Determination of treatment options for Class III malocclusions in adult patients based on cephalometric values: a systematic review

The current systematic review was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). The review protocol was submitted in March 2020 and registered with the OSF/Center for Open Science. The review was also submitted in January 2020 and was successfully registered with the National Institute of Health Research Database (https://www.crd.york.ac.uk/prospero/; trial registration number: PROSPERO CRD42020165164 in July 2020.

The PICO for the study was:

Population: Adult patients aged 18 years and older presenting with a Class III malocclusion

Included were randomised controlled trials (RCTs), quasi-randomised, prospective studies, retrospective cohort, and cross-sectional studies in the present search. Non-human studies (animal or laboratory studies) and studies on patients with craniofacial malformations were excluded. No language restrictions were applied and all articles were included. Google translator was used to translate the title and abstract of non-English articles.

A comprehensive electronic database search was conducted until the 18th September 2021 with no language nor publication date restrictions. Science Direct, PubMed, Ovid MEDLINE, Web of Science, Scopus and Open Grey were databases searched using Medical Subject Headings and free-text words for PubMed which were optimised for each database (Table I). A hand search of three orthodontic journals, namely the Angle Orthodontist, the European Journal of Orthodontics and the American Journal of Orthodontics and Dentofacial Orthopedics was conducted between Jan 2005 until September 2021. In addition, the reference lists of all the included studies were screened. If necessary, corresponding authors were contacted to obtaining clarifications or for additional data extraction.

The literature search, study inclusion, methodology quality assessment and data extraction were performed independently and in duplicate by two authors (N.N.Z. and N.A.A.). Any conflicts were resolved by contacting the third, fourth and fifth authors (S.S., W.M.C. and M.M.S.F.). The articles were first screened by title and abstract. Relevant articles were then screened by full text and compared against the inclusion/exclusion criteria for a final selection.

Data extraction was carried out by two review authors independently (N.N.Z. and N.A.A.) using a standardised data extraction form. The data extraction was cross-checked by the third author (S.S.). Data extraction applied the following items: demographic characteristics of the studies, cephalometric cut-off value between orthodontic camouflage and orthognathic surgery in Class III malocclusions and post-treatment complications.

The risk of bias was analysed by using the RoB Cochrane tool for RCTs and prospective non-randomised studies. For retrospective cohort and cross-sectional studies, the risk of bias was examined with the aid of a modified version of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist comprising seven items related to: (1) study design, (2) study setting, (3) participants criteria, (4) sample size, (5) variable description, (6) outcome measurements, and (7) statistical analysis. The quality of the studies was categorised as weak (3 or less), moderate (4 or 5) and high quality (6 or more) by two pairs of independent reviewers in duplicate (N.N.Z. & N.A.A.) and (S.S. & W.M.C.). Any disagreements were resolved by discussion and consultation with the fifth author to reach a consensus (M.M.S.F.).

The studies were eligible for quantitative synthesis if two or more reported the same outcome, and used the same measurement unit and tools. The data were combined using MedCalc Software Ltd (Belgium-version 19.4.1) for quantitative analysis. Otherwise, the results were summarised qualitatively. Statistical heterogeneity was inspected using the I-squared index, assigning values of 25%, 50% and 75% corresponding to low, moderate and high heterogeneity. According to the I² test, the random-effect model was applied to studies with more than 50% heterogeneity.¹⁶

No subgroup analysis was undertaken.

The comprehensive search yielded a total of 13033 articles of which 3414 were duplicates. The remaining 9619 articles were screened by title and abstract. A total of 9598 articles were excluded by title and 9 articles were excluded by abstract. A final sample of 12 articles were then screened by full text after which 6 studies met the inclusion criteria. The results from the database search are presented in Table I. Figure 1 illustrates the study selection and identification process. Six articles^4,13,17–22 were excluded as two studies^4,13 recruited participants below 18 years of age, one study¹⁷ was a review article, and three studies^16–20 did not define the cephalometric values differentiating orthognathic surgery and camouflage.

Figure 1.

The characteristics of the selected studies are presented in Table II. No RCTs or quasi-randomised nor prospective clinical trials were identified. All of the six currently included studies^{14,15,21–24} were retrospective in nature. The total sample sizes of the studies ranged from 47 to 175. The mean age of the patients was greater than 18 years of age in four studies,^15,22–24 two studies^14,21 did not specify the mean age, but included adult patients in the studies. No data were reported on the gender of the patients in two studies,^21,22 three studies^14,15,24 had more female than male patients and one study²³ recruited equal numbers of male and female patients.

Two studies^15,24 were judged to be of high quality and four studies^14,21–23 were of moderate quality (Table III). All of the moderate quality studies did not completely report on the research parameters and sample size calculations.

At the time of registration with OSF/Center for Open Science, there was no other identified systematic review which investigated the cephalometric cut-off values and complications associated with orthodontic camouflage and orthognathic surgery in Class III malocclusion cases. There is still clinical uncertainty in deciding the appropriate management of borderline orthognathic Class III cases which has prompted the need to identify potential threshold indicators. The decision was mostly dependent on the clinician’s experience and limited evidence. The current review aimed to provide an insight into the cephalometric indicators that might be used as a guide in the decision-making process associated with borderline Class III malocclusion management, its complications and aesthetic perceptions.

A systematic analysis conducted in 2011 that aimed to predict the outcome of orthodontic treatment in Class III malocclusion cases was identified during the literature search; however, it concentrated on the orthodontic/orthopaedic care of growing subjects, which was outside the scope of this review.²⁵ Overall, only retrospective studies were found to match the selection criteria and were appraised using the modified version of the STROBE checklist which ensured a clear presentation of general reporting for descriptive observational cross-sectional studies.

Commonly identified skeletal sagittal parameters used to discriminate treatment modalities in Class III cases are the ANB angle and Wits appraisal. Due to doubts on the validity of the ANB angle in measuring sagittal jaw relations,^26,27 the Wits appraisal has been supplemented. The literature revealed that the surgical group tended to have a smaller ANB angle and Wits appraisal value compared to the camouflage group. Other studies found that the Wits appraisal was the most useful predictor to distinguish patients prior to treatment whether independently^15,23 or in combination with other parameters.^14,22 A recent study which performed a discriminant analysis found that those with a Wits appraisal of more than −5.8 mm may be successfully treated by orthodontics alone.¹⁵ An earlier study using a ROCs analysis found −11.18 mm to be the Wits threshold.²³ The purpose of the Wits appraisal is to dissociate the maxillamandibular relationship from the cranial base²⁷ and so the rotation of the jaw complex relative to the cranial base does not affect the severity of the jaw disharmony. The identification of a functional occlusal plane can be a challenge and considered a potential source of error, and so the validity of the Wits appraisal deserves to be questioned.

The Holdaway angle in an ideal face is 7°–15° when facial convexity is normal.²⁸ The more forward the position of soft tissue Pogonion in relation to the upper lip, the more concave is the lateral profile, indicating a more severe Class III discrepancy. Previous studies have suggested that the Holdaway angle is a decisive factor.^15,21 The studies suggested that a patient with a greater Holdaway angle than the threshold, reflected a more convex profile which could be successfully treated by orthodontics alone. After following a stepwise discriminant analysis, one study found that the Holdaway angle was the sole decisive parameter.²¹

A Class III malocclusion is likely to have a level of dental compensation which helps to maintain function and masks the underlying skeletal discrepancy. Lower incisor angulation was found to be a discriminating factor in a study which used the ROC analysis²³ with a cut-off value of 80.8°. The use of overjet, overbite, maxillary-mandibular ratio and gonial angle as cut-off discriminants was recommended by only one study.²³ A study by Kerr et al.¹³ advocated the use of the ANB angle to determine treatment options differentiating camouflage or surgery. However, the study was excluded from the present review because the control group consisted of young patients who were still exhibiting potential growth. None of the other studies reported the use of the ANB angle as a determining cut-off factor.

Asymmetry is also a consideration for surgical intervention in Class III malocclusion cases as these patients may present with a level of mandibular deviation. One study²² considered the implication of mandibular asymmetry in distinguishing possible treatment between the two groups by applying a multivariate equation. Both studies that used the discriminant analysis equation had a high success rate (91.3– 92%) for correctly classifying the predicted treatment plan.^14,22 The addition of the transverse component to the discriminant analysis equation improved the outcome of correctly classifying the surgical groups from 86.4% to 92.7%. The input of selected parameters into an equation may increase the accuracy of the prediction compared to a single parameter. However, this takes a little longer because each measurement must be entered into the formula to determine the outcome.

It is appreciated that the decision to choose surgery in borderline cases is not solely reliant on cephalometric values and involves clinical judgement as well as the patient’s wishes. Published research has indicated that enhancing dental and facial aesthetics is the most common reason people seek orthodontic treatment.^29,30 The secondary outcome of the present review found one article³⁰ that studied the orthodontic perception of facial profile attractiveness in borderline surgical Class III malocclusion cases and found that facial profile attractiveness was significantly improved in both groups and no difference in the final outcome was apparent. The surgical group had a lower attractiveness score and a slightly severe Class III molar relationship at the outset and therefore had a greater improvement during treatment. This indicates that there are additional clinical factors that may guide the judgement of clinicians when deciding between camouflage treatment or surgery.

Significant clinical heterogeneity was found in the present review due to the variability of the applied inclusion criteria for the recruitment of the Class III malocclusion cases. The majority of the studies involved three criteria or more. The similarity of the parameters in borderline malocclusions complicates treatment planning which leaves treatment choices to the preference of the clinician.³¹ The inclusion criteria for the parameters such as overjet, Wits analysis and the ANB angle in many of the studies were uni-directional (e.g. ANB < 0°, Wits < −1 mm). It is generally accepted that the farther away the value of the characteristic is from the norm, the more likely a surgical plan would be adopted. Therefore, the selection parameters for the inclusion criteria in borderline cases should have a minimum and a maximum limit that can be present in both surgical and camouflage groups which was implemented in one study.¹⁵ There is a large variation in the inclusion criteria for the Wits appraisal ranging from −7.5 to −1 mm. The variation likely depended on the population norm as Asian populations tend to have a greater negative value compared to Caucasian norms. Minor methodological diversity was noted in the present review with the use of a multivariate analysis such as the discriminant analysis^14,15,21,22 and the ROC analysis²³ applied in an attempt to identify parameters and associated critical values to distinguish appropriate treatment options in borderline Class III malocclusion cases.

The present review identified only retrospective studies which rendered the level of evidence of moderate to high quality. The complications of either treatment modality or patient reported outcome could not be addressed since the included studies did not explore this area. There was an inability to perform a Grading of Recommendations Assessment, Development and Evaluation (GRADE) analysis because this analysis is only limited for interventions and a diagnostic systematic review.

The current review intended to report the complications of orthodontic camouflage and orthognathic surgery (gingival recession, root resorption, dehiscence, tooth mobility, tooth sensitivity and black triangles) and aesthetic perceptions by laypersons/orthodontists. However of concern, a lack of substantive evidence was identified.

Future randomised controlled trials should be planned to further investigate the cephalometric parameter thresholds in managing borderline Class III malocclusion cases. Researchers are encouraged to report all findings including the success of treatment outcomes, and patient-reported outcomes along with associated adverse effects.