Association between pterygomaxillary separation during surgically assisted rapid maxillary expansion and three-dimensional morphological and morphometric changes of the mandibular condyle: a cone-beam computed tomography study.

Abstract

Objectives: This study aimed to evaluate the association between pterygomaxillary separation (PMS) performed during surgically assisted rapid maxillary expansion (SARME) and three-dimensional morphological and morphometric changes of the mandibular condyle using cone-beam computed tomography (CBCT).

Materials and methods: This retrospective CBCT-based study included 50 independent adult patients who underwent SARME, with 25 patients in each PMS group. Both condyles of each patient were segmented and measured separately to preserve side-specific morphological information, yielding 100 condylar imaging observations from 50 patients. Preoperative and postoperative CBCT data were used to perform condylar segmentation and generate three-dimensional models. Condylar volume, surface area, root mean square distance (RMS), mean signed distance (MSD), minimum and maximum surface distances, apposition/resorption distribution, and right-left condylar asymmetry were assessed. To account for the correlation between bilateral condyles from the same patient, the principal condyle-level outcomes were analyzed using generalized estimating equations (GEE) with patient as the clustering unit and robust standard errors. Condylar volume and surface area were evaluated longitudinally using the PMS group × time interaction, and a patient-level sensitivity analysis based on bilateral condylar averages was additionally performed.

Results: Baseline condylar volume and surface area values were generally similar between the PMS(+) and PMS(-) groups (p > 0.05). After SARME, measurable morphological and morphometric changes were observed in the condyle in both groups. After accounting for within-patient clustering, the estimated between-group difference in longitudinal condylar volume change was 47.61 mm³ (95% CI, - 1.31 to 96.53; p = 0.056). Nominal cluster-adjusted between-group associations were observed for RMS (estimated difference, 0.067 mm; 95% CI, 0.004 to 0.131; p = 0.038; q = 0.113) and minimum surface distance (estimated difference, - 0.240 mm; 95% CI, - 0.434 to - 0.046; p = 0.015; q = 0.092). However, neither remained statistically significant after Benjamini-Hochberg FDR correction. Moreover, these surface-distance estimates could not be demonstrated to exceed imaging and measurement uncertainty and were therefore regarded as exploratory observations of uncertain methodological and clinical significance. The longitudinal condylar volume interaction was also not statistically significant after correction (p = 0.056; q = 0.113), and no statistically detectable differences were observed for surface area change, MSD, or maximum surface distance. Patient-level sensitivity analysis based on bilateral condylar averages yielded a similar overall pattern. Apposition/resorption distributions were summarized descriptively without formal condyle-level inference.

Conclusions: After accounting for within-patient clustering and multiple comparisons, none of the principal imaging outcomes remained statistically significant. Nominal between-group associations were observed for RMS and minimum surface distance before FDR correction, but their adjusted q values exceeded 0.05. Patient-level sensitivity analyses yielded a similar overall pattern. These findings should therefore be interpreted as exploratory imaging observations of uncertain methodological and clinical significance rather than established evidence of condylar remodeling; moreover, they should not be attributed specifically to PMS because residual confounding by indication cannot be excluded.

Clinical relevance: Exploratory CBCT-based surface-distance observations differed nominally according to PMS status, but none of the principal imaging outcomes remained statistically significant after FDR correction, and the observed surface-distance differences could not be demonstrated to exceed imaging and measurement uncertainty. Their clinical significance remains uncertain, and residual confounding by indication precludes attributing these observations specifically to PMS.