Association between condylar positional changes and postsurgical mandibular relapse following surgery-first approach in skeletal class III patients

Abstract

Objective

This study aimed to investigate three-dimensional (3D) positional changes of the mandibular condyles following bimaxillary orthognathic surgery performed using the surgery-first approach (SFA) in skeletal Class III patients. The associations between condylar positional changes, other clinical variables, and postsurgical mandibular skeletal relapse were also assessed.

Methods

Twenty-five adult skeletal Class III patients who underwent bimaxillary orthognathic surgery with the SFA were included retrospectively. Cone-beam computed tomography (CBCT) scans were acquired at three time points: preoperatively (T0), immediately postoperatively (T1), and at 1-year follow-up (T2). 3D translational and rotational condylar displacements, as well as dental and skeletal changes and postoperative stability, were quantified using analysis software. Repeated-measures analysis of variance (ANOVA), Pearson or Spearman correlation tests, and multiple linear regression were used for statistical analysis.

Results

A significant anterosuperior mandibular relapse was observed at the 1-year follow-up (T2). Immediately after surgery (T1), the condyles showed significant posterior and inferior translational displacement, accompanied by clockwise pitch and medial yaw rotation. While these positional changes tended to return toward their preoperative position by T2, no significant correlation was found between the magnitude or direction of condylar displacement and mandibular skeletal relapse. In contrast, the initial sagittal position and the surgical change in the upper incisors were significant predictors of relapse. Specifically, the initial anteroposterior position accounted for 20.3% of the variance in sagittal mandibular relapse, indicating that greater dental compensation of the upper incisors was associated with increased mandibular relapse. The modest proportion of explained variance suggests that other unmeasured factors also play important roles in postsurgical stability.

Conclusions

Although postoperative condylar positional changes commonly occur following the SFA in skeletal Class III patients, these changes do not appear to be a primary contributor to mandibular skeletal relapse. Instead, postoperative mandibular stability seems more closely associated with surgical compensatory patterns, particularly upper incisor protrusion. While the SFA demonstrates acceptable skeletal stability, meticulous evaluation and appropriate control of dental decompensation during treatment planning are crucial to minimize postoperative relapse.

Trial registration

The study was registered at the Chinese Clinical Trial Registry on October 24, 2025 (Identification number: ChiCTR2500111068).

Background

Class III malocclusion is relatively common, and severe cases often require combined orthodontic and orthognathic treatment to correct the underlying skeletal discrepancy and to achieve stable occlusion and facial harmony. Traditionally, treatment has followed the conventional approach (CA), in which a pre-surgical orthodontic phase is completed to establish a relatively stable occlusion immediately after surgery [1]. However, this sequence often prolongs overall treatment duration and may transiently worsen facial esthetics, which is a major concern for many patients [2].

In recent decades, the surgery-first approach (SFA) has been introduced [3]. Contrary to the CA, the orthognathic surgery precedes orthodontic treatment in SFA, which has provided immediate improvement in facial appearance, shortened total treatment time, and may potentially increase the efficiency of postoperative tooth movement through the regional acceleratory phenomenon [3]. These advantages have made the SFA an attractive alternative to the CA for both clinicians and patients. Nevertheless, concerns remain regarding the stability and predictability of SFA outcomes, since the surgical occlusion established in SFA is inherently less stable than that achieved after presurgical orthodontics in CA. Consequently, postsurgical orthodontic treatment must resolve residual occlusal discrepancies, during which changes in condylar position and skeletal relationships may occur [4].

Previous studies have suggested that early skeletal relapse, particularly within the first 6–8 weeks after surgery, may be related to proximal segment rotation, condylar sagging, or instability at the osteotomy site [5–7]. To address these issues, several researches have evaluated the effects of treatment approaches (CA versus SFA) and surgical modalities, including bilateral sagittal split ramus osteotomy (BSSRO) or intraoral vertical ramus osteotomy (IVRO), with or without LeFort I osteotomy, on condylar positional changes in skeletal Class III patients, with or without facial asymmetry [5, 8–13]. Overall, the existing literature suggests broadly comparable perioperative and postoperative condylar displacements in Class III patients treated with CA or SFA, regardless of whether one-jaw (BSSRO) [8] or two-jaw surgery (LeFort I with BSSRO [9] or IVRO [10]) was performed. In those patients with facial asymmetry, studies involving BSSRO (either one-jaw or two-jaw surgery) similarly reported no significant differences in proximal segment or condylar positional changes between CA and SFA, or between the deviated and non-deviated sides [11, 12]. More recently, Park et al. reported greater rotational displacement of the condyle on the lesser-setback side in asymmetry cases treated with SFA; however, the magnitude of change was relatively small (< 5°) and unlikely to be clinically meaningful [13].

Despite these findings, the evidence remains limited regarding the relationship between postoperative condylar positional changes and skeletal relapse following SFA. Furthermore, it remains unclear whether the magnitude or direction of condylar displacement, or changes in any skeletal or dental variables, influence the stability of mandibular setback. Therefore, the aim of this study was to use cone-beam computed tomography (CBCT) to investigate postoperative condylar positional changes in skeletal Class III patients undergoing two-jaw surgery with the SFA, and to determine whether these changes are associated with mandibular skeletal relapse or the amount of mandibular setback.

Materials and methods

This retrospective study was approved by the Institutional Review Board of Peking University School and Hospital of Stomatology (No. PKUSSIRB-2025115202). Informed consent was obtained from all participants. A priori sample size estimation was performed to detect a clinically meaningful difference in skeletal stability outcomes (B-VRP) with 80% power and a two-sided α of 0.05 [14], indicating that at least 6 participants were required.

Patients

Adult patients with skeletal Class III malocclusion who underwent SFA with bimaxillary orthognathic surgery, with or without genioplasty, at Peking University School and Hospital of Stomatology between 2016 and 2022 were recruited.

The inclusion criteria were as follows: (1) Adult skeletal Class III patients (ANB < 0°, Wits < -1 mm) treated with two-jaw surgery; (2) Virtual surgical planning (VSP) consistent with the SFA protocol; (3) Availability of CBCT scans at three time points: before surgery (T0), 1 week after surgery (T1), and 1 year after surgery (T2).

The exclusion criteria included: (1) Cleft lip/palate or syndromic craniofacial deformities; (2) History of temporomandibular joint disorders; (3) Significant facial asymmetry (> 4 mm).

Clinical procedures

Preoperative intraoral scans (IOS), 3D facial scans, and CBCT images were obtained at the initial visit (T0). VSP was performed using IVSP-Orthognathic (version 1.0, WeiSiPuLian Technology Co., Ltd, Beijing, China). IOS data replaced dental structures in the CBCT images, and facial scans replaced the CBCT facial soft tissue. The maxillomandibular complex was then repositioned in three dimensions to achieve favorable jaw relationships and facial symmetry. Following established orthodontic principles and recommended SFA guidelines [15], a “treatable” occlusion was jointly determined by the surgeon and orthodontist [16]. Digital splints were designed and fabricated using 3D printing. The intermediate splint guided maxillary repositioning, and the final splint established the definitive mandibular position. Surgery included LeFort I osteotomy, BSSRO, and genioplasty when indicated. Rigid internal fixation was performed using titanium plates and monocortical screws: four microplates at the bilateral pyriform apertures and zygomaticomaxillary crests for the maxilla, and two miniplates at the mandibular osteotomy sites for BSSRO; two additional miniplates were used when genioplasty was performed. Temporary anchorage devices (TADs) were inserted intraoperatively to facilitate intermaxillary ligation, postsurgical elastic traction, and postoperative orthodontic biomechanics.

All patients were treated with fixed orthodontic appliances. Brackets were placed 1–2 weeks before surgery, and postoperative orthodontic treatment began approximately 1 month after surgery. All surgeries were performed by two experienced surgeons (X.J.L. and Z.L.L.).

Data acquisition and 3D measurements

All serial CBCT scans were obtained at three time points (T0, T1, and T2) using the same machine (NewTom VGi; Cefla S.C., Verona, Italy). The scanning matrix was 400 × 400 with 16-bit grayscale depth, a slice thickness of 0.3 mm, and a field of view of 24 cm × 19 cm. Patients were positioned with the midsagittal plane perpendicular to the floor and the Frankfort plane parallel to it, maintaining centric occlusion. Data were exported in Digital Imaging and Communications in Medicine (DICOM) format to Dolphin Imaging (version 11.9; Dolphin Imaging and Management Solutions, Chatsworth, CA).

Reference planes and superimposition

Reference planes were established on the T0 images (Fig. 1): (1) Horizontal reference plane (HRP), the Frankfort horizontal (FH) plane passing through bilateral infraorbital points and the superior point of the right external auditory canal; (2) Midsagittal plane (MSP), a plane through Basion (Ba) and Nasion (N), perpendicular to HRP; (3). Vertical reference plane (VRP), a plane through Ba, perpendicular to both HRP and MSP. The T1 and T2 CBCT images were superimposed on T0 using voxel-based registration at the cranial base, followed by manual slice-by-slice verification. All measurements were performed relative to the T0 reference planes.

Fig. 1.

The reference planes (VRP, HRP, and MSP) were constructed in Dolphin 3D Imaging software

Measurement of condylar displacement

Condylar linear and angular measurements were calculated using Dolphin 3D. Ba served as the origin of the 3D coordinate system, defined as positive in the left, anterior, and superior directions. Landmark definitions and measurement variables are listed in Table 1 and illustrated in Figs. 1, 2, 3 and 4.

Table 1.

Definition of landmarks and measurement variables

Variables	Definition		Bilateral
Landmarks
B	The innermost point on the contour of the mandible between the incisor and bony chin
U6	The midpoint of the bilateral maxillary first molar mesial buccal cusps		√
Ba	The most anterior midpoint of foramen magnum; set as the original point
N	The most anterior midpoint of the nasofrontal suture
A	The deepest midpoint in the curved bony outline from the base to the alveolar process of the maxilla
Pog	The most anterior midpoint of symphysis of mandible
Me	The most inferior point on the symphysis outline
Co	The most superior point of the condyle head		√
Sn	The most inferior point of the sigmoid notch		√
Cp	Tip of coronoid process		√
C-lat	The most lateral point of the condyle head		√
C-mes	The most mesial point of the condyle head		√
Planes
HRP	Horizontal reference plane, the Frankfort horizontal (FH) plane passing through bilateral infraorbital points and the superior point of the right external auditory canal
VRP	Vertical reference plane, a plane through Ba, perpendicular to both HRP and MSP.
MSF	Midsagittal plane, a plane through Ba and N, perpendicular to HRP
Skeletal and dental measurements
A-VRP (mm)	The distance of A to the vertical reference plane
A-HRP (mm)	The distance of A to the horizontal reference plane
B-VRP (mm)	The distance of B to the vertical reference plane
B-HRP (mm)	The distance of B to the horizontal reference plane
Mandibular plane angle (°)	The angle between HRP and the plane of bilateral Co and Me
Overbite (mm)	The distance from the incisal edge of the upper central incisors to the incisal edge of the lower central incisors on the VRP
Overjet (mm)	The distance from the incisal contacts of the upper central incisors to the surface of lower incisors on the HRP
U6-HRP (mm)	The distance of U6 to the horizontal reference plane
U1-VRP (mm)	The distance of the mesial contact of upper incisors
U1-HRP (°)	The angle between HRP and upper incisor axis
Condylar measurements
SJS (mm)	Vertical distance from the most superior aspect of the glenoid fossa to the condyle		√
AJS (mm)	Vertical distance from the most prominent anterior aspect of the condyle to the glenoid fossa		√
PJS (mm)	Vertical distance from the most prominent posterior aspect of the condyle to the glenoid fossa		√
Yaw (°)	The angle between the C-lat and C-mes connection and MSP		√
Roll (°)	The angle between the C-lat and C-mes connection and HRP		√
Pitch (°)	The angel between the Co-Sn connection and HRP		√

Fig. 2.

Landmarks registered on the condyle: Co, Cp, Sn, C-mes, and C-lat

Fig. 3.

3D angular measurements of the condylar position, including roll, yaw, and pitch

Fig. 4.

Sagittal-reformatted CBCT image showing the measurements of AJS (blue line), PJS (green line), and SJS (red line)

Statistical analysis

All statistical analyses were performed using SPSS Statistics (version 27; IBM, Armonk, NY, USA). Intra-examiner reliability was assessed by repeating measurements on 10 randomly selected CBCT datasets after a 4–week interval, and intraclass correlation coefficients (ICCs) were calculated. Data normality was evaluated using the Shapiro-Wilk test. Time-dependent skeletal changes were assessed using repeated-measures analysis of variance (RM ANOVA) or the Friedman test. Pearson or Spearman correlation analyses were used to examine associations between variables and the amount of jaw relapse, serving as the basis for variable selection. Multiple linear regression analysis was then performed to identify predictors of mandibular postsurgical stability. Although a priori power analysis supported a smaller sample, our sample size (N = 25) is modest for multiple regression. To mitigate overfitting, we reported adjusted R² and limited predictors based on theoretical rationale. A P value of < 0.05 was considered statistically significant.

Results

A total of 25 patients (6 males and 19 females) with a mean age of 29.6 ± 3.7 years (range, 24–37 years) were included in the study (Table 2). Intra-examiner reliability was excellent, with all ICCs exceeding 0.90.

Table 2.

Patient baseline characteristics

Characteristics
Age(y)	29.6 ± 3.7
Gender(n)	Males,7; females, 18
SNA (°)	82.4 ± 4.6
SNB (°)	86.4 ± 4.6
ANB (°)	-3.7 ± 3.5
Mandibular plane angle (°)	50.6 ± 4.2

Maxillary and mandibular relapse

During the 1-year postsurgical period, the maxillary position showed no statistically significant changes in either the vertical or sagittal dimensions. However, the mandible exhibited a significant forward (1.7 ± 2.0 mm) and upward relapse (-1.8 ± 2.6 mm) (P < 0.01) (Table 3).

Table 3.

Comparison of treatment changes with surgical movement (T1-T0) and postsurgical changes (T2-T1)

	T0	T1	T2	T1-T0	P	T2-T1	P
Skeletal measurements
A-VRP (mm)	84.4 ± 4.5	87.9 ± 5.1	87.6 ± 4.9	3.5 ± 2.2	< 0.001***	-0.3 ± 1.2	0.228
A-HRP (mm)	30.9 ± 2.9	31.9 ± 4.1	31.9 ± 3.2	1.1 ± 3.2	0.106	-0.1 ± 2.2	0.906
B-VRP (mm)	91.2 ± 6.3	85.9 ± 6.4	87.3 ± 6.7	-5.2 ± 3.0	< 0.001***	1.7 ± 2.0	0.002**
B-HRP (mm)	71.3 ± 7.8	71.0 ± 6.0	69.4 ± 5.9	-0.3 ± 4.3	0.728	-1.8 ± 2.6	0.002**
Mandibular plane angle (°)	50.6 ± 4.2	52.1 ± 3.6	51.3 ± 3.7	1.4 ± 2.5	0.006**	-0.9 ± 1.3	0.003**
Dental measurements
Overbite (mm)	0.2 ± 2.1	0.6 ± 1.8	1.2 ± 1.2	-0.4 ± 2.6	0.481 ƚ	-0.7 ± 1.9	0.104
Overjet (mm)	-1.6 ± 4.0	5.2 ± 2.2	4.2 ± 1.3	6.4 ± 2.9	< 0.001***	-0.9 ± 2.4	0.065
U6-HRP (mm)	48.7 ± 3.7	48.7 ± 3.6	48.5 ± 3.9	0.0 ± 1.8	0.991	-0.2 ± 0.9	0.323
U1-HRP (mm)	118.5 ± 6.6	113.9 ± 6.6	117.9 ± 7.6	-4.6 ± 5.0	< 0.001***	4.0 ± 4.1	< 0.001***
U1-VRP(°)	92.0 ± 5.6	94.0 ± 5.5	94.7 ± 5.8	2.1 ± 2.5	< 0.001***	0.7 ± 1.5	0.036*

Data are presented as the mean ± standard deviation

T Time, T0, preoperative; T1, immediately postoperative; T2, more than one year postoperatively

^ƚ, chi-square test; otherwise, paired t-test

*, P<0.05; **, P<0.01; ***, P<0.001

Translational movements of the condylar head

Table 4 summarizes the comparison of 3D condylar position and angulation throughout different time points. Immediately after surgery, the anterior joint space (AJS) increased significantly on both sides (P < 0.01), whereas the posterior joint space (PJS) showed no significant change (P > 0.05). The superior joint space (SJS) remained unchanged on the right but increased significantly on the left, indicating a posterior and inferior shift of the condyles.

Table 4.

Comparison of bilateral condylar positional changes at T0, T1 and T2, presented as either mean ± SD or median (IQR)

	Con-Rt				Con-Lt				Con-(Rt-Lt)
	T0	T1	T2	P	T0	T1	T2	P	T0	T1	T2	P
Condylar space
AJS	2.6 ± 0.6	3.1 ± 0.8	2.6(0.7)	0.002**a T0-T1, 0.006** T1-T2, 0.014* T0-T2, 1.000	2.4 ± 0.9	3.4 ± 1.2	2.6(1.3)	0.001**a T0-T1, 0.001** T1-T2, 0.085 T0-T2, 0.537	0.5(1.0)	1.0(1.0)	0.5(0.4)	0.875 a
SJS	2.3(0.8)	2.8(1.5)	2.4(0.9)	0.076a	2.6 ± 0.8	3.3 ± 1.4	2.7 ± 0.7	< 0.001*** T0-T1, < 0.001*** T1-T2, 0.001** T0-T2, 0.545	0.4(0.7)	0.8 ± 0.7	0.5(0.6)	0.177a
PJS	2.2 ± 0.7	2.6(1.5)	2.0(0.9)	0.232a	2.2(0.8)	2.4 ± 1.1	2.4 ± 0.6	0.867a	0.5(0.9)	0.5(0.5)	0.4(0.5)	0.380 a
Condylar angle
Pitch	45.7 ± 4.2	47.7 ± 5.3	47.4 ± 4.6	0.008** T0-T1, 0.003** T1-T2, 0.383 T0-T2, 0.002**	46.2 ± 4.6	47.4 ± 4.7	47.3 ± 4.5	0.013** T0-T1, 0.013** T1-T2, 0.868 T0-T2, 0.016**	1.7(1.8)	2.5(3.3)	2.1(3.0)	0.603 a
Roll	6.3(9.5)	3.6(8.1)	3.6(9.5)	0.376a	4.4(9.1)	4.8(10.8)	7.7(11.3)	0.932a	4.1(6.0)	3.3(6.2)	4.1(7.2)	0.887 a
Yaw	75.8 ± 6.4	73.0 ± 7.1	74.8 ± 7.4	0.015* T0-T1, 0.006** T1-T2, 0.076 T0-T2, 0.281	74.5 ± 6.3	71.8 ± 6.9	71.0 ± 7.4	0.019* T0-T1, 0.009** T1-T2, 0.015* T0-T2, 0.537	3.5(5.5)	2.9(5.4)	4.3(6.6)	0.961 a

SD standard deviation, IQR interquartile range, Con-Rt right condyle, Con-Lt left condyle, Con-(Rt-Lt) bilateral difference in condylar spaces and angles

^a, Friedman test, otherwise RM ANOVA

*, P<0.05; **, P<0.01; ***, P<0.001

The position of the condylar center (CC) showed no significant difference between pre-surgery and 1-year post-surgery, suggesting a tendency for the condyles to return to their original positions. Overall, no significant differences were found in bilateral joint space measurements at three time points (T0, T1, and T2).

Rotational movements of the condylar head

Rotational changes of the condyles are shown in Table 4. Immediately post-surgery, both condyles demonstrated significant clockwise pitch and medial yaw rotation compared with pre-surgery values (P < 0.05).

At 1-year post-surgery, significant clockwise pitch rotation persisted bilaterally (P < 0.05), and the left condyle additionally showed significant medial yaw rotation (P < 0.05). No significant difference in yaw, roll, or pitch rotation was observed between T1 and T2. Bilateral condylar rotational changes did not differ significantly across the three time points.

Variables associated with mandibular relapse

To identify factors influencing skeletal stability after the SFA, correlation analyses were performed using initial measurements, surgical condylar changes, and various skeletal and dental variables. As shown in Table 5, mandibular sagittal relapse (B-VRP, T2-T1) demonstrated a significantly positive correlation with the initial upper incisal edge position (U1- VRP, T0; P < 0.05) and a negative correlation with surgical changes in the upper incisal edge position (U1- VRP, T1-T0; P < 0.05). Additionally, stepwise multiple linear regression analysis was conducted to determine the strongest predictors of instability (Table 6). This analysis indicated that greater upper incisor protrusion at T0 was significantly associated with increased anteroposterior mandibular relapse (adjusted R² = 0.203).

Table 5.

Correlation between B-VRP and other variables

Variables	B-VRP (T2-T1)
	T0					T1-T0
A-VRP (mm, T0)	0.329					-0.134
A-HRP (mm, T0)	-0.195					0.021
B-VRP (mm, T0)	0.205					-0.349
B-HRP (mm, T0)	-0.030					0.247
U1-VRP (mm)	0.486*					-0.430*
U1-HRP (°)	0.280					-0.118
U6-HRP (mm)	-0.038					0.079
Overjet	0.239					-0.050
Overbite	0.222a					0.044
	Con-Rt	Con-Lt		Con-(Rt-Lt)		Con-Rt	Con-Lt	Con-(Rt-Lt)
AJS	0.124		0.257a		-0.053a	-0.174a	-0.201	-0.179a
PJS	0.186		0.099a		0.263a	-0.260a	0.074	0.061a
SJS	-0.138a		-0.155		0.232a	-0.246a	-0.019	0.042a
Pitch	-0.048		-0.253		-0.265a	0.050	0.253	-0.167a
Roll	0.219a		0.152a		-0.075a	0.369	0.253	0.353a
Yaw	-0.148		-0.100		0.000a	-0.029a	0.153a	-0.359a

^a, Spearman test, otherwise Pearson test

*, P<0.05

Table 6.

Linear stepwise regression analysis for 1-year mandibular sagittal relapse (B-VRP, T2-T1)

Variables	Unstandardized coefficients β	P value	VIF
B-VRP (T2-T1)
U1-VRP (mm, T0)	0.164	0.014*	1.000
U1-VRP (mm, T1-T0)	-0.249	0.079	1.066

VIF variance inflation factor

*, P< 0.05

B-VRP (T2-T1), adjusted R²=0.203

Discussion

Although the SFA offers clear advantages, including immediate esthetic improvement, reduced treatment duration, and enhanced patient satisfaction, it has not yet achieved universal adoption. The principal primary concern remains long-term skeletal and occlusal stability. Potential contributors to postoperative relapse include the magnitude of surgical movement, control of the proximal segment, fixation method, masticatory muscle tension, postoperative orthodontic mechanics, and residual growth or remodeling [17–19]. This study used CBCT-based 3D analysis to evaluate maxillomandibular relapse and condylar displacement. This methodology enables precise linear and angular measurements and reliable superimposition for longitudinal assessment.

A certain degree of skeletal relapse is generally recognized as unavoidable following orthognathic surgery, regardless of surgical sequencing. In the present study, the maxillary position remained stable throughout the one-year follow-up period; however, the mandible demonstrated a tendency toward anterosuperior relapse. Although the mean sagittal relapse of 1.7 mm reached statistical significance (P < 0.01), its clinical relevance warrants careful interpretation. In orthognathic surgery, a threshold of ≥ 2 mm is commonly regarded as clinically significant for skeletal relapse [20], as it is considered the minimum magnitude likely to be perceived by patients or to influence occlusal stability. The mean relapse observed in this cohort (1.7 mm) falls slightly below this threshold, suggesting that although the anteroposterior change was statistically detectable, it may be within a clinically acceptable range. From a functional perspective, such a minor skeletal change is unlikely to create significant occlusal interferences or compromise long-term occlusal interdigitation, particularly in the absence of associated dental compensations. Regarding patient perception, given that soft tissue changes often mask underlying skeletal shifts of this magnitude, the impact on facial aesthetics or patient satisfaction is likely minimal. Therefore, despite its statistical significance, the clinical importance of this 1.7 mm relapse appears limited, and the procedure can be considered relatively stable in this context. No significant associations were found between postoperative mandibular relapse and the initial maxillary position or surgical movement, which aligns with previous findings [14, 21, 22].

Changes in condylar position following mandibular setback have been implicated in various temporomandibular joint (TMJ) dysfunctions, such as clicking, pain, disc displacement, limited mouth opening, and condylar resorption. In this study, condylar positional changes were observed in all dimensions. Immediately postoperatively, the condyles shifted posteroinferiorly and demonstrated backwards and inward rotation. At the one-year follow-up, most parameters tended to return to their preoperative positions, consistent with earlier SFA and CA research involving both BSSRO and IVRO [8, 13, 14]. Only the inwards rotation of the proximal segment persisted at one year, a finding similar to those reported by other investigators [5, 8]. Postoperative condylar displacement may be attributed to several factors: (1) imprecise intraoperative positioning or seating of the proximal segments [23, 24] (though digital planning and 3D-printed splints may mitigate this risk [25]); (2) altered masticatory muscle balance following surgical detachment or repositioning; (3) intra-articular edema resulting from surgical manipulation of the proximal segment; (4) effects of general anesthesia and muscle relaxants; and (5) altered tongue posture due to reduced intraoral space [26]. Despite these factors, correlation analyses in this study showed no significant association between condylar positional changes and mandibular relapse. Additionally, bilateral TMJ positional differences were not significant during the perioperative period or at the one-year mark. This contrasts with the findings of Jiang et al., who identified downward condylar displacement as a significant predictor of sagittal relapse [15]. This discrepancy may be attributed to the inclusion of premolar extraction cases in their cohort. In such cases, the greater overjet observed immediately after surgery, which is a result of initial upper incisor dental compensation, may prompt patients to protrude their mandible to facilitate mastication, thereby increasing relapse risk.

Upper incisor position emerged as a critical variable associated with postoperative skeletal changes. Correlation analyses identified both the sagittal labiolingual position and the surgical movement of the upper incisor as significant predictors of mandibular relapse. Specifically, stepwise regression analysis showed that the initial labiolingual position of the upper incisor (T0) explained 20.3% of the variance in mandibular relapse at 1 year, with greater protrusion associated with greater relapse. One possible explanation is that pronounced upper incisor proclination at baseline (T0) necessitates substantial postsurgical decompensation; if this movement is insufficient or the resulting dental relationship is unstable, the patient may be more susceptible to relapse. While this study found no significant correlation between sagittal relapse and the magnitude of mandibular setback or condylar positional change, consistent with the findings of de Villa et al. [27]. However, this relationship remains controversial, as other studies suggested that larger setbacks may predispose SFA patients to greater instability [14, 28–30], highlighting the need for careful case selection and management. This modest explanatory power suggests that the unexplained variance likely reflects the multifactorial nature of relapse, which may involve additional variables not captured in this study, such as masticatory muscle adaptation and the quality of postoperative orthodontic management. Future studies incorporating a broader range of potential predictors, including soft tissue factors and muscle function, are warranted to better understand the complex mechanisms underlying skeletal relapse following the surgery-first approach.

This study has several limitations. First, the one-year follow-up may not capture delayed degenerative changes, long-term remodeling, or late orthodontic effects. An extended evaluation is needed to determine the long-term progression of these findings. Second, this study lacked a control group treated via the CA. Establishing such a group retrospectively was not feasible because of the clinical caseload during the study period: most patients treated by the primary orthodontist were SFA cases, and including historical CA patients would introduce chronological bias. However, previous research suggests that perioperative condylar changes are comparable between SFA and CA, indicating similar biomechanical responses regardless of the surgical sequence [5, 8, 14]. Third, although the sample size (N = 25) was sufficient for primary comparisons, it is modest for multiple regression, raising concerns about sampling error, overfitting, and generalizability. To mitigate this, we used adjusted R² and verified low VIFs; nonetheless, future multicenter randomized clinical trials with larger samples are needed to confirm our findings.

Conclusions

In patients with skeletal Class III malocclusion treated with SFA involving bimaxillary surgery, postsurgical relapse primarily occurred in the sagittal and vertical dimensions of the mandible. Most condylar positional changes tended to return to their preoperative position within 1 year. Mandibular stability was significantly associated with the initial dental compensation and the surgical movement of the upper incisors. Specifically, the degree of initial upper incisor protrusion emerged as a key predictor of sagittal relapse.

Abbreviations

3D

Three-dimensional

SFA

Surgery-first approach

CA

Conventional approach

CBCT

Cone-beam computed tomography

RAP

Regional acceleratory phenomenon

BSSRO

Bilateral sagittal split ramus osteotomy

IVRO

Intraoral vertical ramus osteotomy

VSP

Virtual surgical planning

IOS

Intraoral scans

TAD

Temporary anchorage device

DICOM

Digital Imaging and Communications in Medicine

HRP

Horizontal reference plane

MSP

Midsagittal plane

VRP

Vertical reference plane

ICC

Intraclass correlation coefficient

RM ANOVA

Repeated-measures analysis of variance

AJS

Anterior joint space

PJS

Posterior joint space

SJS

Superior joint space

CC

Condylar center

TMJ

Temporomandibular joint

Authors' contributions

Zhou Y.: Conceptualization, Data curation, Formal analysis and Writing-Original Draft; Ye G., Data curation; Liu X. and Li Z.: Resources and Supervision; Zou B.: Conceptualization, Methodology and Writing- Review &amp; Editing.

Funding

This research was supported by Peking University School and Hospital of Stomatology (No. YS020217).

Data availability

The datasets used and analyzed in the present study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of Peking University School and Hospital of Stomatology (No. PKUSSIRB-2025115202;24.09.2025) and conducted in accordance with the principles of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Informed consent was obtained from all participants.

Oral informed consent was obtained from all participants prior to their inclusion in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.