Dental and skeletal effects of non-surgical Class III treatment on cone-beam computed tomography: A systematic review

Search strategy

The review followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Electronic searches in PubMed-Medline and Web of Science combined MeSH terms and keywords including “Class III malocclusion,” “skeletal Class III treatment,” “facemask,” “orthopedic protraction,” “camouflage orthodontics,” and “cone-beam computed tomography.” Boolean operators AND/OR were applied appropriately. No language or date restrictions were initially imposed. Reference lists were hand-searched for additional studies. Search strategy details aligned with the PRESS checklist^[11,12] are in the Supplementary Material.

Inclusion and exclusion criteria

Study selection and data extraction

Records were imported into Systematic Review Accelerator.^[11,12] Two reviewers independently screened titles/abstracts, then full-text articles. Disagreements were resolved by discussion or third reviewer consultation. The PRISMA flow diagram [Figure 1] details the selection process. Standardized forms extracted: study characteristics, patient demographics, treatment protocols, control groups, follow-up duration, and key outcomes (SNA, SNB, ANB, WITS, incisor inclinations, and temporomandibular joint (TMJ) changes).

Close

Figure 1:

Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 flow diagram for systematic reviews.

Export to PPT

Quality assessment

RCTs were evaluated using the Cochrane Risk of Bias 2 (RoB) tool; observational studies using the Newcastle-Ottawa Scale. Two reviewers performed assessments independently, resolving discrepancies by consensus. Studies were rated as low, moderate, or high RoB [Table 1].

Table 1: Risk of bias assessment.

Study (Author, Year)	Design	Domain 1 (Random/selection)	Domain 2 (deviations/comparability)	Domain 3 (missing data/ascertainment)	Domain 4 (outcome measurement)	Domain 5 (selective reporting)	Overall risk
Almuzian et al., 2019[18]	Observational				—	—
Onem Ozbilen et al., 2019[25]	RCT				—	—
Papadopoulou et al., 2017[21]	Observational				—	—
Loca-apichai and Jein-Wein Liou, 2022[28]	Observational				—	—
Sitaropoulou et al., 2020[30]	Observational				—	—
Fischer et al., 2018[19]	RCT				—	—
Özbilen et al., 2021[27]	Observational				—	—
Paredes et al., 2020[26]	Observational				—	—
Ding et al., 2023[23]	Observational				—	—
Chen et al., 2022[22]	Observational				—	—
Hino et al.2013[6]	Observational				—	—
Liang et al., 2021[17]	RCT
De Clerck et al., 2012[14]	Observational				—	—
Nguyen et al., 2011[13]	Observational				—	—
Nguyen et al., 2014[20]	Observational				—	—
Guo et al., 2020[24]	Observational				—	—
Lee et al., 2016[29]	Observational				—	—
Khwanda et al., 2022[16]	RCT

: Low risk of bias, : Some concerns or moderate risk, : High risk of bias. For RCTs. Domain 1: Bias from randomization process, Domain 2: Bias due to deviations from intended interventions, Domain 3: Bias due to missing outcome data, Domain 4: Bias in measurement of the outcome, domain 5: Bias in selection of the reported result. For observational studies, Domain 1: Selection of study groups, Domain 2: Comparability of groups, Domain 3: Ascertainment of outcome. “—”: Indicates a domain not applicable to that study design. RCT: Randomized controlled trial

Data synthesis

Data were synthesized qualitatively and quantitatively with descriptive summaries in [Tables 2 and 3]. Heterogeneity in treatment protocols, patient age, and outcome measures precluded formal meta-analysis; results are presented by treatment modality for clinical decision-making.

Study selection

Database search yielded 780 records. After removing duplicates, 703 titles/abstracts were screened. Following a full-text evaluation of 42 articles, 18 studies met the inclusion criteria.^[15] Included studies were published from 2011–2022.

Study characteristics

Among 18 studies: two RCTs,^[16,17] seven prospective controlled studies,^{[6,13,14,16-21]} and nine retrospective analyses.^[22–30] A total of 494 Class III patients were evaluated. Sample sizes ranged from 10-46 patients. Most studies focused on growing children/adolescents (ages ~8–15 years) undergoing interceptive treatment; several included young adults undergoing camouflage therapy. Active treatment durations ranged from 6 months to 2 years; some included post-treatment follow-up of 6 months to 2–3 years.

Interventions

Non-surgical Class III treatment modalities comprise four main approaches with distinct biomechanical mechanisms.

RoB

Both RCTs had some concerns (lack of blinding). Observational studies scored moderate on NOS, with common weaknesses being a lack of control groups or baseline differences. Most used standardized measurement methods [Table 1].

Outcome measures

The results summarized in [Table 2] are stratified by treatment modality.

Maxillary changes (SNA)

• Most effective (≥2.5°): Alt-RAMEC/FM 2.73 ± 0.99°;^[25] RME/FM 2.68 ± 1.10°;^[33,35] and BAMP 2.65 ± 1.02°.^[17,18,20]

• Moderate (1.5-2.5°): Tooth-borne expansion/protraction 1.85 ± 1.23°;^[27] elastics/Alt-RAMEC/MARME 1.87 ± 1.06°.^[18]

• Limited (≤1.5°): MEAW (non-significant);^[24] chin cup (-0.34 ± 0.74°) [Figure 2].^[27]

Close

Figure 2:

SNA angle changes by treatment modality.

Export to PPT

Mandibular changes (SNB)

• Most pronounced reduction (≥0.8°): Alt-RAMEC/FM (-1.02 ± 0.99°);^[25] RME/FM (-0.89 ± 0.74°);^[30] and tooth-borne/FM (−0.80 ± 0.51°).^[27]

• Modest reduction (0.4-0.6°): BAMP (−0.58 ± 0.86°).^[17,18,20]

• Minimal/No effect: MEAW (−0.02 ± 1.56°) [Figure 3].^[24]

Close

Figure 3:

SNB angle changes by treatment modality.

Export to PPT

Interarch sagittal (ANB, WITS)

• ANB - Most effective (≥3°): Alt-RAMEC/hybrid MARME/elastics (3.95 ± 0.57°).^[18]

• ANB - Moderate (2–3°): BAMP (2.42 ± 1.22°);^[17,18,20] Alt-RAMEC/FM (2.26 ± 0.89°);^[25] and RME/FM (2.17 ± 1.14°).^[30]

• ANB - Limited: MEAW (0.78 ± 1.78°).^[24]

• WITS - Most effective (≥6 mm): Alt-RAMEC/FM (7.59 ± 2.14 mm).^[25]

• WITS - Moderate (4–6 mm): RME/FM (5.13 ± 1.76 mm);^[30] and BAMP (4.87 ± 1.24 mm) [Figure 4].^[17,18,20]

Close

Figure 4:

WITS measure changes by treatment modality.

Export to PPT

Advantages of CBCT over 2D cephalometry

This review evaluated non-surgical Class III treatment using CBCT, which offers critical advantages over conventional 2D lateral cephalography. While most included studies reported conventional measurements (SNA, SNB, ANB, and WITS) for comparability with existing literature, the utilization of CBCT imaging provided methodological improvements and unique assessment capabilities that extend beyond traditional cephalometric analysis.

CBCT eliminates fundamental limitations inherent to 2D cephalometry, including projection errors from geometric magnification (varying 7–15% depending on patient-film distance), superimposition of bilateral structures complicating landmark identification, and head positioning variability affecting angular measurements.^[7,8] These technical improvements translate to enhanced measurement reproducibility for longitudinal assessment, with studies reporting superior intra- and inter-examiner reliability for CBCT-derived measurements compared to conventional radiographs.^[7,8] This increased precision is particularly valuable for detecting small but clinically significant changes in growing patients where treatment effects must be distinguished from normal development.

The most important unique contribution is TMJ evaluation, which represents the defining advantage of CBCT over two-dimensional imaging. Glenoid fossa remodeling quantification – demonstrating anterior eminence modification (1.38 ± 1.03 mm) and posterior wall resorption (-1.34 ± 0.60 mm)^[14] – provides direct visualization of adaptive bone remodeling that is completely invisible on lateral cephalograms where the TMJ is obscured by overlapping cranial base structures. Similarly, three-dimensional condylar displacement patterns (44% backward, 40% backward and downward, and 16% backward and upward)^[15] can be precisely quantified and categorized in CBCT datasets through multi-planar analysis and 3D reconstructions, whereas conventional cephalometry offers only indirect, two-dimensional projections of condylar position with significant measurement error. These findings provide crucial evidence that Class III orthopedic treatment induces adaptive rather than pathological TMJ remodeling, directly addressing long-standing clinical concerns about potential iatrogenic joint damage from FM and protraction therapies.^[14,29]

In addition, CBCT enables transverse dimension assessment^{[22,26,27,30]} – completely impossible on sagittal cephalograms – allowing discrimination between skeletal and dental contributions to expansion (e.g., MSE achieving 96.58% skeletal effect with angular measurements versus 60.16% with linear measurements, demonstrating measurement method importance). Asymmetry evaluation, including mandibular deviation quantification (mean reduction 0.43mm with bilateral condylar remodeling),^[30] provides clinically relevant information for treatment planning that cannot be obtained from anteroposterior radiographs alone. Dentoalveolar bone thickness changes^[27] and voxel-based superimposition for precise three-dimensional displacement quantification^[7,9,10] represent additional CBCT capabilities impossible with two-dimensional imaging, though these advanced analytical methods remain underutilized in current literature.

Comparison with existing two-dimensional systematic reviews^[31-39] reveals similar sagittal skeletal change magnitudes (SNA approximately 2.0–2.7° and SNB approximately 0.5–1.0°), suggesting that CBCT-derived measurements of conventional angular parameters are consistent with traditional cephalometric findings and confirming comparability across imaging modalities for these established metrics. This concordance validates the reliability of both measurement approaches for sagittal assessment while simultaneously highlighting CBCT’s value: Similar accuracy for conventional measurements plus unique capabilities for TMJ evaluation, transverse assessment, volumetric analysis, and multi-dimensional treatment visualization. Therefore, CBCT adds the critical dimensions of enhanced measurement reliability through elimination of projection artifacts, comprehensive TMJ documentation providing evidence of treatment safety, transverse dimension assessment enabling skeletal versus dental differentiation, and multi-dimensional evaluation supporting a more complete understanding of treatment effects.

Treatment modality comparison and clinical implications

While surgery remains definitive treatment for severe adult skeletal Class III malocclusion, various non-surgical protocols demonstrated significant short-term improvements, particularly in growing patients, with substantial modality differences reflecting distinct biomechanical mechanisms.^[32,33]

For maxillary advancement, Alt-RAMEC/FM (2.73 ± 0.99°), RME/FM (2.68 ± 1.10°), and bone-anchored protraction (2.65 ± 1.02°) produced the most substantial SNA increases, while camouflage approaches (MEAW) showed minimal skeletal effect. The similarity between Alt-RAMEC and conventional RME protocols (difference <0.1°) suggests that the theoretical advantages of alternating expansion-constriction cycles may not translate to dramatically superior skeletal outcomes in clinical practice, though both achieve clinically meaningful maxillary advancement.^{[25,27,30,33,35]}

Skeletal anchorage produced favorable maxillary movement with minimal dental tipping,^[6,13,14,17] confirming that elimination of dental anchorage allows more purely skeletal effects by applying forces directly to skeletal bases. However, maxillary advancement magnitude remained comparable to properly applied FM therapy with dental anchorage, suggesting tooth-borne appliances can achieve substantial orthopedic effects when protocols are optimized. The primary advantage of skeletal anchorage may therefore lie in reducing relapseable dental compensations rather than dramatically increasing skeletal response magnitude.^[6,14]

Dentoalveolar changes varied substantially, with MEAW producing the greatest lower incisor retroclination (8.8 ± 4.03°), consistent with dentoalveolar compensation rather than skeletal modification.^[24] This contrasts sharply with bone-anchored protraction, where lower incisor changes remained minimal, reflecting the difference between camouflage (dental) and orthopedic (skeletal) treatment philosophies.

For adults refusing surgery (70–86% of Class III surgical candidates in some populations^[40-42]), non-surgical alternatives, including MARPE for transverse deficiencies^[5,22] and camouflage with lower extraction^[1,24] may achieve acceptable occlusion, though skeletal improvements remain limited. Realistic patient expectations are critical, with goals focused on functional occlusion rather than complete skeletal normalization.

TMJ findings – The unique 3D contribution

TMJ assessment represents CBCT’s defining advantage for evaluating Class III treatment, providing evidence of adaptive rather than pathological joint changes impossible to document through two-dimensional cephalometry.

Consistent glenoid fossa remodeling – bone apposition anteriorly (anterior eminence modification 1.38 ± 1.03 mm), resorption posteriorly (posterior wall −1.34 ± 0.60 mm)^[14,29] – demonstrates fossa adaptation to posteriorly repositioned condyle following protraction and growth redirection. This represents physiologically appropriate, not pathological, remodeling. The biological mechanism involves balanced bone deposition and resorption creating space for the condyle in its new position, with remodeling magnitude proportional to mandibular repositioning degree.^[14,29] This adaptive response occurs primarily during active growth but demonstrates plasticity even in young adults, as evidenced by measurable adaptations following MARPE treatment.^[22] The temporal sequence shows initial condylar displacement during active protraction, followed by gradual fossa remodeling over subsequent months, suggesting treatment effects stabilize through biological adaptation rather than merely mechanical repositioning.^[14]

Three distinct condylar displacement patterns (44% backward, 40% backward/downward, and 16% backward/upward)^[15] reveal individual variation in mandibular adaptation, possibly relating to growth pattern, treatment mechanics, or biological response differences, though studies did not stratify by vertical pattern. The predominance of posterior displacement (combined 84%) aligns with therapeutic objectives of mandibular growth restraint and maxillary advancement, while variability in the vertical component may reflect individual differences in mandibular rotation. Patients showing backward/upward displacement may represent those with counterclockwise rotation, potentially associated with more favorable facial profile outcomes, though this requires verification through vertical dimension stratification.^[15]

MEAW in adults produced no significant TMJ changes,^[24] while bone-anchored protraction in growing patients induced substantial displacement and remodeling,^[14,15] highlighting the importance of growth potential. This contrast illustrates the fundamental difference between dentoalveolar compensation (achieving occlusal correction without skeletal TMJ involvement) and orthopedic skeletal modification (necessitating joint adaptation). However, MARPE induced measurable adult condylar repositioning,^[22] suggesting some adaptive capacity persists beyond growth, though the magnitude remained less than in growing patients. This has important implications for adults refusing surgery, indicating transverse maxillary expansion can produce skeletal adaptation even in mature individuals, though expectations must be calibrated accordingly.

No studies reported adverse TMJ findings (condylar resorption, degenerative changes, and TMD symptoms), providing safety reassurance across all modalities evaluated, though limited follow-up (maximum 2–3 years) precludes definitive long-term conclusions. This absence of pathological findings across 494 patients and diverse protocols provides preliminary safety evidence, though extended follow-up remains critical as degenerative changes may manifest years post-treatment. Future research should incorporate systematic TMD assessment, clinical examination, and long-term CBCT evaluation (minimum 5-year follow-up) to definitively establish safety profiles.^{[14,15,22,24,29]}

Clinical implications

• (1) Multiple modalities (Alt-RAMEC/FM, RME/FM, and bone-anchored protraction) achieve comparable maxillary advancement (~2.5-2.7°), suggesting clinician/patient preference regarding appliance type, duration, and invasiveness can guide selection among equivalently effective approaches.

• (2) Skeletal anchorage minimizes dentoalveolar side effects but does not dramatically increase skeletal response versus properly applied tooth-borne protraction, though it may enhance stability by reducing relapseable dental compensations.^[14]

• (3) Transverse expansion with bone-borne anchorage produces predominantly skeletal effects; CBCT documentation is crucial for distinguishing skeletal from dental expansion.^[26]

• (4) TMJ adaptation occurs consistently with orthopedic Class III treatment in growing patients, with patterns consistent with adaptive rather than pathological remodeling, though long-term follow-up remains necessary.

• (5) For adults declining surgery, camouflage can achieve acceptable occlusion through dentoalveolar compensation, though skeletal correction remains limited; CBCT assessment quantifies dental versus skeletal correction components.

Limitations

Lack of untreated control groups in most studies (only two of 18 included controls) limits distinguishing treatment effects from natural growth, particularly for SNB, where mandibular growth continuation could confound interpretation.^[30]

Ethnicity was rarely reported (estimated 60% Asian, 40% Western based on author affiliations/locations). This is significant because Class III etiology patterns differ substantially between ethnic groups (maxillary deficiency is more prevalent in Asians; mandibular prognathism in Caucasians).^[1,2] Future research should explicitly report ethnicity and stratify outcomes.

Follow-up duration was limited (active treatment 6 months to 2 years; post-treatment retention rarely exceeding 2–3 years). Long-term stability remains uncertain, particularly for growing patients, where residual mandibular growth may negate treatment effects.^[21] Future research requires a minimum 5-year follow-up.

Vertical skeletal pattern was rarely stratified (only one study excluded low-angle patients^[24]). Class III treatment effects and growth patterns differ substantially between high/low-angle patients; future research should stratify by vertical dimension.

Despite CBCT utilization, most studies reported primarily conventional 2D measurements rather than 3D volumetric analysis. Few reported volumetric changes^[26] or utilized voxel-based superimposition,^{[7,9,10,13-15,20]} representing missed opportunities. Standardized 3D protocols are needed.^[10]

Patient compliance was rarely reported, introducing potential bias as FM/functional appliance effectiveness depends heavily on cooperation.