider that the mandibular arch does not function alone, and that it does function by occluding with the maxillary arch. Occlusion is not a static condition and the intra-arch relationships of individual teeth, the inter-arch relationships of the teeth, and the positional relationships of the teeth to the maxilla and the mandible change continually throughout life.  Therefore, the only constant in occlusion is continual change (Begg, 1954).  Our study only addressed the change in the mandibular arch, which Graber identified as the culprit arch, where a greater likelihood exists for tipping by means of convergence of the canine and the second premolar crowns (Graber, 1971).  Conversely, he observed that the maxillary canines and premolars more often than not paralleled themselves, with autonomous adjustments (Graber, 1971).  Future studies utilizing this method should include the maxillary arch to confirm Graber’s findings and to see if any of the same patterns exist. The third spherical outcome, the radius encompassing only the buccal cusps of the first molars and the four incisors (curve of Spee) was not statistically significant at T1 among the three groups.  A possible explanation for this could be that the buccal cusps of the lower first molars stay in the same position, and all the movement comes from lingual tip.  The lingual cusps of the lower first molars are included in the other two sphere outcomes (radius overall and radius of the four cusps of the lower first  77 molars), which demonstrated significant differences.  This suggests more change to the curve of Wilson, which also contributes to the changes seen in the sphere of Monson (overall radii).  Based on the MLR models at T1, it was made clear that the grouping effect and values at T0 were major predictors in the differences observed.  The MLR at T2 revealed that the observed values at T1 contributed as major predictors to the differences as did gender. These results confirm the expectations that extracting teeth and allowing natural drift altered the occlusal curves and tipping patterns.  These extraction-induced changes require time to correct by uprighting and achieving proper tooth tip during comprehensive orthodontic treatment.  This is especially true for the mandibular canine, as its larger conical root is challenging and time consuming to move in order to reach the appropriate desired final position.   Comparing the differences in the occlusal curves from the casts between the two groups at T0, differences observed between the SE and control groups were not anticipated, nor favourable for our purposes. The differences observed at T0 between the SE and control groups were mostly likely a result of the extraction of the primary canines in the SE group.  This was done prior to taking the initial records for the SE group, and therefore, prior to extractions of the primary first molars and permanent first premolars.  These extractions would normally result in altered tooth movement, often in a distal direction for the incisors, which likely led to the significant differences observed in 2/3 of the radii outcomes (the overall radii and the radii of the lower first molars).    78 At T1, we expected to see statistically significant differences between the SE subjects and the other two groups.  Controls and LPE have had no intervention, while the SE cases have previously had their extractions (primary first molars and permanent first premolars) and have now finished their post extraction drift period.   The controls and LPE groups showed no significant differences in the three radii outcomes measured.  However, the SE group demonstrated a significant difference from the two other groups (LPE group has not yet undergone extractions) with regards to two radii: the overall radii and the radii of the lower first molars.  There were no significant differences among the three groups in the curve of Spee outcome.  The shortened arch length in the SE cases, would yield smaller radii and explains these observations. Post orthodontic treatment, at T2, the LPE group had significantly larger radii versus the SE group with respect to overall radii and the radii of the lower first molars, but no difference was observed in the radii of buccal cusps of lower first molars and four incisors.  The LPE group’s arch form could be a result of a lesser amount of arch and tooth collapse (that could result from extractions), which is maintained as the child grows and transitions from the mixed to the permanent dentition.  The alveolar bone housing the teeth naturally resorbs in height and width when a tooth is extracted.  In SE cases, this reduced arch length is one of the accepted consequences with the greater benefit of reducing the crowding and risk of impaction in such TSALD cases.   Although significant differences were observed between the SE and LPE groups in the overall radii and the radii of lower molars at T2, this does not indicate that one group was successfully treated and the other was not.  The reduced alveolar development contributing to a shorter arch length in the SE cases may be responsible for these  79 observations.  The clinical implications of smaller versus larger radii probably do not directly correlate with treatment difficulty.  So long as the TSALD is managed appropriately, with a technique that is biologically sound, neither SE, nor LPE, should be considered to be compromised treatment. It became evident from the radii results that the most significant differences among the individual groups, as well as between them, was a result of the lower first molars.  Therefore, we devised a method to measure the molar tip in the lingual direction (cephalometric radiographs only show changes in the anteroposterior direction).   By measuring the linear distances from the ML and DL cusp tips to a reference plane formed by the MB and DB cusp tips of both right and left first molars, we were able to isolate the magnitude and direction in which the first molars rolled to the lingual.  This increase in lingual inclination contributed to the increased curve of Wilson.  The control group showed decreases in ML and DL distances, which indicates molar uprighting from T0-T1 in the control group.  This may be apparent uprighting as it is possible that the lower molars erupt to the lingual and then they level out as they reach the occlusal plane.  The comparison for molar tip at T0 between the control and SE group demonstrated a significantly more lingually positioned molar in the SE group, for both ML and DL cusps.  At T1, the SE group had the largest outcomes for ML and DL cusps from the reference plane, followed by the LPE, with the controls having the most leveled ML and DL cusps to the reference plane.  The SE group had the most lingually rolled molars as compared to the control and LPE groups at T1, which was illustrated by both of the molar’s lingual cusps having greater distances from the reference plane.  This confirms that SE does indeed deepen the curve of Wilson.  The control and LPE  80 groups at T1 were not statistically different for both the ML and DL outcome values.  This is most likely a result of the LPE group not yet having the premolars extracted, nor the opportunity for the molars to roll lingually, as these cases are almost immediately placed in full fixed appliances and leveled.  In the SE group, from T0-T1, the ML and DL cusps had non-significant increases in linear distance from the reference plane.  From T1-T2, the SE group had significant decreases in the ML and DL distances, indicating that the orthodontic treatment served to upright the lingually tipped molar.  The LPE group showed significant decreases in both ML and DL distances to the reference plane from T1-T2, which similarly to the SE group, indicate that orthodontic treatment led to the molar being uprighted in a buccolingual direction. Finally, at T2, The SE and LPE groups had non-significant differences in both ML and DL cusp distances from the reference plane.  This result helped confirm that the appropriate orthodontic treatment in these cases uprighted and leveled the molars, which parallels ABO CRE criteria.   Comparing the differences in incisor, canine, and molar movements on cephalometric radiographs between the SE and control groups at T0, they showed no significant differences in incisor and canine angulations.  However, at this time, a statistically significant difference was observed in molar angulation.  One must remember that the SE cases at T0 have already had an intervention: their primary canines were extracted.  With this past intervention in mind, it could explain the significant differences observed in the outcome of molar movements.  In addition, given inter-individual variation, this is a realistic and plausible outcome, and was accounted for in the regression analysis.  81 Evaluation of tooth angulation relative to the palatal plane at T1 among the control, SE, and LPE groups illustrated a pattern for the incisors and canines.  Both incisor and canine angulations were not statistically significant among the controls and SE at T1, but were statistically significant between those two groups and the LPE group.  The SE sample had a more upright lower incisor while the LPE sample had an overall more acute incisor and canine angulation.  This leads to the potential conclusion that due to the increased anterior crowding in these subjects, the incisors and canines had increased mesial crown tip (tip labially) as observed on the cephalometric radiographs.  This supports the notion that without early intervention in severely crowded cases, lower canines will tend to erupt buccally and further push the incisors labially.  However, at T1, the molar angulation did not follow this pattern.  The SE and LPE groups had no significant differences in molar angulation, while they both differed significantly from the controls, whose molars tended to have obtuse angulations relative to the palatal plane, indicative of more distal molar crown tip.  This may be attributed to the nature of the population making up the control sample as well as the quality and density of the buccal bone in the molar area.  The almost identical molar angulation value calculated for the SE and LPE groups could possibly be a result of the natural eruption path of the molars, and the anterior crowding having a minimal effect on the posterior teeth’s ability to tip.  While the SE group is now post intervention and natural drift, the lack of significant molar change is most likely due to the space from the extraction being used up by the anterior teeth unraveling and allowing space for the permanent canines to erupt by tipping distally into the mandibular first premolar extraction site.   Patterns of facial divergence were not considered in this study, and this  82 may also play a role in molar movement, as hyperdivergent growth could increase the likelihood of mesially tipped molar crowns.  For future studies, facial patterns of growth should be considered as a criteria for comparison to evaluate if there is any difference. Regarding the treated cases, it is important to remember that the same two orthodontists treated all of the cases in the experimental sample.  The tooth angulations post comprehensive orthodontic treatment (T2) between the SE and LPE groups revealed a significant difference in incisor and canine angulations, but no statistically significant difference, once again, in molar angulations, which is consistent with the previous time point (T1) calculations. This outcome was expected as the incisors and canines in the LPE group showed more acute angulations, indicating greater mesial tip on the incisors and canines at T1.  The majority of the space created from extracting the lower first premolars in the SE and LPE groups most likely ends up being occupied by the crowded permanent canines.  Any excess space in the anterior region once the canines are in the arch, is often used to retrocline and retract flared mandibular incisors.  The amount of incisor uprighting and the amount of leveling is limited by the arch form, desired overjet, and most importantly by the resulting lip support and profile effects.    The suggested advantages of SE versus LPE include improved relationships of the dentition to the surrounding hard and soft tissues, and a reduction in active treatment time (Graber, 1971; Wagner and Berg, 2000; O’Shaughnessy, 2011).   The period of drift following SE is meant for self-alignment, which can reduce the first stage of comprehensive treatment: leveling and aligning  (Graber, 1971).   Graber suggested that SE could reduce active therapy by 50%, while O’Shaughnessy observed a reduction  83 of 4-6 months in his sample (Graber, 1971; O’Shaughnessy, 2011).  O’Shaughnessy’s study corroborates Little’s findings that SE may require more observation visits but fewer months of active therapy with fixed appliances (Little et al, 1987; O’Shaughnessy, 2011).   Wagner and Berg’s data support the idea of post-SE physiological drift minimizing future treatment duration and complexity (Wagner and Berg, 2000).  However, they also mentioned a similar end result could be obtained later, with LPE and a longer duration in fixed appliances (Wagner and Berg, 2000).  A greater active treatment reduction time would have been anticipated, but it is only 4-6 month (as observed by O’Shaughnessy), most likely a result of the amount of time required to correct the root divergences created by tooth tipping in the mandibular arch (Graber, 1971; O’Shaughnessy, 2011).  The crown convergence/root divergence of the mandibular canines and second premolars after extraction of the first premolar are to be corrected, as root parallelism is an objective of orthodontic treatment (Graber, 1971).    Our findings quantify these tipping changes and compare them to the LPE group to determine any significance.  Perhaps a removable appliance with guide planes during the guidance of eruption and physiological drift period would benefit the SE patient.   While, improved anterior alignment probably contributes to the 4-6 month decrease in active treatment duration, if an appliance could reduce the amount of tipping permitted, it is possible that greater than 4-6 months could be saved. In the present study, an algorithm of best fit for producing the 3-D spheres to represent different occlusal curves was developed and the resulting radii measurements of the spheres were performed.  This was then related to a 2-D cephalometric analysis to help determine the directional tipping behaviour of specific  84 teeth in the lower arch.  There have been two published papers that address novel ways of analyzing the occlusal surface in 3-D (Ferrario et al., 2003; Nam et al., 2013). Ferrario et al. aimed to describe human occlusal curvature in 3-D from a mathematical and statistical point of view (Ferrario, Sforza, Poggio, Serrao, & Colombo, 1999), while Nam’s purpose was to generate the 3-D occlusal curvatures of Monson’s sphere and curve of Wilson using virtual dental casts and custom designed software to quantify them (Nam et al., 2013).  Ferrario was the first investigator to obtain x, y, z coordinates of cusp tips using a 3-D digitizer and the first to demonstrate a geometric-mathematical analyses of a spherical model of the occlusal curvature (Ferrario et al., 2003).  The cephalometric analysis on tooth tip post extractions is a more prolific topic and has been studied over the last 50 years (Glauser, 1973; Weber, 1969; Yoshihara et al., 2000).  The reasons why palatal plane was selected as the reference plane for the cephalometric analysis are twofold: it can be located easily with great precision and therefore is easily reproducible, as well as its location lies in close proximity to the areas under consideration.  Broadbent observed that palatal plane appeared to maintain a parallel relation over the growth range in the population he studied (Broadbent, 1937).  Furthermore, Brodie also found that palatal plane maintains a constant angular relationship with anterior cranial base (SN) (Brodie, 1953).  In addition, Yoshihara used palatal plane as his reference plane to assess SE cases (Yoshihara et al., 2000). From the literature, the advantages of using palatal plane as a reference plane include: 1 No influence of growth changes at Nasion 2. No influence of rotation of the jaws   85 3. Inclination of the occlusal plane by dental effects is excluded  (Premkumar, 2011) In the present study using the cephalometric data obtained, and looking across the time points for the three samples, certain patterns were observed.  The controls had a general non-significant decrease (-0.12 degrees) in the angulation of the incisors (very slightly tipped mesial), and an increase in canine (+2.55 degrees) and molar angulation (+0.37 degrees), indicative of distal tip, from T0-T1.  This can be interpreted as a minimal change in the incisor and a more significant distal tip or uprighting of the canines (to achieve root parallelism) with a slight uprighting of the molar.     According to our data for the SE group, the mean lower first molar angulation at T0 was 71.58±5.27, and tipped slightly mesially post drift at T1 to 70.05±5.21.  When comparing the control and SE groups from T0-T1, our results for incisor angulation were in accordance with Kau et al., whose randomized control trial focused on primary canine extractions (Kau et al., 2004).  Kau also found no significant differences in the incisor inclinations between the extraction and non-extraction groups after the two-year observation period (Kau et al., 2004).  Their study observed a decrease in arch perimeter in the extraction group, which he attributed to the mandibular molar tipping forward or mesially while the incisor angulation was maintained (Kau et al., 2004).  At T2, there was no significant difference observed in the molar mesial migration in the SE and LPE groups, which is in agreement with Ringenberg who observed that the mesial migration of molars during the period of drift in the SE group proved of no consequence in the final results of treatment.  These teeth were moved forward a comparable magnitude during active treatment in the LPE group due to anchorage loss (Ringenberg,  86 1967).  While we measured relative to the palatal plane, this result conflicts with Weber, who reported a trend observed in SE cases with lower first premolar enucleation, whereby the lower first molars had a definite tendency to upright (more distal inclination) in relation to the line from nasion to B point (Weber, 1969).  On average, Weber reported 2.9 degrees of molar uprighting over the two and a half-year observation period.  He also noted that the space closure was divided into individual tooth movements; mesial molar movement contributed to one third of the closure and the distal canine movement was responsible for the remaining two thirds (Weber, 1969).  Weber’s findings overall indicate mesial molar translation with concomitant uprighting (Weber, 1969).  Yoshihara, whose reference plane was used in our study, found that the first molar tipped mesially (0.29 degrees/year) after the deciduous canine was extracted, and then proceeded to tip distally 0.33 degrees/year from the time of the first premolar extraction until the end of an undefined observation period (Yoshihara et al., 2000).  Our control sample also exhibited minimal change relative to the first molars.  At T0 the mean angulation was 74.42±3.22, and at T1 was 74.79±3.80.   The aforementioned studies (Kau et al., 2004; Weber, 1969; Yoshihara et al., 2000) are limited by small sample sizes, and these results can be more attributed to individual variation, ethnicity, and skeletal patterns.  With steeper mandibular plane angles, the molars would have a greater tendency for mesial tipping versus those subjects with flatter mandibular plane angles.   The canine movement may be another possibility for these conflicting results.  In our study, the canine change in the SE group was significant from T0-T1.  Specifically, it tipped back into the extraction site, increasing its distal inclination from 67.32±6.68 to 74.02±5.70.    87  There are several limitations to our study.  The retrospective study design is inferior to clinical randomized blinded trials and obviously limits causal inferences.  Ethics considerations and a lack of good informed consent would make this design difficult.  Secondly, there is some measurement error, which was 0.711 to 0.834 for the cast data, and 0.808 to 0.956 for the cephalometric data.  However, the significant differences observed at T0 may be explained by inherent differences within the matched SE and control samples.  The sample size for having three groups, 30 subjects in each was calculated a priori.  A sample of this size has also the potential to reflect and adjust for the differences in inter-individual variation in each of the groups.  Therefore, the inter-group variation, if it exists, could make it possible to find differences among the groups.    In selecting our control sample, we chose to use the Oregon longitudinal control sample because it was easily available (nearest in proximity and immediate permission of use was granted) and included the necessary records required for our research purposes.   Relative to other existing longitudinal growth studies (Burlington, Bolton-Brush, Iowa), the Oregon sample was smaller and more limited in Class I samples (30).  Our criteria included 30 Class I controls, and therefore we were able to go with Oregon’s sample.  A larger sample size would have been beneficial to eliminate the effects of inter-individual variation within the groups, thereby potentially enabling the outcomes measured to show more differences.  In selecting the individual cases for study, we did not consider any skeletal relationships, and attempted to minimize possible confounders by choosing to include only Class I dental malocclusion cases.  We chose not to consider the effects of the directionality of growth patterns, and rather  88 focused solely on changes to the dentition. However, sagittal jaw relationships are difficult to evaluate because of rotations of the jaws during growth, vertical relationships between the jaws and the reference planes, and a lack of validity for the many proposed methods of their evaluation (Nanda, 1971; Nanda & Merrill, 1994).  All of the above could affect the outcomes we studied, as could the slope of the articular eminence, craniofacial morphology, and the amount of overbite and overjet, which have been related to the morphological arrangement of the teeth in the sagittal plane (Farella, 2002; Ahmed et al, 2011).  Shannon and Nanda found that the only skeletal measurement that correlated to the depth of the curve of Spee was the Frankfurt horizontal to mandibular plane angle (FMA), which decreased as the curve of Spee depth increased (Shannon & Nanda, 2004).  The amounts of overbite and overjet have also been shown to significantly influence the variation of the curve of Spee in the mandibular arch.  Specifically, the depth of the curve of Spee increases with an increase in overbite and overjet (Ahmed et al, 2011; Cheon, 2008).  With regards to craniofacial growth, Farella et al found that the curve of Spee is more dramatic in short-face individuals and less dramatic in long-face individuals (Farella, 2002).  Similarly, Ahmed et al found that the curve of Spee was deepest in Class II division 2 malocclusions that presented with deep overbites and short face heights versus (Ahmed et al, 2011).  Another important note is that in order to avoid potential measurement bias we did not eliminate cases with intraoral restorations, which would include occlusal surfaces and some cusp tips and therefore possibly impact our digitized landmarks on the casts (12 cusps).    89 With regards to the methodology, we chose to not consider the second molars for our outcomes.  This was decided because at the start of treatment, namely T0 and T1, the second molars were unerupted in the control and SE samples.  It is important to note that the second molar eruption generally contributes to the deepening of the curve of Spee (Marshall, 2008).   Contrary to the literature, our curve of Spee relied on the first molar.  The duration of the drift period following SE was thereby dictated by  the eruption of the second molars, which was coincident with fixed appliance treatment.  Our method of measuring the curve of Spee attempted to adhere to Baydas’ method (Baydas et al., 2004).  This commonly used method averages the right and left perpendicular distances between the deepest cusp tip and a flat plane that was laid on top of the mandibular dental cast, touching the incisal edges of the central incisors and the distal cusp tips of the most posterior teeth in the lower arch (Baydas et al., 2004).  The most posterior teeth in the arch of our sample at T0 and T1 were the first molars.  The difference in method was that this study used spherical radii to reflect the depth encompassed by the landmarks selected, and not just one point reflected by the deepest perpendicular distance, as per Baydas.  The landmarks that were selected are unique to the literature and only resemble Nam’s 3-dimensional study (Nam et al, 2013).  They were designed specifically to isolate the different curves, as we aimed to identify and measure the individual curvatures’ contributions to changes observed in SE cases.  A mathematical limitation in the methods used may exist due to measuring a curve.  The technique of sphere-fitting is a 3D surface-based technique and cannot differentiate various spatial components within a cloud of points.  This study emphasized intra-arch measurements and therefore, it must be considered that the only possible reference  90 was itself and secondly, it was not a defined property, but rather a descriptor.  This study considered the mandibular arch independently, recognizing that it does interact with the maxillary arch in order to allow for masticatory function.  In addition, the cephalometric radiographs that were utilized provide a two-dimensional measure of a three-dimensional object.  While it is common practice for orthodontists to use the traditional two-dimensional descriptors, this type of analysis is restricted to the anteroposterior and superoinferior dimensions alone.  This limitation proves it a questionable procedure for a three-dimensional skull.  With that in mind, the curve of Spee remains a loose 2D conceptualization, and as such will likely continue to pose problems in any kind of meaningful quantification.  Machine error could result in different degrees of magnification and distortion by means of patient size, machine settings, and patient placement in the cephalostat.  The Oregon longitudinal control sample was not all radiographed by the same machine, as technological advances in the field yielded better models over the time course of the data collection.  The cephalostats were not uniform among the experimental groups from private practice either (more recent patients were radiographed using a new digital cephalostat).  The use of absolute angular measurements helped eliminate bias.  They should not be dependent on those factors related to the cephalostat.  A major limitation with respect to the method was in the SE and control samples at T0.  The crowding of these two samples was not assessed.  It was assumed that the SE cases were severely crowded and that the controls represented the normal amount of crowding seen in the general population.   Unfortunately, at T0, which marked an early stage of development, the permanent canine was often unerupted in the control and SE groups.  The primary  91 investigator had to estimate its cusp tip and root tip via the tooth bud position.  Placement of the landmarks for the long axis of the canine in these cases was done by approximation based on the forming cusp tip and the dental papilla.   The cephalometric intra-class correlation coefficient was satisfactory, ranging from 0.808 to 0.956. The clinical implications of this study provide a better understanding of how SE affects the developing arches and dentition.  These effects are best explained by quantifying the consequences that ensue following SE by two means.  One is by defining the radii formed by specific landmarks (teeth) to visualize arch changes.  The second measure was to the tip of the specific teeth in order to determine the direction of crown movement.  Generally speaking, spontaneous changes that occur during the driftodontics period following SE, have been shown to be therapeutic in terms of relieving crowding.   Considering individual variability, there is also a change in the way this relief is manifested, as the permanent teeth erupt into the newly available space.  The eruption trend observed in the SE cases in this study was accompanied by tooth tipping, which can be due to excess space created from the extractions or from ectopic eruption patterns (which may in turn be a result of the initial crowding).  Comprehensive orthodontic treatment, which may not always be a requirement once SE is done, is often the choice to achieve a more ideal alignment, close residual extraction spaces and improve tooth position.   The orthodontic treatment in this study illustrated that reversal of the movements during drift in the SE cases were accomplished; incisors and canines were positioned more mesially, and the molars more distally.   This is seen clinically as the lower incisor and canine tipping back into the premolar extraction  92 space during drift, followed by being tipped more forward during fixed appliances treatment. The treatment efficiency argument of SE may be clarified by the changes observed in the occlusal curves, mainly secondary to tooth tipping.  This was especially evident in this study with both the ML and DL cusps of the first molars in the SE group revealing significant tip.  These teeth require time and diligence to be uprighted with the proper root torque obtained in fixed appliances.  Typically, a deep curve of Spee is corrected by any combination of extrusion of the premolars, uprighting of the molars, and flaring or intrusion of the incisors (Shannon & Nanda, 2004).  This study did not measure premolar extrusion in the treated samples at T2, if any, which would introduce a vertical component.   It is very unlikely that the limiting factor in orthodontic treatment would be extrusive movements.  Continuing in the sagittal plane, mandibular incisor proclination was not considered, which according to Pandis et al’s prospective clinical trial, accounted for the majority of curve of Spee leveling correction (Pandis, 2010).  Specifically, for every 1 mm of leveling, the mandibular incisors were proclined 4 degrees, without altering arch width (Pandis, 2010).  Germane et al utilized mathematical models to evaluate the effect of leveling the curve of Spee on arch circumference (Germane et al., 1992). They found that the amount of arch length required to level the curve of Spee was non- linear, and for their sample, consistently less than one to one for curves of Spee less than 9 mm (Germane et al., 1992).   They concluded that common orthodontic practice of allocating 1 mm of arch circumference for every millimeter of leveling the curve of Spee was a gross overestimation (Germane et al., 1992).    93 For SE cases, if the 4-6 month braces-free reduction can be justified as a health, economic, or psychosocial benefit, it could be argued that the early extractions before comprehensive treatment was advantageous in the form of less time (months, minutes, number of visits) with the orthodontist while in fixed appliances (O’Shaughnessy et al, 2011).   O’Shaughnessy’s study compared SE and LPE cases using the PAR index, and SE cases were lower.  This indicates that the SE cases in his particular study were easier for the orthodontist to finish versus the LPE group.  However, treatment time was not a consideration in this study.  This could have been valuable in verifying the tooth movements required versus fixed appliance treatment duration.  The SE and LPE sample selected for this study and O’Shaughnessy’s sample were from the same orthodontic practice, but the samples were not coincident, and therefore his results cannot be directly applied. The results from this study reject both of the study’s null hypotheses.  Significant differences were observed, and this information suggests a general guideline to anticipated SE-related occlusal curvature changes and tooth movement.  Other clinical factors that can contribute to the success of SE include changes in alignment, marginal ridge discrepancies, overbite, overjet, and anteroposterior relationships.  In order to validate present findings, future directions should include a prospective study incorporating a larger, more uniform sample for comparison purposes, ideally having initial records for the SE cases prior to extractions of the primary canines, as well as measure the maxillary arch and skeletal parameters (FMA) to get a more complete picture of changes in the occlusal curvatures.  It is possible that while SE is a tool for early  94 interceptive treatment, it does not negate the need for fixed appliances to correct the occlusion to established orthodontic standards.  95 Chapter  5: Conclusion Interceptive treatment with SE could be undertaken in carefully selected individual cases that meet specific criteria: early mixed dentition, no congentially missing teeth, severe crowding/TSALD, Class I skeletal relationship, and good facial proportions.  In this study’s sample, SE patients had steeper occlusal curves (smaller radii) at T1 and T2. The SE group’s changes were reflected by smaller Monson’s spheres and curves of Wilson after the period of tooth drift compared to the control and LPE groups, and after orthodontic treatment as compared to the LPE group.  No significant differences were observed in the curve of Spee among the three groups.  The molar tip of the SE group in the buccolingual direction demonstrated an increase in lingual tip from T0-T1, followed by a decrease from T1-T2 with the leveling from the orthodontic appliances.  The SE group had significantly greater molar lingual tip at T0 and T1 as measured in linear distances from the ML and DL cusp tips to the reference plane formed by the buccal cusps of molars.  The SE group’s incisors and canines had a tendency to tip distally and upright from T0-T1, while the molars had a tendency to tip mesially.  Whilst in orthodontic treatment, from T1-T2, the incisors and canines tipped mesially while the molars tipped distally, reversing the effects of physiological drift secondary to SE.  In general, orthodontic treatment, post SE, involved incisor and canine proclination, as well as molar uprighting.   96  Bibliography   Adaskevicius, R., & Svalkauskiene, V. (2011). Measurement of the Depth of Spee's Curve using Digital 3D Dental Models. Electronics and Electrical Engineering, 109(3): 53–56. Ahmed, I., Nazir, R., Erum, G.-E., & Ahsan, T. (2011). Influence of malocclusion on the depth of curve of Spee. The Journal of Pakistan Medical Association, 61(11): 1056–1059. Alexander, R. G. (2001). The Principles of the Alexander Discipline. Seminars in Orthodontics, 7(2): 62–66. The American Board of Orthodontics [online].  http://www.americanboardortho.com/professionals/downloads/Case_Management_Form_Instructions.pdf.  Andrews, L. F. (1972). The Six Keys to Normal Occlusion. American Journal of Orthodontics and Dentofacial Orthopedics, 62(3): 296–309. Arya, B. S., Savara, B. S., & Thomas, D. R. (1973). Prediction of first molar occlusion. American Journal of Orthodontics, 63(6): 610–621. Baume L.J. (1950). Physiological tooth migration and its significance for the development of occlusion. II. The biogenesis of accessional dentition. Journal of Dental Research 29(3):331-337. Baydas, B., Yavuz, I., Atasarai, N., Ceylan, I., & Metin Dagsuyu, I. (2004). Investigation of the Changes in the Positions of Upper and Lower Incisors, Overjet, Overbite, and Irregularity Index in Subjects with Different Depths of Curve of Spee. The Angle  97 Orthodontist, 74(3): 349–355. Begg R.P. (1954). Stone age man’s dentition: with reference to anatomically correct occlusion, the etiology of malocclusion, and a technique for its treatment. American Journal of Orthodontics 40(4): 298-312. Bernstein, R. L., Preston, C. B., & Lampasso, J. (2007). Leveling the curve of Spee with a continuous archwire technique: A long term cephalometric study. American Journal of Orthodontics and Dentofacial Orthopedics, 131(3): 363-371. Birkeland, K., Boe, O.E., Wisth, P.J. (2000). Relationship between occlusion and satisfaction with dental appearance in orthodontically treated and untreated groups. A longitudinal study. European Journal of Orthodontics, 22: 509-518. Bishara, S. E., & Jakobsen, J. R. (1997). Profile changes in patients treated with and without extractions: Assessments by lay people. American Journal of Orthodontics and Dentofacial Orthopedics, 112(6): 639–644. Bishara, S. E., Jakobsen, J. R., Treder, J. E., & Stasl, M. J. (1989). Changes in the maxillary and mandibular tooth size-arch length relationship from early adolescence to early adulthood: A longitudinal study. American Journal of Orthodontics and Dentofacial Orthopedics, 95(1): 46-59. Boley, J. C. (2002). Serial extraction revisited: 30 years in retrospect. American Journal of Orthodontics and Dentofacial Orthopedics, 121(6): 575–577. Bowman, S. J., & Johnston, L. E. (2000). The Esthetic Impact of Extraction and Nonextraction Treatments on Caucasian Patients. The Angle Orthodontist, 70(1): 3–10.  Brandt, S., & Tweed, C. H. (1967). JPO Interviews Dr. Charles H. Tweed, Part 1. Journal of Clinical Orthodontics, 1(12): 142–148.  98 Braun, M. L., & Schmidt, W. G. (1956). A cephalometric appraisal of the curve of spee in Class I and Class II, Division 1 occlusions for males and females. American Journal of Orthodontics, 42(4): 255-278 Braun, S., Hnat, W. P., & Johnson, B. E. (1996). The curve of Spee revisited. American Journal of Orthodontics and Dentofacial Orthopedics, 110(2): 206-210. Broadbent, B. H. (1937). Bolton Standards and Technique in Orthodontic Practice. The Angle Orthodontist, 7(4): 209–233.  Brodie, A. G. (1953). Late growth changes in the human face. The Angle Orthodontist, 23(3): 146–157.  Carcara, S., Preston, C. B., & Jureyda, O. (2001). The relationship between the curve of spee,relapse, and the Alexander discipline. Seminars in Orthodontics, 7: 90–99. Casko, J. S., Vaden, J. L., Kokich, V. G., Damone, J., James, R. D., Cangialosi, T. J., et al. (1998). Objective grading system for dental casts and panoramic radiographs. American Journal of Orthodontics and Dentofacial Orthopedics, 114(5): 590-599. Cerny R. (2005). What are the three most important reasons for orthodontic treatment? Are they aesthetics, aesthetics and aesthetics? Australian Orthodontic Journal 21:156-160. Cheon, S.H, Park, Y.H., Paik, K.S., Ahn S.J., Hayashi K., Yi, W.J., and Lee, S.P. (2008). Relationship between the curve of Spee and dentofacial morphology evaluated with a 3- dimensional reconstruction method in Korean adults.  American Journal of Orthodontics and Dentofacial Orthopedics 133(5): 640.e7-640.e14. Creekmore, T. D. T. (1982). Teeth want to be straight. Journal of Clinical Orthodontics, 16(11): 745–764.  99 Dawson, P. D. (1988). Evaluation, Diagnosis and Treatment of Occlusal Problems, 2e (2nd ed.). Mosby. Demirjian, A., & Goldstein, H. (1976). new-systems-for-dental-maturity. Annals of Human Biology, 3(5): 411–421. Demirjian, A., Goldstein, H., & Tanner, J. M. (1973). dental-age-assessment. Human Biology, 45(2): 211–227. Dewel, B. F. (1954). Serial extraction in orthodontics: Indications, objectives, and treatment procedures. American Journal of Orthodontics, 40(12): 906–926.  El-Dawlatly, M. M., Fayed, M. M. S., & Mostafa, Y. A. (2012). Deep overbite malocclusion: Analysis of the underlying components. American Journal of Orthodontics and Dentofacial Orthopedics, 142(4): 473-480. Farella, M., Michelotti, A., Van Eijden, T. M. G. J., & Martina, R. (2002). The curve of Spee and craniofacial morphology: a multiple regression analysis. European Journal of Oral Sciences, 110(4): 277–281. Ferrario, V. F., Sforza, C., & Miani, A., Jr. (2003). Statistical Evalusation of Monson's Sphere in Healthy Permanent Dentitions in Man. Archives of Oral Biology, 42(5): 365–369. Ferrario, V. F., Sforza, C., Poggio, C. E., Serrao, G., & Colombo, A. (1999). Three-dimensional dental arch curvature in human adolescents and adults. American Journal of Orthodontics and Dentofacial Orthopedics, 115(4): 401–405. Fudalej, P., & Årtun, J. (2007). Mandibular Growth Rotation Effects on Postretention Stability of Mandibular Incisor Alignment. The Angle Orthodontist, 77(2): 199–205.  Germane, N., Staggers, J. A., Rubenstein, L., & Revere, J. T. (1992). Arch length  100 considerations due to the curve of Spee: A mathematical model. American Journal of Orthodontics and Dentofacial Orthopedics, 102(3): 251-255. Glauser, R. O. (1973). An evaluation of serial extraction among Navajo Indian children. American Journal of Orthodontics, 63(6): 622-632. Goldberg, A. I., Behrents, R. G., Oliver, D. R., & Buschang, P. H. (1994). Facial divergence and mandibular crowding in treated subjects. The Angle Orthodontist, 83(3): 381–388.  Graber, T. M. (1971). Serial extraction: A continuous diagnostic and decisional process. American Journal of Orthodontics and Dentofacial Orthopedics, 60(6): 541–575. Grainger R.M. (1967). Orthodontic treatment priority index. National Center for Health Statistics Series 2, no. 25. Washington, DC: Public Health Service. Gresham H. A manual of orthodontics (1957). Christchurch, New Zealand: N.M. Peryer. Harris, E.F., Corruccini, R.S. (2008). Quantification of Dental Occlusal Variation: A Review of Methods. Dental Anthropology, 21(1):1-11. Hellsing, E. (1990).  Increased overbite and craniomandibular disorders-a clinical approach.  American Journal of Orthodontics and Dentofacial Orthopedics, 98(6):516- 522. Hotz, R. P. (1970). Guidance of eruption versus serial extraction. American Journal of Orthodontics, 58:1–20. Kaifu, Y., Kasai, K., Townsend, G. C., & Richards, L. C. (2003). Tooth wear and the “design” of the human dentition: A perspective from evolutionary medicine. American Journal of Physical Anthropology, 122(S37): 47–61.  Kau, C. H., Miotti, F. A., & Harzer, W. (2004). Extractions as a form of interception in the developing dentition: a randomized controlled trial. Journal of Orthodontics, 31, 107– 101 114. Kennedy, D. B., Joondeph, D. R., Osterberg, S. K., & Little, R. M. (1983). The effect of extraction and orthodontic treatment on dentoalveolar support. American Journal of Orthodontics, 84(3): 183-190. Kim, E., & Gianelly, A. A. (2003). Extraction vs Nonextraction: Arch Widths and Smile Esthetics. The Angle Orthodontist, 73(4): 354–358.  Kjellgren, B. (1947-1948). Serial extraction as a Corrective Procedure in Dental Orthopaedic Therapy. Transactions of the European Orthodontic Society, 134–160.  Koyama, T. (1979). A comparative analysis of the curve of Spee (lateral aspect) before and after orthodontic treatment--with particular reference to overbite patients. Journal Nihon University School of Dentistry, 21: 25-34. Kumar, K. P. S., & Tamizharasi, S. (2012). Significance of curve of Spee: An orthodontic review. Journal of Pharmacy & Bioallied Sciences, 4(Suppl 2), S323–8. doi:10.4103/0975-7406.100287 Kuroda, T., Motohashi, N., Tominaga, R., & Iwata, K. (1996). Three-dimensional dental cast analyzing system using laser scanning. American Journal of Orthodontics and Dentofacial Orthopedics, 110(4): 365-369. Kusnoto, B., & Evans, C. A. (2002). Reliability of a 3D surface laser scanner for orthodontic applications. American Journal of Orthodontics and Dentofacial Orthopedics, 122(4): 342–348.  Leifert, M. F., Leifert, M. M., Efstratiadis, S. S., & Cangialosi, T. J. (2009). Comparison of space analysis evaluations with digital models and plaster dental casts. American Journal of Orthodontics and Dentofacial Orthopedics, 136(1):16.e1-16.e4.  102 Lynch, C. D., & McConnell, R. J. (2002). Prosthodontic management of the curve of Spee: Use of the Broadrick flag. Journal of Prosthodontic Dentistry, 87, 593–597. Marshall, S. D., Caspersen, M., Hardinger, R. R., Franciscus, R. G., Aquilino, S. A., & Southard, T. E. (2008). Development of the curve of Spee. American Journal of Orthodontics and Dentofacial Orthopedics, 134(3): 344–352. McLaughlin, R. P., & Bennett, J. C. (2003). Finishing with the preadjusted orthodontic appliance. Seminars in Orthodontics, 9(3): 165-183. Meredith, H. V. (1959). Recent studies on growth of the body and face. American Journal of Orthodontics, 45(2): 110-124. Moorrees, C. F. A., Fanning, E. A., & Gron, A.-M. (1963). The Consideration of Dental Development In Serial Extraction. American Journal of Orthodontics.  33(1): 46–59. Moorrees, C. F. A., Lebret, L. M. L., & Frohlich, F. J. (1969). Growth studies of the dentition: A review. American Journal of orthodontics 55(6): 600–616. Motohashi, N., & Kuroda, T. (1999). A 3D computer-aided design system applied to diagnosis and treatment planning in orthodontics and orthognathic surgery. The European Journal of Orthodontics, 21(3): 263–274. Nam, S.-E., Park, Y.-S., Lee, W., Ahn, S.-J., & Lee, S.-P. (2013). Making three-dimensional Monson's sphere using virtual dental models. Journal of Dentistry, http://dx.doi.org/10.1016/j.jdent.2013.01.003. Nanda, R.S. (1971). Growth changes in skeletal-facial profile and their significance in orthodontic diagnosis. American Journal of Orthodontics 59:501-13. Nanda, R.S, Merrill, R.M. (1994). Cephalometric assessment of sagittal relationship between maxilla and mandible.  American Journal of Orthodontics 105(4):328-44.  103 Osborn, J.W. (1993). Orientation of the masseter muscle and the curve of Spee in relation to crushing forces on the molar teeth of Primates. American Journal of Physical Anthropology 92 (1):99-106.  O'Shaughnessy, K., Koroluk, L. D., Phillips, C., & Kennedy, D. B. (2011). Serial extractions versus late premolar extractions. American Journal of Orthodontics and Dentofacial Orthopedics, 139:510-6. Owens, S. E., Dykhouse, V. J., Moffitt, A. H., Grubb, J. E., Greco, P. M., English, J. D., et al. (2006). The case management form of the American Board of Orthodontics. American Journal of Orthodontics and Dentofacial Orthopedics, 129(3): 325–329.  Owens, S., Buschang, P.H., Throckmorton, G.S., Palmer, L., English, J. (2002). Masticatory performance and areas of occlusal contact and near contact in subjects with normal occlusion and malocclusion. American Journal of Orthodontics and Dentofacial Orthopedics 121 (6): 602-609. Pandis, N., Polychronopoulou, A., Sifakakis, I., Makou, M., Eliades, T. (2010). Effects of levelling of the curve of Spee on the proclination of mandibular incisors and expansion of dental arches: a prospective clinical trial. Australian Journal of Orthodontics 26 (1):61-5. Papandreas, S., Buschang, P. H., Alexander, R. G., & Kennedy, D. B. (1993). Physiologic drift of the mandibular dentition following first premolar extractions. The Angle Orthodontist, 63(2):127–134. Premkumar, S. (2011). Textbook of Craniofacial Growth. Jaypee Brothers Medical Publishers. Preston, C. B., Maggard, M. B., Lampasso, J., & Chalabi, O. (2008). Long-term effectiveness  104 of the continuous and the sectional archwire techniques in leveling the curve of Spee. American Journal of Orthodontics and Dentofacial Orthopedics: 133(4): 550–555.  Proffit, W. R., Fields, H. W., Jr, & Sarver, D. M. (2012). Contemporary Orthodontics, 5e (5 ed.). Mosby. Prosthodontics, A. O. (1994). The Glossary of Prosthodontic Terms. Mosby. Redlich, M., Weinstock, T., Abed, Y., Schneor, R., Holdstein, Y., & Fischer, A. (2008). A new system for scanning, measuring and analyzing dental casts based on a 3D holographic sensor. Orthodontics & Craniofacial Research, 11(2): 90–95.  Ringenberg, Q. M. (1967). Influence of serial extraction on growth and development of the maxilla and mandible. American Journal of Orthodontics, 53(1): 19-26.  Schumacher, G. H. (1985). Die Funktionelle Anatomie des orofazialen Systems. Shannon, K. R., & Nanda, R. S. (2004). Changes in the curve of Spee with treatment and at 2 years posttreatment. American Journal of Orthodontics and Dentofacial Orthopedics, 125:589-96. Shaw W.C. (1981). The influence of children’s dentofacial appearance on their social      attractiveness as judged by peers and lay persons. American Journal of Orthodontics and Dentofacial Orthopedics, 79:399-415. Shaw, W.C., Gabe, M.J., Jones, B.M. (1979). The expectations of orthodontic patients   in South Wales and St. Louis, Missouri. British Journal of Orthodontics, 6:203-205. Shaw, W.C., Rees, G., Dawe, M., Charles, C.R. (1985). The influence of dentofacial appearance on the social attractiveness of young adults. American Journal of Orthodontics and Dentofacial Orthopedics, 87:21- 26. Sicher, H. (1949). Oral Anatomy. St. Louis: Mosby.  105 Simons, M. E., & Joondeph, D. R. (1973). Change in overbite: A ten-year postretention study. American Journal of Orthodontics and Dentofacial Orthopedics, 64(4): 349–367. Sinclair, P. M., & Little, R. M. (1983). Maturation of untreated normal occlusions. American Journal of Orthodontics, 83(2): 114-123. Sohmura, T., Kojima, T., Wakabayashi, K., & Takahashi, J. (2000). Use of an ultrahigh-speed laser scanner for constructing three-dimensional shapes of dentition and occlusion. The Journal of Prosthetic Dentistry. Mosby-Year Book. Spee, F. G., Biedenbach, M. A., Hotz, M., & Hitchcock, H. P. (1980). The gliding path of the mandible along the skull. Ferdinand Graf Spee (1855-1937), prosector at the Anatomy Institute of Kiel. The Journal of the American Dental Association, 100(5): 670–675. Stephens, C. K., Boley, J. C., Behrents, R. G., Alexander, R. G., & Buschang, P. H. (2005). Long-term profile changes in extraction and nonextraction patients. American Journal of Orthodontics and Dentofacial Orthopedics, 128:450-7. Stevens, D. R., Flores-Mir, C., Nebbe, B., Raboud, D. W., Heo, G., & Major, P. W. (2006). Validity, reliability, and reproducibility of plaster vs digital study models: Comparison of peer assessment rating and Bolton analysis and their constituent measurements. American Journal of Orthodontics and Dentofacial Orthopedics, 129(6): 794–803.  Traebert, E.S., and Peres, M.A. (2007). Do malocclusions affect the individual’s oral health-related quality of life? Oral Health Prevention in Dentistry, 5:3-12. Vaden, J. L., Dale, J. G., & Klontz, H. A. (2000). The Tweed-Merrifield Edgewise Appliance: Philosophy, Diagnosis and Treatment. (T. M. Graber & R. I. Vanarsdall, Eds.) (34 ed.). Recherche. Wagner, M., & Berg, R. (2000). Serial Extraction or Premolar Extraction in the Permanent  106 Dentition? Journal of Orofacial Orthopedics, 61(3): 207–216. Weber, A. D. (1969). A longitudinal analysis of premolar enucleation. American Journal of Orthodontics. 56(4): 394-402. Wilson, J. R., Little, R. M., Joondeph, D. R., & Doppel, D. M. (1999). Comparison of soft tissue profile changes in serial extraction and late premolar extraction. The Angle Orthodontist, 69(2): 165–172. Yoshihara, T., Matsumoto, Y., Suzuki, J., Sato, N., & Oguchi, H. (2000). Effect of serial extraction alone on crowding: Spontaneous changes in dentition after serial extraction. American Journal of Orthodontics and Dentofacial Orthopedics, 118: 611–616.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0167464/manifest

Comment

Related Items