Text

Objectives and questions: Custom acetabular implants are increasingly used for reconstruction of severe acetabular defects in revision total hip arthroplasty (rTHA). Earlier geometry-based design approaches often relied on simplified shapes, potentially requiring additional intraoperative bone preparation. Newer bone-conforming designs aim to replicate patient-specific defect morphology more closely and may reduce the extent of bone removal.

Aim: To compare planned and achieved bone removal between geometry-based and bone-conforming custom acetabular implants.

Material and methods: This retrospective cohort study included patients undergoing rTHA for Paprosky type III acetabular defects treated with custom-made, three-dimensional (3D) printed titanium acetabular implants. Twenty-nine patients received geometry-based implants and twenty received bone-conforming implants. All implants were designed using preoperative computed tomography (CT)-based 3D reconstructions. Bone removal was quantified by comparing preoperative surgical plans with postoperative CT-based 3D models using validated image segmentation and registration workflows. Planned bone removal volume was derived from the preoperative plan, and achieved bone removal volume was calculated from postoperative imaging. Group comparisons were performed using non-parametric statistical tests (Mann-Whitney U test). Statistical significance was set at p < 0.05.

Results: Bone-conforming implants demonstrated consistently lower bone removal requirements across both planned and achieved measurements. Median planned bone removal was reduced by 60%, measuring 2 cm³ (interquartile range [IQR] 1–3) in the bone-conforming group compared with 5 cm³ (IQR 3–13) in the geometry-based group (p < 0.0001). The geometry-based group demonstrated greater variability in planned bone resection, indicating less predictable host bone preservation. Postoperative analysis confirmed that these differences translated into intraoperative outcomes. Median achieved bone removal remained significantly lower in the bone-conforming group at 3 cm³ (IQR 1–6) compared with 6 cm³ (IQR 4–8) in the geometry-based group (p = 0.0118), representing an approximate 50% reduction in bone loss. The narrower interquartile range observed in the bone-conforming group suggests improved reproducibility between planned and achieved resection.

Discussion and conclusions: Bone-conforming custom acetabular implants significantly reduce bone removal and demonstrate improved consistency between planned and achieved resections compared with geometry-based designs, supporting their use in complex acetabular reconstruction.