Three-dimensional computed tomography evaluation of early mandibular reconstruction after costochondral graft in children With unilateral Pruzansky–Kaban type IIB/III hemifacial microsomia

Abstract Background Quantitative evidence on early mandibular reconstruction after costochondral graft (CCG) in children with severe hemifacial microsomia (HFM) remains limited, particularly regarding dimension-specific postoperative changes, the distinction between immediate reconstructive gain and short-term maintenance, and their relationship with surgical age. This study evaluated early postoperative mandibular changes after CCG in children with unilateral Pruzansky–Kaban type IIB/III HFM using three-dimensional computed tomography (3D-CT). Methods This retrospective single-center imaging study included 22 children (16 type IIB and 6 type III) younger than 18 years who underwent primary CCG reconstruction and had complete maxillofacial CT data before surgery (T0), at the first postoperative CT examination (T1), and at early follow-up (T2; median T1–T2 interval, 12.0 months). Bilateral ramus height (RH), body length (BL), and ramus volume (RV) were measured on three-dimensional reconstructions. Reconstructed-to-contralateral ratios, immediate gains, and secondary descriptive maintenance indices were calculated. Surgical age was analyzed primarily as a continuous variable and secondarily in the < 6 years, 6–9 years, and ≥ 9 years groups; Holm adjustment was applied separately to the four group comparisons and the four correlations. Results Reconstructed-side RH and RV increased immediately after surgery and remained above baseline at T2. RHR increased from 0.56 ± 0.10 at T0 to 0.70 ± 0.11 at T1 and 0.72 ± 0.12 at T2; RVR increased from 0.55 ± 0.16 to 0.66 ± 0.15 and then decreased to 0.61 ± 0.16. The median BLR was 0.84 at T0, 0.83 at T1, and 0.82 at T2; values at T2 were lower than those at T0 and T1 (adjusted p = 0.0016 and 0.0027, respectively), although the absolute differences were small. None of the four age-group comparisons was significant. Surgical age showed nominal correlations with RVG ( r = 0.458, p = 0.032) and RVMI ( r = − 0.453, p = 0.034), but neither remained significant after Holm adjustment (both adjusted p = 0.129). Conclusions CCG was associated with immediate and early improvement in reconstructed-side ramus height and local osseous ramus volume, whereas BLR showed a small but statistically significant decrease from a median of 0.84 at T0 to 0.82 at T2. These findings describe early postoperative skeletal change and should not be extrapolated to skeletal maturity or long-term graft stability, and they do not establish an optimal surgical age. Exploratory age-related trends were observed, but require confirmation in larger cohorts.