Longitudinal outcomes of bone-anchored maxillary protraction treatment

Abstract

Introduction:

Bone-anchored maxillary protraction (BAMP) is a well-tolerated treatment for Class III orthopedic effects during adolescence. BAMP aims to improve Class III skeletal relationships with minimal dentoalveolar compensations, in the hope of reducing the need for orthognathic surgery. Our aim was to evaluate the short- and long-term effects of BAMP on measures of tooth and jaw positions and definitive treatment outcomes in patients with Class III malocclusion treated with (+) and without (−) BAMP during adolescence.

Methods:

Patients treated with BAMP (n = 45) in adolescence were compared with an adolescent Class III control group (n = 39) before comprehensive treatment. Lateral cephalograms were used to quantify skeletal and dental relationships at initial records (T0, before BAMP or growth observation) and at comprehensive initial records (T1) and final comprehensive records (T2). Patients were categorized by their definitive treatment modality: surgical, with 1-jaw and 2-jaw subcategories, or camouflage, with extraction and nonextraction subcategories.

Results:

Data indicate that BAMP+ patients had improvements in skeletal measures, including SNA (P = 0.001), ANB (P < 0.0001), and Wits appraisal (P = 0.0001), compared with controls after phase 1 treatment. Between T1 and T2, changes in ANB (P = 0.019), Wits appraisal (P = 0.001), and B-N Vert (P = 0.02) indicate a greater worsening of Class III malocclusion in the BAMP+ group, suggesting additional growth. At the end of comprehensive treatment, BAMP+ patients had a higher ANB (P = 0.026) and A-N Perp (P = 0.012) than controls, whereas controls had greater maxillary incisor protrusion (P = 0.0072) and proclination (P = 0.047). There was no difference in the prevalence or type of orthognathic surgery. Among patients treated with orthodontic camouflage, controls (55.17%) underwent extractions more often than BAMP+ patients (21.05%, P = 0.003).

Conclusions:

There is modest improvement in maxillary position associated with BAMP, with a limited effect on dentoalveolar outcomes. BAMP did not influence the comprehensive treatment modality (surgical vs camouflage). However, controls with camouflage were over twice as likely to have extractions compared to BAMP+ patients. Our findings suggest that BAMP may reduce extractions, with a limited effect on the frequency of orthognathic surgery.

The treatment of Class III malocclusion presents a significant challenge for orthodontists due to the prolonged nature of mandibular growth. ¹ Traditional treatment modalities emphasize early, phase 1 intervention with reverse pull headgear (RPHG) for orthopedic maxillary protraction, followed by comprehensive treatment with orthodontic camouflage or orthognathic surgery at growth completion. ^1,2 Depending on severity, Class III nonsurgical treatments compromise dentoalveolar health and esthetics with excessive proclination or retroclination, risking fenestration, dehiscence, and gingival recession. ^1,3,4 Despite functional and esthetic needs, many patients with Class III malocclusion do not undergo orthognathic surgery due to its invasiveness, high cost, and associated surgical risks. ^5,6

Bone-anchored maxillary protraction (BAMP) is a well-tolerated, modestly invasive treatment similar to third molar removal, using skeletal anchorage to achieve Class III orthopedic effects during adolescence. ⁷ BAMP is often performed around puberty, before growth cessation. The technique involves surgical placement of titanium miniplates on the infrazygomatic crest and mandibular canine region, with transmucosal hooks to which Class III elastics are attached for intermaxillary traction. ⁸ De Clerck et al ⁹ were the first to investigate the dentofacial effects of BAMP on 21 patients and found a 4 mm mean increase in maxillary advancement, with significant increases in overjet (3.8 mm) and improved molar relationships (4.8 mm) compared with those in untreated controls. Cone-beam computed tomography 3-dimensional imaging studies also demonstrate favorable skeletal changes after BAMP treatment, including anterior displacement of the maxilla, zygoma, and maxillary incisors. ^10,11 In addition, De Clerck et al ¹² found BAMP to have restraining effects on the mandible, with distal displacement of the posterior ramus, gonial angle closure, and glenoid fossa remodeling.

BAMP treatment appears to reduce the severity of Class III skeletal relationships without significant dentoalveolar compensations, potentially reducing the need for, or the magnitude of, subsequent orthognathic surgery after growth. ^7,8,13 Although the findings are promising, patients were only followed until the completion of BAMP treatment, when most subjects had significant growth remaining.^{9, 10,12} Therefore, it is unknown whether these changes were maintained upon completion of treatment, after the pubertal growth spurt. Furthermore, these studies did not address the definitive treatment modalities of the patients. A recent randomized controlled trial addressed some shortcomings by evaluating the skeletal and dental changes with BAMP, as well as the perceived need for orthognathic surgery 3 years after treatment. Mandall et al ¹⁴ found smaller skeletal changes than previous studies, reporting a 0.6° increase in ANB (°) and a 1.4 mm mean increase in overjet in BAMP patients after 18 months from baseline. The perceived surgical need for BAMP patients was 48%, compared with 75% for controls. ¹⁴ However, the assessment of surgical needs was determined at 15 years of age, which is somewhat premature, as mandibular growth can continue into the late teens and early 20s. ^15,16

Ultimately, patients and their families pursue and finance BAMP in the hope of avoiding orthognathic jaw surgery. However, there is a gap in knowledge as to whether BAMP treatment significantly influences comprehensive treatment modality and final treatment outcomes at growth cessation. The purpose of this study had 2 parts: (1) to evaluate the short- and long-term effects of BAMP on objective measures of jaw and tooth position for patients with Class III malocclusion relative to controls and (2) to compare the prevalence of different comprehensive treatment modalities in patients with Class III malocclusion who received BAMP and those who did not (controls). We hypothesized that BAMP treatment is associated with improved jaw and tooth positions at treatment completion and a reduction in the frequency of orthognathic surgery, relative to controls.

MATERIAL AND METHODS

This longitudinal, retrospective study assesses the short- and long-term effects of BAMP appliances on skeletal and dental relationships and final treatment outcomes in patients with Class III malocclusion. This study was approved by the University of North Carolina (UNC) Institutional Review Board (No. 23–0233). Subjects were identified from a database of patients who were treated between 2002 and 2024 by 1 of 7 UNC-trained oral and maxillofacial surgeons (OMS) or by an orthodontist at UNC or 1 of 5 nearby private practices. Orthodontic charts and records were screened by 2 calibrated examiners (C.E., L.J.). Patients included had skeletal Class III malocclusion, were aged 9–15 years, and received 4 BAMP plates (BAMP+). All patients were managed jointly by their OMS and orthodontist. The standardized protocol, as specified by the OMS, involved incremental loading with 100–250 g of force per side via full-time bilateral elastic traction, except when eating and brushing (Fig 1). After 1–2 months, patients were evaluated to have their elastic size modified based on clinical change, plate stability, and jaw asymmetry. Full-time elastic wear was typically continued until positive overjet was achieved. After full-time elastic wear for retention, BAMP+ patients were asked by their OMS to wear elastics only at night until plate removal. In most patients, plate removal occurred concurrently with third molar extractions to minimize surgical procedures. Control patients with Class III malocclusion (BAMP−) were selected from a pool of patients with skeletal Class III malocclusion who were qualified for or were referred for BAMP but declined treatment, with growth observation before comprehensive orthodontics (Fig 2; Table I). A minority of controls (n = 2) and BAMP+ patients (n = 9) received additional appliances (concurrent with observation or BAMP+) when indicated for space maintenance (eg, Nance holding arch) or transverse discrepancies (ie, expansion), which would have minimal to no impact on sagittal jaw position; expansion and holding the transverse dimension were performed over a 5–6 month period. All patients must have subsequently completed comprehensive phase 2 treatment with complete orthodontic records. Neither BAMP nor control patients were excluded for receiving phase 1 treatment before data collection for this study began at T0. Enrollment criteria are shown in Table I. Phase 2 mechanics (between T1-T2) included fixed appliances or clear aligners and intraoral, interarch elastics as prescribed by the orthodontist; space closure was reciprocal without additional temporary skeletal anchorage devices.

Fig 1.

A, Schematic diagram of bone anchors in the infrazygomatic and mandibular canine regions and Class III elastic positioning. B, Elastic traction (100–200 g per side; 200–400 g of total force) is applied bilaterally, 24 hours a day until positive overjet is achieved. C, Frontal intraoral photograph showing bone anchor and elastics positions at the beginning of BAMP treatment with full-time elastic wear. This patient presented with edge-to-edge incisors. D, Frontal intraoral photo showing positive overjet after treatment.

Fig 2.

A and B, Study Flowchart and timeline. Patients identified from a database at UNC and treated at UNC Orthodontics or 5 nearby private practices. The initial sample of 245 BAMP+ and 1017 control patients was reduced to a final sample size of 45 BAMP+ and 39 BAMP− after screening for inclusion/exclusion criteria. Cephs were traced and analyzed at initial records before BAMP treatment or growth observation (T0), before comprehensive treatment (T1), and at the end of comprehensive treatment (T2). For patients whose comprehensive treatment included orthognathic surgery, presurgical ceph was used in place of T2 ceph to accurately represent jaw and tooth positions before surgery.

Table I.

Enrollment criteria

	Inclusion criteria	Exclusion criteria
BAMP+	• Skeletal and dental Class III (Wits appraisal <0 mm) • BAMP plates placed and followed by UNC oral surgeon • Age 10–15 y at the start of BAMP • Completion of BAMP treatment with pretreatment lateral ceph • Completion of phase 2 comprehensive treatment with posttreatment lateral ceph	• History of cleft lip or palate or craniofacial syndromes • Prior jaw surgery or significant craniofacial trauma • Reverse pull headgear (RPHG) in use at time of T0 • Incomplete records –missing T0 or T2 ceph • Did not receive phase 2 orthodontic treatment • Total observation period ≤2.3 y • Severe asymmetry or open bite of ≥2.6 mm at T0
Control	• Skeletal and dental Class III (Wits appraisal <0 mm) • Declined phase 1 BAMP treatment or placed on growth observation before comprehensive treatment, with pretreatment lateral ceph • Age 10–15 y at the start of growth observation • Completion of phase 2 treatment with pre- and posttreatment lateral ceph	• Same as above

Lateral cephalograms (cephs) were evaluated at the following timepoints: initial records at the time of BAMP recommendation (T0), comprehensive initial records (T1), and final records at the end of comprehensive treatment (T2). At T0, all patients were required to have completed prior phase 1 RPHG treatment, if applicable. For patients whose comprehensive treatment included orthognathic surgery, their presurgical ceph was used in place of their T2 ceph to accurately represent jaw and tooth positions before surgery. For these patients, the postsurgical timepoint was defined as T3. The initial sample of 245 BAMP+ patients and 1017 control patients was reduced to a final sample of 45 BAMP+ and 39 BAMP− patients after screening for enrollment criteria (Fig 2; Table I).

Lateral cephs taken at T0, T1, and T2 were traced by a single examiner using Dolphin Imaging software (Dolphin Imaging and Management Solutions, Chatsworth, Calif). Cephalometric measures are shown in Figure 3 and include angular measurements (°): SNA, SNB, ANB, U1-SN, U1-NA, U1-FH, IMPA, L1-NB, FMIA, FMA, and SN-GoGn and linear measurements (in millimeters): A-N Vert, B-N Vert, Pog-N Vert, A-N Perp, B-N Perp, Pog-N Perp, Co-ANS, Co-Gn, U1-NA, L1-NB, Wits appraisal, Overjet, and Overbite.

Fig 3.

A, Cephalometric analysis performed using the following landmarks: (S) Sella; (N) Nasion; (A) A point; (B) B point; (Co) Condylion; (Go) Gonion; (Gn) Gnathion; (ANS) anterior nasal spine; (Pog) Pogonion; (P) Porion; (Or) Orbitale; planes: (FH) Frankfort horizontal (P-Or); mandibular (Go-Gn); occlusal. IMPA is defined as L1-mandibular plane, FMIA is defined as L1-Frankfort horizontal plane. Long axis of U1 and L1 relate the inclination and millimeter distance of the maxillary and mandibular incisors to the NA and NB lines using the Steiner analysis. B, Serial superimposition of a representative BAMP treatment outcome. Black line, pretreatment (T0); red line, posttreatment (T2).

For patients treated with BAMP concurrently with comprehensive treatment, only T0 and T2 records were analyzed. After initial tracing, 20% of the subjects in each group were selected by a random number generator for reanalysis for a concordance correlation reliability test (Supplementary Table I).

The primary outcomes of interest were skeletal and dentoalveolar changes induced by BAMP, as indicated by cephalometric measures from lateral tracings at each timepoint. Secondary outcomes included definitive comprehensive treatment modalities in BAMP+ vs controls. Patients were categorized by their definitive comprehensive treatment modality as receiving surgical treatment or nonsurgical camouflage treatment. For patients receiving surgical treatment, they were further categorized based on surgery type (1-jaw vs 2-jaw). Patients treated nonsurgically were categorized by whether camouflage was with or without extractions. The total observation period was defined as the time between initial records (T0) and final records after comprehensive treatment (T2). Total comprehensive treatment time was defined as the period from the date of bonding for comprehensive treatment to the date of debonding. If exact dates were not identified during the chart review, T1 and T2 timepoints of initial and final records were used as a proxy.

Statistical analysis

The primary exposure of interest was BAMP, dichotomized as with or without BAMP treatment. Descriptive statistics for demographic variables at T0 for the BAMP+ and control groups are shown in Table II.

Table II.

Participant demographics

	BAMP+ (n = 45)	Control (n = 39)
	n (%) or mean ± SD	n (%) or mean ± SD	P value
Sex			0.274
Male	20 (44.44)	22 (56.41)
Female	25 (55.56)	17 (43.59)
Race			0.431
White	34 (75.56)	33 (84.62)
Black	4 (8.89)	4 (10.26)
Asian	6 (13.33)	2 (5.13)
Other	1 (2.22)	0 (0.00)
Age
T0	11.96 ± 1.07	12.60 ± 1.16	0.006*
T1	14.86 ± 1.71	15.69 ± 1.85	0.031*
T2	17.16 ± 1.66	18.03 ± 1.95	0.036*
CVM
T0	2.70 ± 1.15	2.79 ± 1.36	0.744
T1	4.38 ± 0.90	4.49 ± 0.73	0.554
T2	5.2 ± 0.59	5.08 ± 0.69	0.414
Phase 1 RPHG Treatment before BAMP or control growth observation	12 (26.67)	5 (12.82)	0.115

Note. P values were calculated using chi-square tests to compare frequencies of race, sex, and the use of RPHG in the treatment and control groups. The other category includes patients of unknown race. P values were calculated using the Wilcoxon 2-sample rank sum test (Mann-Whitney U test) to compare age at each time point in the treatment and control groups. P values for mean CVM stage were calculated using independent t tests. The significance threshold was set at P <0.05.

SD, standard deviation.

^*Indicates a significant difference between the treatment (BAMP+) and control groups.

General linear models, adjusted for gender, race, age, and cervical vertebral maturation (CVM), Wits appraisal, and ANB at T0, were used to test for significant differences in growth between groups (with and without BAMP) from pre- to posttreatment records. Based on a linear mixed-effects model, changes between timepoints within each group were calculated using pairwise t tests to compare the least squares means; the changes in each outcome between timepoints (T0, T1, and T2) were tested with the application of a contrasts matrix of group by time with the F-test. Differences between groups at each timepoint were calculated and compared using t tests based on the linear regression model.

Outcomes for comprehensive treatment modalities were assessed dichotomously based on whether patients had orthognathic surgery or not in each group. Patient groups were also compared categorically according to whether they had orthognathic surgery in one or both jaws, or had nonsurgical treatment with or without extractions. The chi-square test or Mantel-Haenszel chisquare exact test was used to evaluate whether treatment groups differed in categorical variables. The significance threshold was P < 0.05. A clinically significant difference was defined a priori as a difference of 2 mm or greater in mean linear measurement and 2° or more for angular measures. ^17,18 To evaluate the reliability of tracing, intrarater concordance correlation coefficients were calculated (Supplementary Table I).

Population mean and variance for cephalometric values were derived from De Clerck et al.⁹ Using these data, we calculated that 8–31 participants per group were needed for 90% power, depending on the cephalometric measure used. The cephalometric values included in these calculations were: A-VertT (sample size: 12, effect size: 2.31), B-VertT (sample size: 31, effect size: −1.53), Overjet (sample size: 8, effect size: 7.62), and Wits appraisal (sample size: 8, effect size: 3.37). To ensure adequate power with our longer-term timepoints, the maximum number of patients with records that met the inclusion criteria were included in this analysis. With 39 and 45 people per group, respectively, this study was adequately powered to detect differences.

RESULTS

The study cohort consisted of 45 BAMP+ patients and 39 controls. The treatment and control groups were well matched demographically (Table II). At baseline, there were significant differences in A-N Perp, B-N Perp, and Pog-N Perp, although the samples were otherwise well matched (Table III, Supplementary Table II). Multiple skeletal variables were intentionally included for redundancy, and similar measures were not significantly different. There were significant differences in age between the BAMP+ and control groups at each timepoint, with the BAMP+ group being 8–11 months younger. However, the groups were similar in skeletal age, as there were no differences in CVM stage at any timepoint, preand postsurgically (Tables II, III, and IV; Supplementary Tables III and IV). Intrarater concordance correlation coefficients indicated high repeatability in cephalometric tracings (Supplementary Table I).

Table III.

Comparison of cephalometric measures and CVM staging at T0

	BAMP (n = 45)	Control (n = 39)
Ceph measures	Mean ± SD	Mean ± SD	P value
SNA (°)	81.20 ± 3.76	80.57 ± 4.27	0.668
SNB (°)	82.24 ± 3.25	81.94 ± 3.90	0.692
ANB (°)	−1.05 ± 2.15	−1.38 ± 2.13	0.804
A-N Vert (mm)	−2.18 ± 5.50	−2.76 ± 4.45	0.945
B-N Vert (mm)	−2.16 ± 7.76	−2.20 ± 6.67	0.911
Pog-N Vert (mm)	−1.98 ± 8.99	−1.63 ± 7.01	0.913
A-N Perp (mm)	0.13 ± 3.87	−1.19 ± 3.49	0.028*
B-N Perp (mm)	1.96 ± 5.31	0.33 ± 5.48	0.032*
Pog-N Perp (mm)	2.82 ± 5.97	1.35 ± 6.22	0.046*
Co-ANS (mm)	88.09 ± 5.34	88.50 ± 4.95	0.883
Co-Gn (mm)	122.88 ± 7.72	124.85 ± 8.50	0.754
U1-SN (°)	107.82 ± 6.43	107.00 ± 8.74	0.598
U1-NA (°)	26.62 ± 6.67	26.43 ± 7.10	0.683
U1-NA (mm)	5.36 ± 2.96	5.32 ± 2.29	0.659
U1-FH (°)	116.73 ± 6.42	115.31 ± 7.74	0.221
IMPA (°)	82.59 ± 5.71	81.37 ± 6.77	0.951
L1-NB (°)	19.64 ± 6.68	18.73 ± 8.25	0.783
L1-NB (mm)	3.98 ± 2.60	3.53 ± 2.96	0.865
Wits appraisal (mm)	−6.20 ± 2.88	−6.33 ± 2.73	0.804
Overjet (mm)	0.04 ± 2.22	0.03 ± 2.45	0.750
FMIA (°)	71.52 ± 6.59	71.53 ± 8.47	0.270
FMA (°)	25.89 ± 3.72	27.11 ± 5.25	0.055
SN-GoGn (°)	32.33 ± 4.95	33.08 ± 5.47	0.359
OB (mm)	0.70 ± 1.76	0.98 ± 1.56	0.405
CVM stages, n (%)			0.793
Stage 1	7 (15.91)	9 (23.08)
Stage 2	14 (31.82)	8 (20.51)
Stage 3	10 (22.73)	9 (23.08)
Stage 4	11 (25)	8 (20.51)
Stage 5	2 (4.55)	5 (12.82)
Stage 6	0 (0.00)	0 (0.00)

Note. P values were calculated using a linear regression model t test to compare T0 cephalometric variables between the treatment and control groups. The Mantel-Haenszel chi-square test was used to compare frequencies of CVM stage between the 2 groups. The significance threshold was set at P <0.05.

SD, standard deviation.

^*Indicates a significant difference between the treatment (BAMP+) and control group at T0.

Table IV.

Comparison of cephalometric measures and CVM staging at T1 and T2 timepoints

	T1			T2
	BAMP (n = 45)	Control (n = 39)		BAMP (n = 45)	Control (n = 39)
Ceph measures	Mean ± SD	Mean ± SD	P value	Mean ± SD	Mean ± SD	P value
SNA (°)	82.80 ± 4.02	80.16 ± 4.01	0.012*	83.0 ± 4.11	81.09 ± 4.72	0.071
SNB (°)	83.13 ± 3.26	82.79 ± 4.04	0.647	83.64 ± 4.01	83.08 ± 4.65	0.485
ANB (°)	−0.34 ± 3.01	−2.61 ± 2.23	<0.001*	−0.66 ± 2.62	−2.00 ± 2.52	0.026*
A-N Vert (mm)	−2.21 ± 6.10	−2.72 ± 4.50	0.902	−0.20 ± 6.49	−2.29 ± 4.51	0.053
B-N Vert (mm)	−3.07 ± 8.94	−0.21 ± 7.16	0.126	0.82 ± 10.62	−0.31 ± 7.03	0.280
Pog-N Vert (mm)	−2.50 ± 10.29	1.29 ± 8.66	0.145	2.48 ± 11.92	1.76 ± 8.50	0.308
A-N Perp (mm)	1.38 ± 4.41	−1.65 ± 3.54	0.001*	1.97 ± 4.78	−0.45 ± 3.60	0.012*
B-N Perp (mm)	3.01 ± 6.09	1.84 ± 6.16	0.215	4.47 ± 7.82	2.71 ± 6.20	0.195
Pog-N Perp (mm)	4.44 ± 7.22	3.64 ± 7.33	0.261	6.7 ± 8.75	5.29 ± 7.17	0.251
Co-ANS (mm)	92.49 ± 7.41	91.16 ± 6.82	0.183	92.92 ± 7.52	92.44 ± 7.27	0.392
Co-Gn (mm)	129.65 ± 10.13	133.12 ± 9.70	0.304	132.40 ± 12.57	134.92 ± 12.20	0.739
U1-SN (°)	109.82 ± 6.29	109.47 ± 7.98	0.629	113.40 ± 7.87	114.01 ± 8.36	0.379
U1-NA (°)	27.03 ± 6.85	29.32 ± 6.26	0.258	30.41 ± 6.95	32.92 ± 7.26	0.047*
U1-NA (mm)	5.93 ± 2.89	7.11 ± 2.42	0.142	6.42 ± 2.73	7.70 ± 2.47	0.007*
U1-FH (°)	118.36 ± 6.40	117.77 ± 7.09	0.404	122.22 ± 7.89	122.59 ± 7.74	0.453
IMPA (°)	84.14 ± 7.16	81.25 ± 6.97	0.195	84.27 ± 7.94	82.05 ± 5.80	0.630
L1-NB (°)	20.85 ± 7.70	18.35 ± 8.51	0.484	20.84 ± 8.36	19.52 ± 7.28	0.746
L1-NB (mm)	4.64 ± 3.03	3.53 ± 3.16	0.317	4.15 ± 3.07	3.73 ± 2.45	0.626
Wits appraisal (mm)	−4.88 ± 3.96	−7.42 ± 3.43	<0.001*	−5.56 ± 3.87	−6.31 ± 4.18	0.536
Overjet (mm)	0.83 ± 2.55	−0.12 ± 2.70	0.043*	1.41 ± 2.98	1.16 ± 3.39	0.772
FMIA (°)	70.82 ± 7.85	72.73 ± 9.24	0.829	71.63 ± 9.25	72.15 ± 8.21	0.465
FMA (°)	25.05 ± 4.02	26.02 ± 5.66	0.067	24.09 ± 4.67	25.81 ± 6.26	0.067
SN-GoGn (°)	30.93 ± 4.85	31.67 ± 6.24	0.260	30.40 ± 5.53	31.94 ± 7.22	0.115
OB (mm)	0.56 ± 1.71	0.62 ± 2.21	0.935	0.94 ± 1.21	0.86 ± 2.19	0.678
CVM stages, n (%)
Stage 1	0 (0.00)	0 (0.00)	0.581	0 (0.00%)		0.482
Stage 2	1 (2.50)	0 (0.00)		0 (0.00%)
Stage 3	5 (12.50)	2 (5.41)		0 (0.00%)
Stage 4	15 (37.50)	18 (48.65)		4 (8.89%)
Stage 5	16 (40)	14 (37.84)		28 (62.22%)
Stage 6	3 (7.50)	3 (8.11)		13 (28.89%)

Note. P values were calculated using a linear regression model t test to compare cephalometric variables at T1 and T2 between the treatment and control groups. The Mantel-Haenszel chi-square test was used to compare frequencies of CVM stage between the 2 groups. The significance threshold was set at P < 0.05.

SD, standard deviation.

^*Indicates a significant difference between the treatment (BAMP+) and control group at T1 or T2.

Significant treatment changes occurred between T0-T1 for the BAMP+ cohort. SNA and A-N Perp increased 1.8°/1.5 mm more in the treatment group, indicating a more forward position of the maxilla at T1 compared with controls (Fig 4; Table V). BAMP treatment did not appear to affect or restrict mandibular growth, as indicated by significant increases in SNB in both the groups. There were no significant differences in mean change of mandibular skeletal measures including SNB, B-N Vert, and B-N Perp (Table V; Supplementary Table II). Consistent with these findings, ANB demonstrated an improvement toward Class I in the BAMP+ group and worsening toward Class III in the control group, resulting in a statistically significant change of 1.8° (Table V). A similar pattern was observed for Wits appraisal in each group, with an improvement in the BAMP+ group toward Class I (+1.21 mm) and worsening in the control group toward Class III (−1.05 mm). Overjet also increased by 0.93 mm in the treatment group, a significant change when compared with controls. Notably, the change in incisor position (U1-NA) was significantly different between the groups, with a greater mean change in controls (+1.74 mm) compared with BAMP patients (+0.58 mm) (Table V). There were no other significant changes between groups from T0-T1 for dentoalveolar measures.

Fig 4.

A-H, Cephalometric changes over time (T0-T2) in the BAMP+ (blue) and control (orange) groups. *Indicates a statistically significant difference between 2 time points within groups. ^Indicates a statistically significant difference between the treatment (BAMP+) and control group at a given time point. I, Frequency of final treatment approaches (surgical, camouflage extraction, and camouflage nonextraction) in comprehensive treatment. J, Approach details for surgical intervention and extraction patterns. P <0.05 significance threshold.

Table V.

Change in cephalometric measures from T0-T1 and T1-T2 between BAMP and control groups

	T0-T1				T1-T2
	BAMP (n = 45)	Control (n = 39)			BAMP (n = 45)	Control (n = 39)
Ceph measures	Mean ± SD	Mean ± SD	Difference	P value	Mean ± SD	Mean ± SD	Difference	P value
SNA (°)	1.65 ± 1.83	−0.15 ± 2.39	1.75	0.001*	0.24 ± 1.70	0.68 ± 2.14	−0.48	0.304
SNB (°)	0.94 ± 1.78	1.00 ± 2.18	−0.07	0.887	0.59 ± 2.06	0.22 ± 2.09	0.34	0.493
ANB (°)	0.72 ± 1.75	−1.13 ± 1.28	1.80	<0.001*	−0.35 ± 1.36	0.44 ± 1.31	−0.80	0.020*
A-N Vert (mm)	0.13 ± 5.27	0.10 ± 5.13	0.14	0.919	1.79 ± 5.76	0.21 ± 5.35	2.38	0.090
B-N Vert (mm)	−0.65 ± 8.42	1.95 ± 8.37	−2.35	0.270	3.59 ± 9.82	−0.17 ± 6.91	5.09	0.020*
Pog-N Vert (mm)	−0.19 ± 9.82	2.89 ± 9.77	−2.74	0.268	4.59 ± 11.29	0.36 ± 8.00	5.77	0.022*
A-N Perp (mm)	1.34 ± 2.41	−0.39 ± 2.56	1.52	0.010*	0.42 ± 2.09	1.07 ± 2.63	−0.89	0.122
B-N Perp (mm)	1.18 ± 4.02	1.48 ± 4.29	−0.50	0.594	1.22 ± 3.99	0.96 ± 4.90	−0.09	0.932
Pog-N Perp (mm)	1.79 ± 4.92	2.27 ± 5.18	−0.60	0.591	1.93 ± 4.73	1.73 ± 5.59	−0.20	0.877
Co-ANS (mm)	4.30 ± 5.50	2.78 ± 5.26	1.62	0.200	0.22 ± 4.69	1.25 ± 5.75	−1.27	0.301
Co-Gn (mm)	6.79 ± 6.21	8.48 ± 7.80	−1.95	0.243	2.71 ± 6.44	1.57 ± 8.26	1.02	0.553
U1-SN (°)	2.18 ± 5.92	2.31 ± 5.54	0.18	0.893	3.04 ± 9.10	4.83 ± 6.73	−2.61	0.176
U1-NA (°)	0.54 ± 6.36	2.46 ± 5.50	−1.55	0.255	2.80 ± 8.81	4.15 ± 6.72	−2.13	0.262
U1-NA (mm)	0.58 ± 2.15	1.74 ± 2.34	−1.07	0.044*	0.36 ± 2.55	0.73 ± 2.41	−0.83	0.179
U1-FH (°)	1.83 ± 5.80	2.11 ± 5.66	−0.07	0.955	3.17 ± 8.62	5.25 ± 6.93	−3.09	0.111
IMPA (°)	1.55 ± 4.72	0.00 ± 4.35	2.16	0.052	−0.02 ± 6.29	0.45 ± 5.92	−1.20	0.426
L1-NB (°)	1.16 ± 4.84	−0.17 ± 4.42	1.57	0.163	−0.23 ± 6.06	0.86 ± 6.28	−1.70	0.270
L1-NB (mm)	0.54 ± 1.61	0.03 ± 1.75	0.52	0.207	−0.56 ± 1.87	0.13 ± 2.26	−0.94	0.056
Wits appraisal (mm)	1.21 ± 2.53	−1.05 ± 2.27	2.40	<0.001*	−0.65 ± 2.03	1.05 ± 2.54	−1.93	0.001*
Overjet (mm)	0.93 ± 1.85	−0.09 ± 1.46	1.00	0.017*	0.48 ± 2.31	1.23 ± 3.19	−0.98	0.158
FMIA (°)	−0.58 ± 6.15	0.96 ± 5.13	−1.89	0.159	0.90 ± 6.52	−0.22 ± 6.35	1.55	0.338
FMA (°)	−0.96 ± 3.01	−0.96 ± 2.72	−0.25	0.689	−0.90 ± 2.54	−0.24 ± 2.52	−0.36	0.569
SN-GoGn (°)	−1.48 ± 2.77	−1.41 ± 2.50	−0.38	0.556	−0.70 ± 2.46	0.37 ± 2.81	−0.98	0.115
OB (mm)	−0.11 ± 2.12	−0.30 ± 1.48	0.28	0.521	0.50 ± 1.77	0.24 ± 2.36	0.17	0.751

Note. P values were calculated using F-tests to compare change over time between groups based on a linear mixed model. The significance threshold was set at P < 0.05.

SD, standard deviation.

^*Indicates a significant difference between the mean change of the treatment (BAMP+) and control groups for a given time interval.

At T1, there were significant differences between the groups in SNA, ANB, A-N Perp, Wits appraisal, and overjet, with values indicating a less severe Class III malocclusion in the BAMP+ cohort (Table IV).

When comparing changes from T1-T2 between the BAMP+ and control groups, significant differences were noted for ANB angle, B-N Vert, Pog-N Vert, and Wits appraisal. All measures indicated a greater worsening of the Class III malocclusion in the BAMP+ group (Fig 4; Tables IV and V). Of these variables, only the changes in B-N Vert and Pog-N Vert were found to be clinically significant. Nonetheless, BAMP patients finished orthodontic treatment at T2 with an ANB 1.3 ° greater than that of controls and less severe A-N Perp values (Table IV). Changes between T1-T2 in overjet, incisor angulation, and position were not significantly different between groups. However, at T2, controls had significantly greater U1-NA (mm) and U1-NA (°) values, indicating greater compensation of the maxillary incisors (Table IV).

During the whole study period (T0-T2), significant positive treatment effects of BAMP were observed for SNA and ANB when comparing changes between groups (Supplementary Table V). The BAMP+ group had approximately a 1° greater improvement in SNA and ANB than the control group. There was also a significant change in maxillary incisor position, measured at U1-NA.

The maxillary incisors were advanced by 1.73 mm more in the controls than in BAMP+ patients. All of these changes initially occurred over the BAMP treatment phase (T0-T1). In some instances, the magnitude of change was slightly reduced during comprehensive treatment.

Among BAMP+ and control participants who underwent orthognathic surgery, there were no significant differences in presurgical cephalometric measures, though significant improvements in ANB, Wits appraisal, and overjet were seen post-operatively, as expected (Supplementary Tables VI and VII). Because presurgical decompensation can affect cephalometric measures, a secondary analysis was performed with the removal of surgical participants at T2 (Supplementary Tables VIII-X). Results still demonstrated a significant difference between the BAMP+ and control groups at T2 for ANB and U1-NA (mm), but no significant difference was found for A-N Perp nor U1-NA (°). This is explained by the removal of control surgical participants, who had the most severe maxillary deficiency in the study (with an average A-N Perp = −2.45 mm), compared with the surgical BAMP group (with an average A-N Perp = 2.70 mm) and both camouflage groups (BAMP, A-N Perp = 1.83 mm; control, A-N Perp = 0.24 mm). U1-NA (°) average values increased slightly with the removal of the surgical participants, as expected, and approached significance. Changes between T1-T2 were significant between groups for ANB, A-N Perp, and Wits appraisal but not for B-N Vert or Pog-N Vert. Changes over the T0-T2 period remained significant with the removal of the surgical participants, with significant BAMP+ treatment effects observed for SNA, ANB, and U1-NA (mm).

Definitive treatment modalities differed between the BAMP+ and control groups for nonsurgical treatment only, with 55.17% of the control group receiving camouflage treatment with extractions, compared with 21.05% in the BAMP + group (Fig 4, I and J; Table VI). There was no significant difference in extraction patterns between the groups. Of the BAMP+ group, 15.56% were treated with orthognathic surgery, compared with 25.54% of the controls (P = 0.2512). Moreover, there was a greater number of 2-jaw surgeries in the BAMP+ group, but the difference was not significant. The mean observation time from initial records to the end of treatment was approximately 5 years in both groups. The total comprehensive treatment time was also similar between the groups, averaging 2.5 years.

Table VI.

Definitive treatment approach for BAMP and control groups

	BAMP+ n (%)	Control n (%)	P value
Surgery	7 (15.56)	10 (25.54)	0.251
1-Jaw	4 (57.14)	9 (90.00)	0.250
2-Jaw	3 (42.86)	1 (10.00)
Camouflage	38 (84.44)	29 (74.36)
Extractions	8 (21.05)	16 (55.17)	0.004*
4 Bilateral extractions	3 (37.50)	6 (37.50)	0.838
2 Lower premolar extractions	3 (37.50)	5 (31.25)
1 Lower premolar extraction	1 (12.50)	3 (18.75)
1 Lower incisor extraction	1 (12.50)	1 (6.25)
2 Lower premolar and 1 lower incisor extractions	0 (0.00)	1 (6.25)
Nonextraction	30 (78.95)	13 (44.83)
Total observation period (y), mean ± SD	5.20 ± 1.33	5.41 ± 1.71	0.540
Time in phase 2 treatment (y), mean ± SD	2.50 ± 1.04	2.41 ± 0.94	0.720

Note. P values were calculated using chi-square tests to compare frequencies of surgery, camouflage, and extraction treatment modalities in the BAMP+ and control groups. P values to compare the type of surgery and extraction pattern were calculated using Mantel-Haenszel chi-square tests. The significance threshold was set at P < 0.05.

SD, standard deviation.

^*Indicates a significant difference between the treatment (BAMP+) and control groups.

DISCUSSION

At the end of treatment, BAMP+ patients showed significantly greater ANB and A-N Perp values compared with controls, indicating a less severe Class III skeletal relationship. Although BAMP patients started treatment with higher mean A-N Perp values, its increase over treatment was significantly greater than that in controls and was maintained through the end of treatment. There were no significant dentoalveolar effects in the BAMP+ group, but controls had greater incisor protrusion, measured at U1-NA (mm), and increased proclination at U1-NA (°) at the end of treatment.

Removal of surgical participants from both groups yielded no significant differences between groups for A-N Perp and U1-NA (°) at T2 and for B-N Vert and Pog-N Vert over T1-T2. These differences can be explained by the removal of participants who were likely the patients with the most severe Class III malocclusion, with lower A-N Perp and higher B-N Vert and Pog-N Vert values. U1-NA (°) average values increased slightly, as expected, and approached significance. The data suggest that patients treated with orthognathic surgery still have some Class III maxillary incisor compensation, as both BAMP+ and control surgical patients had greater than normative incisor proclination, as indicated by U1-SN and U1-NA (°). The loss of significance in B-N Vert and Pog-N Vert between T1-T2 may be due to slightly greater vertical growth in the control surgical participants, resulting in a net reduction in mandibular sagittal projection. Most notably, changes over the T0-T2 period remained significant with the removal of the surgical participants, with significant BAMP treatment effects observed for SNA, ANB, and U1-NA (mm).

There were no differences in surgical prevalence or approach between the 2 groups, but when treated with orthodontic camouflage, twice as many extractions were performed in the control group.

In the literature, maxillary protraction using miniplate anchorage produces variable results. A systematic review reported a large range of 1.61–8 mm advancement at A point across studies with BAMP treatment.¹⁹ De Clerck et al⁹ found 4 mm of advancement using a similar BAMP protocol as used here. Although different cephalometric variables were used, De Clerck’s measurements indicate a larger magnitude of change than our data suggest, especially when considering the 6.7 mm improvement in Wits appraisal and 3.8 mm improvement in overjet seen in the De Clerck BAMP group compared to controls. However, this data was collected from prepubertal patients, most of whom were still CVM 1–3 at the end of phase 1 BAMP treatment and had considerable mandibular growth remaining. ⁹

Our results align more closely with those of Mandall et al¹⁴ who found, in a randomized controlled trial with 57 patients, “a small but favorable improvement in ANB and overjet” in Class III children treated with BAMP. They reported an average increase in SNA of 1.1° with BAMP, which, when compared to controls, resulted in a 0.7° and 0.6° greater improvement in SNA and ANB, respectively.¹⁴ In their study, skeletal effects were maintained over the observation period; this differs from our findings, as continued mandibular growth counteracted some beneficial effects of BAMP on ANB and Wits appraisal seen at T1 (Fig 4; Tables IV and V). Despite the apparent lack of mandibular growth restriction during BAMP treatment, our results indicate that BAMP+ patients demonstrate more marked mandibular growth throughout comprehensive treatment compared to controls. This may be due to the cessation of full-time elastic wear on bone anchors resulting from either provider instruction or patient burnout. Regardless, the net effect, a mean decrease of −0.8° in ANB in BAMP+ patients, is clinically undetectable and likely had no impact on treatment outcomes. The improvement in Wits appraisal observed in the control group during the comprehensive phase may be the result of dentoalveolar compensations. Intermaxillary elastic forces applied to the teeth (as opposed to bone anchors) may have contributed to a counter-clockwise rotation of the occlusal plane, which can make Wits appraisal appear less Class III.¹

A notable finding was the limited benefit BAMP had in terms of reducing Class III dentoalveolar compensations, specifically mandibular incisor angulations (IMPA, L1-NB, and FMIA), relative to controls. This is surprising because BAMP is thought to be advantageous for its lack of compensating effects on the dentition; however, similar compensations are seen in the mandibular incisors at the end of phase 1 and phase 2 comprehensive treatment between groups. De Clerck et al⁹ and Cevidanes et al¹³ reported a significant improvement in mandibular incisor retroclination after BAMP treatment, but not a reduction in maxillary incisor proclination. Conversely, our results show a significant mean change (+1.7 mm) in maxillary incisor position of the control group at the end of phase 1. It is possible that this increased protrusion is the result of phase 1 treatment without skeletal anchorage to correct an anterior crossbite, which was maintained through comprehensive treatment. There were also small but clinically significant differences in U1-NA (°) at T2, indicating greater maxillary incisor proclination in the control group at the end of treatment.

It is worth noting that the differences between the BAMP+ and control groups for SNA, ANB, and Wits appraisal at T1 and ANB and A-N Perp at T2, although small, were clinically significant. Our data suggest that even modest improvements in maxillary position with BAMP+ can still affect final camouflage treatment outcomes with fewer extractions.

To our knowledge, our study is the first of its kind to examine definitive treatment outcomes in patients treated with BAMP. There are limited data evaluating the effects of BAMP beyond1 year. Mandall et al ¹⁴ report the perceived rate of orthognathic surgery at 3-year follow-up to be 48% in the BAMP group and 75% in controls, which was much greater than the rate of surgery found in our population. This contrast may be a product of the United Kingdom’s nationalized healthcare system, compared to the largely privatized system in the United States and associated out-of-pocket costs. Criteria for surgical intervention are also multifactorial and can vary depending on patient desires, background, and provider experience and treatment philosophy. At UNC's clinics, we operate on patients diagnosed with severe Class III dental and skeletal conditions (maxillary deficiency or mandibular prognathism) by board-certified OMS and orthodontic providers based on facial profile, occlusal, and cephalometric evaluations (negative Wits appraisal, ANB, and overjet). As this study was retrospective in nature, it was not possible to precisely determine the number of patients who had surgery recommended but declined it due to cost or risks. It is highly likely that our prevalence of surgery is lower than the number of patients who were offered surgery as an ideal treatment. It is also possible that the control group was less interested in more invasive and costly treatment options, since many had already declined BAMP as a phase 1 option.

Our study followed patients until CVM stage 5, when most of craniofacial growth is complete, with limited late mandibular growth remaining. ²⁰ A prospective study that follows patients with Class III malocclusion for over 10 years is an ideal next step, though logistically challenging. A limitation of this study was the inability to control for compliance-related issues. BAMP effects may be compromised by limited patient compliance or complications with plates affecting elastic wear, such as mobility or inflammation. There were 5 patients who had BAMP plates removed and replaced during treatment, though a detailed record of more minor complications (such as inflammation around hooks) was not available for all subjects. In addition, several patients were noted to have been debonded early due to noncompliance or hygiene-related issues. There was also some variability in elastic size and force throughout BAMP treatment and during BAMP retention. Despite being adequately powered, we could not control for all potential confounding variables, such as the influence of other early interventions (ie, early expansion or RPHG before T0). It was neither possible, nor ethical, to have a true untreated control group since multiphase treatment is indicated for patients with Class III malocclusion. A future prospective randomized control trial could address this limitation. Future studies should aim to extend these exploratory findings and stratify the treatment group based on the presence and/or rate of expansion as well as the elastic force to bone anchors. Records were collected immediately pre-BAMP and, importantly, after RPHG, if rendered; however, we cannot rule out the possibility that early RPHG may have corrected some skeletal issues and reduced the efficacy of BAMP. Including controls who qualified for BAMP but declined this treatment was the closest approximation in a retrospective, single-center study design.

CONCLUSIONS

1. BAMP treatment resulted in approximately 2° of improvement in maxillary position at SNA compared with controls, with associated changes in ANB and Wits appraisal that suggested an improvement in Class III skeletal relationships toward Class I. By the end of comprehensive treatment, the BAMP group maintained an ANB angle 1.3° greater than controls.

2. BAMP had no effect on mandibular cephalometric measures during phase 1 treatment. Some of the skeletal improvements were diminished by continued mandibular growth during a subsequent comprehensive phase.

3. There were no significant improvements in mandibular incisor compensations induced by BAMP. However, the control group had significantly greater maxillary incisor protrusion at U1-NA (mm) and proclination at U1-NA (°) at the end of the treatment.

4. Our findings suggest that BAMP treatment in Class III adolescents can produce modest improvements in maxillary position. These changes have a limited effect on comprehensive treatment modality (surgical vs camouflage). However, camouflage treatment was more than twice as likely to involve extractions without BAMP intervention.

5. Longitudinal randomized controlled trials are needed to better understand the skeletal and dentoalveolar effects of BAMP and its influence on final treatment modality.

Supplementary Material

SUPPLEMENTARY DATA

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.ajodo.2026.04.018.