Do the porous custom implants have a position consistent with the planning and allow anatomical reconstruction of hip center of rotation in complex acetabular revisions Paprosky III?

Abstract

Introduction: In revision total hip arthroplasty (THA), the advent of porous custom-made triflange acetabular implants with 3D scan planning offers a new perspective to improve implantation accuracy and anatomical restoration of the center of rotation (COR). This issue was investigated using CT-scan as the measurement tool, but in limited series (±10 cases) and without investigating the factors that may influence errors in positioning. Therefore we performed a retrospective study aiming to: (1) assess the placement accuracy of such implants with respect to the preoperative planning, (2) examine whether the volume of bone to be resected in order to apply the implant had an impact on this accuracy, (3) assess if errors in position at surgery had any influence on function, complications and survival.

Hypothesis: Preoperative planning could be accurately reproduced when implanting porous custom-made acetabular implants, and that accuracy would decrease in proportion to the volume of bone to be resected METHOD: Twenty patients undergoing THA revision with porous custom-made acetabular implants were included in this single-center retrospective study. Mean follow-up was 17.9 months ± 9.4 [2-45.1]. Preoperative planning was performed using 3D scanographic modeling. A post-operative CT scan was performed to assess implantation accuracy in terms of orientation and COR restitution. Demographic data, Oxford scores, complications and survival were recorded.

Results: Mean deviation from the preoperative planning in inclination, anteversion and rotation were 4.3 ° ± 2.5, 6.1 ° ± 4.7, and 7 ° ± 4.6, respectively. Restoration of the COR showed a mean deviation of 2.1 ± 1.3 mm anteroposteriorly, 2.5 ± 2 mm mediolaterally and 2.2 ± 1.3 mm proximodistally. In total, 45% (9/20) of implants were positioned with perfect restoration of orientation (±10 °) and COR (±5 mm). The mean planned bone resection was 8.1 ± 4.9 cm³, with placement accuracy and COR restitution decreasing significantly when the volume of bone to be resected exceeded 2.7 cm³. One dislocation was found (5%, 1/20). Survival at last follow-up was 100%, the mean Oxford score at follow-up was 31.7 ± 7.9 [16-52], without being influenced by errors in position or COR restitution.

Conclusion: In total 45% of the implants restored an orientation and a COR as planned, particularly when the volume of bone to be resected is less than 2.7 cm3. Although these are complex cases with large amounts of bone loss, 3D manufacturing could give us hope of greater precision. The link between better precision and low bone resection volume could be an area to develop with the manufacturer in order to improve results.

Level of evidence: III; diagnostic using CT in transversal retrospective study.