Introduction
Class II malocclusion, particularly its unilateral presentation, poses significant challenges in orthodontics because of asymmetrical occlusal relationships and associated midline deviations. Although clear aligner therapy offers esthetic and practical benefits, its predictability in achieving complex movements, such as unilateral molar distalization, remains limited.
Methods
This retrospective study included 51 adult patients treated with clear aligners for Class II subdivision malocclusion taken from the Australasian Aligner Research Database. Digital models at baseline, from the virtual treatment plan (planned), and at the outcome achieved after a single course of aligner treatment were analyzed using Geomagic Control X software to assess unilateral molar Class II correction, overjet changes, and midline discrepancy. A subgroup of 12 patients was further analyzed to evaluate unilateral maxillary molar distalization and associated mesiodistal tipping of the maxillary molars.
Results
On average, 36.8% of the planned molar relationship correction and 23.8% of the planned midline correction were achieved. Overjet increased rather than decreased, representing a 28.8% shortfall in the planned overjet correction. Subgroup analysis indicated that 53% of the planned molar distalization was achieved (mean, 1.6 mm vs 3.0 mm planned), with a strong correlation ( r = 0.98) between distalization and molar tipping, averaging 2.6° per millimeter.
Conclusions
Clear aligner therapy demonstrates limited predictability in achieving unilateral Class II molar correction, overjet reduction, and midline improvement in subdivision malocclusions. Substantial unilateral molar distalization remains challenging and is frequently accompanied by undesirable crown tipping. Age was also identified as a significant predictor of distalization efficiency.
Highlights
-
•
Clear aligners achieved 36.8% of the planned unilateral Class II molar correction.
-
•
Overjet increased rather than decreased (−28.8% of planned correction).
-
•
Achieved molar distalization averaged 1.6 mm vs 3.0 mm planned in the subgroup.
-
•
Each millimeter of distalization was accompanied by ∼2.6° of unwanted molar crown tipping.
-
•
Clear aligners show limited predictability for Class II subdivision.
Clear aligner therapy (CAT) has been widely promoted for its esthetic appeal, comfort, and precise control over tooth movement. However, while CAT has gained popularity, studies consistently demonstrate significant discrepancies between predicted and achieved outcomes for movements commonly associated with Class II correction, including overbite control, ,,, maxillary incisor retraction, posterior maxillary arch expansion, , and distalization of the maxillary posterior teeth. ,
Class II subdivision malocclusion, an asymmetrical variant of Class II relationships, is a recognized clinical presentation in orthodontics, with a reported prevalence of approximately 10% in European populations. It is characterised by a unilateral Class II molar relationship, in which the asymmetry may originate primarily from the mandibular arch because of distal positioning of a mandibular molar (type 1), from the maxillary arch because of mesial positioning of a maxillary molar (type 2), or from a combination of both (type 3), and is often accompanied by dental midline deviations. This asymmetry complicates treatment planning because of unique biomechanical demands. While CAT offers esthetic advantages, its predictability in unilateral molar correction remains unknown.
Addressing such malocclusions requires a detailed understanding of asymmetrical mechanics and their application within each dental arch. Maxillary molar distalization is a commonly used nonextraction approach for correcting Class II malocclusion, particularly in patients with mild-to-moderate discrepancies. Traditional distalizing appliances, such as the pendulum, distal jet, and headgear, are effective but often result in unwanted side effects, including distal tooth tipping, extrusion, and anchorage loss, which can compromise vertical control and increase lower facial height. Applying force closer to the molar’s center of resistance, as achieved with some palatal-acting appliances, improves biomechanics by promoting greater translational movement and reducing undesired tipping. The advent of skeletal anchorage systems, such as mini-implants and palatal implants, has further advanced the predictability and efficiency of distalization by providing stable, compliance-independent anchorage. These systems minimize anterior anchorage loss, incisor proclination, and vertical side effects, enabling more controlled and efficient molar movement with reduced treatment duration and fewer unwanted changes to facial proportions.
CAT can be used to perform molar distalization in the correction of Class II malocclusion. According to current literature, maxillary molars can be distalized by approximately 2.2 mm using aligners with bonded composite resin attachments and Class II elastics, without significant tipping or extrusion; however, the threshold for clinical significance was not defined. Nevertheless, anchorage loss during anterior retraction remains a concern, with reduced distalization efficacy observed when anterior teeth are retracted.
Unilateral maxillary molar distalization with CAT allows correction of molar relationships while preserving midline alignment. However, to the best of our knowledge, there is no evidence on the efficacy and predictability of CAT for this purpose. Biomechanical limitations often require auxiliary aids such as Class II elastics, but their impact on clinical outcomes remains unclear. In addition, discrepancies between digital treatment predictions and actual clinical results raise concerns about CAT efficacy in asymmetrical problems such as Class II subdivision malocclusions.
This study aimed to evaluate the efficacy of CAT for the treatment of Class II subdivision malocclusions in nongrowing patients by comparing predicted outcomes with actual clinical results. We hypothesized that there is no difference between the planned and achieved outcomes for molar relationship, overjet, and maxillary dental midline position with the Invisalign appliance in patients with Class II subdivision malocclusion. The findings will enhance treatment planning, improve predictability, and optimize aligner use for the management of patients with Class II subdivision malocclusion.
Material and methods
The University of Otago Ethics Committee (HD20/004) granted approval for this retrospective study.
The patient sample for this study was selected from the Australasian Aligner Research Database, which is a database of CAT patients treated by 20 orthodontists. The study included a primary analysis of all 51 patients, evaluating planned vs achieved outcomes for molar relationship, overjet, and midline deviation. In addition, a subgroup analysis was conducted on a subset of 12 patients from the original cohort, specifically assessing the accuracy and nature of unilateral maxillary molar distalization. The full patient selection process is detailed in the Supplementary Figure . At the time of data collection, the database comprised approximately 17,000 Invisalign patients treated by orthodontists in private practice across Australia, the United States, and New Zealand. All treating clinicians had at least 10 years of experience in CAT and had treated a minimum of 300 CAT patients each. To minimize measurement and selection bias, strict inclusion and exclusion criteria were applied, and outcome measures were assessed using validated digital tools.
Eligible patients were identified according to the following inclusion criteria: Class II subdivision malocclusion; Planned distal transitional movement of ≥2 mm for 1 maxillary molar, as indicated in the ClinCheck visualization; aged >18 years; availability of digital study models capturing all teeth present at the initial, planned, and achieved stages after a single course of aligner treatment; Treated nonextraction using Invisalign SmartTrack aligners only, between 2013 and 2024; absence of third molars in the maxillary arch.
Records were excluded based on the following criteria: patients who had any treatment that significantly affects multiple tooth size and shape, patients with incomplete or missing digital records, and use of temporary skeletal anchorage devices (TSADs) or fixed orthodontic appliances in conjunction with CAT.
To accurately estimate the amount of molar distalization, eligible participants were further screened for the presence of at least 2 teeth on the contralateral (nonmoving) side that met the following stability criteria: rotation (viewed occlusally) <2°; anteroposterior (AP) movement <0.5 mm.
Available data included patient age, sex, treatment duration, number of aligners, and method of distalization (sequential vs en masse).
Stereolithography files representing the initial dentition (T0), the predicted outcome (T1), and the outcome achieved after a single course of aligner treatment (T2) were exported from ClinCheck.
The primary outcome variables were (1) changes in molar relationship on the Class II side, (2) overjet, and (3) planned end-of-treatment midline discrepancy. The molar relationship was assessed using the AP position of the mesiobuccal cusp tip of the maxillary first molar relative to the buccal groove of the opposing mandibular molar using 3-dimensional digital models. As shown in Figure 1 , a reference plane (y-plane) was constructed perpendicular to the occlusal plane and passing through the buccal groove of the mandibular molar (point b). The AP position of the mesiobuccal cusp (point c) was then measured relative to this plane to classify the relationship as Class I, half-cusp, or full-cusp Class II. Overjet and midline discrepancy values were obtained directly from the ClinCheck interface at T0, T1, and T2. Although the precise measurement algorithms used by the software are proprietary, a previous study has validated the ClinCheck values using metrology-derived values that closely correspond to clinical measurements.
T0 models oriented on the moving side, with the y-plane aligned with the mandibular first molar buccal groove, and a measurement point placed at the mesiobuccal cusp tip of the maxillary first molar. a , Y-plane generated; b , Y-plane aligned with mandibular molar buccal groove; c , A point at mesiobuccal cusp of the maxillary first moving molar, highlighted in yellow .
Secondary outcome variables included the amount and accuracy of maxillary molar distalization and the degree of molar tipping. A subgroup of 12 patients was selected based on stringent inclusion and exclusion criteria, requiring minimal movement on the contralateral (Class I) side, defined as ≤0.5 mm of translation and <5° of rotation in at least 2 teeth, to permit reliable reference-based superimposition ( Fig 2 ). Of the 51 patients, only 12 fulfilled these criteria and were therefore included in the secondary outcome analysis. These were evaluated by digital model superimposition using Geomagic Control X software (3D Systems, Rock Hill, SC), with best-fit surface registration performed using global and fine alignment modes, applying 50 iterations and an 80% point sampling ratio ( Fig 3 ).
Bright blue : flood selected teeth for the best fit alignment using selected data only. In this example, the maxillary right second molar, first molar, and second premolar are selected as a stable reference point.
Best-fit superimpositions of selected data from 4 study participants. The blue models represent the pretreatment condition, green represents the planned tooth movement, and yellow represents the posttreatment outcome after CAT. Note. Image created using Geomagic Control X.
The centroid of the maxillary first molar on the Class II side was located using the flood selection method, a surface-based selection tool in Geomagic Control X that isolates a connected region of the 3-dimensional model based on surface continuity and curvature, to ensure consistent identification of the full crown. and its linear distance from the y-plane was measured at T0, T1, and T2 ( Fig 4 , A and B ).
Examples of measurements taken on corresponding points: A, Pretreatment (T0); B, Posttreatment models (T2). Yellow indicates the geometric center of the maxillary right first molar.
Distalization accuracy was calculated as:
Tipping was quantified by measuring the angle between the cusp line (mesiobuccal to distobuccal) and the y-plane at T0 and T2 ( Fig 5 , A and B ).
Examples of angular measurements of the moving first molar: A, Pretreatment; B, Posttreatment models.
To assess intraexaminer reliability, a random sample comprising 25% of the primary group (13 patients) and 25% of the secondary group (3 patients) was selected using the random number generator in Excel software (Microsoft, Redmond, Wash). The patient samples selected from the primary and secondary groups were mutually exclusive to avoid duplication.
After a 2-week washout period, all primary outcome variables (molar relationship, overjet, and midline) and the secondary outcome variable (distalization accuracy) were remeasured by the same operator (A.H.).
To assess interexaminer reliability, a second examiner (T.W.) independently measured the same sample without access to the original measurements.
The error of the method was then calculated using the intraclass correlation coefficient (ICC) and the Dahlberg formula.
This manuscript was prepared in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for reporting observational studies. ( Supplementary Table ).
Statistical analysis
Data were analyzed using the SPSS software (version 22.0; IBM, Armonk, NY). The normality of data distribution was assessed using the Shapiro-Wilk test. A multivariable analysis was conducted for the main outcome variables, entering sex, age group, number of aligners, and type of distalization as covariates in the model. Linear regression analysis and Pearson’s correlation coefficients were employed to evaluate the accuracy of the 2 different registration methods, using the contralateral side method as the reference standard. The level of statistical significance (type I error) was set at 0.05.
Results
A total of 51 patients (42 females and 9 males) were included in the study. The flow of patients through each stage of the selection process is outlined in the Supplementary Figure. The mean number of aligners used was 49.9 ± 13.4. Descriptive statistics summarizing demographic and treatment variables for both the main and subgroup samples are presented in Table I . Intraexaminer ICC values ranged 0.92-0.95, and interexaminer ICC values ranged 0.92-0.94 for all primary and secondary outcome variables. Measurement error ranged 0.2-0.4 mm for linear measurements and 0.3° to 0.5° for angular measurements.
Table I
Descriptive statistics
| Main Study Sample N = 51 | |
| Age in years [Mean (SD)] | 32.6 (9.1) |
| Aligner number [Mean (SD)] | 49.9 (13.4) |
| Female N (%) | 42 (82.3%) |
| Male N (%) | 9 (17.6%) |
| Malocclusion type | |
| Class II div1, subdivision | 24 (47.1%) |
| Class II div2, subdivision | 27 (52.9%) |
| Type of distalization | |
| En-masse N (%) | 6 (11.8%) |
| Sequential N (%) | 45 (88.2%) |
| Subsample N = 12 | |
| Age in years [Mean (SD)] | 33.3 (9.8) |
| Aligner number [Mean (SD)] | 40.1 (16.1) |
| Female N (%) | 11 (91.7%) |
| Male N (%) | 1 (8.3%) |
| Type of distalization | |
| En-masse N (%) | 2 (16.7%) |
| Sequential N (%) | 10 (83.3%) |
Planned vs achieved corrections after a single course of aligner treatment were assessed for molar relationship, overjet, and midline deviation ( Figs 6 and 7 ). On average, 36.8% of the planned molar relationship correction was achieved, with a mean difference of +2.3 ± 1.0 mm (95% confidence interval [CI], 1.9-2.7 mm). Although the correlation between planned and achieved molar correction was statistically significant (r = 0.37; P = 0.008), the association was weak (R 2 = 0.134).
Deviation from Class I (mm) is plotted for initial, planned, and achieved measurements. In the molar relationship graph, red , 0 mm represents the ideal Class I relationship.
The plot compares the planned vs achieved measurements for 3 variables: molar relationship on the Class II side, overjet, and midline, across 51 patients. Each patient is ranked based on the discrepancy between planned and achieved values, with smaller discrepancies ranked higher. The lines and dots highlight the differences between planned ( green ) and achieved ( blue ) values, showing trends and deviations for each patient.
The mean planned overjet was 2.3 mm, while the achieved value was 3.6 mm. The mean discrepancy was +1.4 ± 1.0 mm (95% CI, 1.0-1.8 mm). This indicates an increase in overjet rather than the intended reduction, corresponding to a −28.8% shortfall in the planned correction. No significant correlation was observed between the planned and achieved values ( P = 0.204).
The mean planned end-of-treatment midline discrepancy correction was 0.3 mm, compared with an achieved end-of-treatment discrepancy of 1.2 mm. The mean difference was +1.0 ± 0.8 mm (95% CI, 0.8-1.2 mm), corresponding to 23.8% of the target. No significant association was found between planned and achieved corrections ( P = 0.135).
Multivariable analysis, conducted on standardized outcome variables expressed as percentages, revealed that age was the only significant predictor of molar relationship correction (β = −0.023; P = 0.004), indicating a 2.3% decrease in efficiency for each additional year of age. Gender, number of aligners, and distalization method were not statistically significant predictors for any outcome variables (all P >0.05).
The subgroup analysis, comprising 12 patients from the original cohort, specifically assessed the accuracy and nature of unilateral maxillary molar distalization. In this subgroup, the planned mean distalization of the maxillary first molar was 3.0 mm, whereas the achieved movement after a single course of aligner treatment was 1.6 mm, representing 53% of the planned amount. The mean discrepancy was 1.4 ± 0.6 mm (95% CI, 1.1-1.7 mm; P <0.001). Despite the underachievement, a strong correlation was observed between planned and achieved values ( r = 0.88) ( Fig 8 ).
