What Is the Risk of Dislocation and Revision in Proximal Femoral Replacement with Dual-mobility Articulation After Two-stage Revision for Periprosthetic Hip Infection?

Abstract

Dislocation is a major complication of revision THA after two-stage exchange for periprosthetic joint infection (PJI). The likelihood of dislocation can be particularly high if megaprosthetic proximal femoral replacement (PFR) has been performed during a second-stage reimplantation. Dual-mobility acetabular components are an established way of reducing the instability risk in revision THA; however, the likelihood of dislocation for dual-mobility reconstructions in the setting of a two-stage PFR has not been studied systematically, although patients with these reconstructions might be at an increased risk. (1) What is the risk of dislocation and revision for dislocation in patients who underwent PFR with a dual-mobility acetabular component as part of two-stage exchange for hip PJI? (2) What is the risk of all-cause implant revision and what other procedures were performed (apart from revision for a dislocation) in these patients? (3) What potential patient-related and procedure-related factors are associated with dislocation? This was a retrospective study from a single academic center including procedures performed between 2010 and 2017. During the study period, 220 patients underwent two-stage revision for chronic hip PJI. Two-stage revision was the approach of choice for chronic infections, and we did not perform single-stage revisions for this indication during the study period. Thirty-three percent (73 of 220) of patients underwent second-stage reconstruction with a single-design, modular, megaprosthetic PFR because of femoral bone loss, using a cemented stem. A cemented dual-mobility cup was the approach of choice for acetabular reconstruction in the presence of a PFR; however, 4% (three of 73) were reconstructed with a bipolar hemiarthroplasty to salvage an infected saddle prosthesis, leaving 70 patients with a dual-mobility acetabular component and a PFR (84% [59 of 70]) or total femoral replacement (16% [11 of 70]). We used two similar designs of an unconstrained cemented dual-mobility cup during the study period. The median (interquartile range) patient age was 73 years (63 to 79 years), and 60% (42 of 70) of patients were women. The mean follow-up period was 50 ± 25 months with a minimum follow-up of 24 months for patients who did not undergo revision surgery or died (during the study period, 10% [seven of 70] died before 2 years). We recorded patient-related and surgery-related details from the electronic patient records and investigated all revision procedures performed until December 2021. Patients who underwent closed reduction for dislocation were included. Radiographic measurements of cup positioning were performed using supine AP radiographs obtained within the first 2 weeks after surgery using an established digital method. We calculated the risk for revision and dislocation using a competing-risk analysis with death as a competing event, providing 95% confidence intervals. Differences in dislocation and revision risks were assessed with Fine and Gray models providing subhazard ratios. All p values were two sided and the p value for significance was set at 0.05. The risk of dislocation (using a competing-risks survivorship estimator) was 17% (95% CI 9% to 32%) at 5 years, and the risk of revision for dislocation was 12% (95% CI 5% to 24%) at 5 years among patients treated with dual-mobility acetabular components as part of a two-stage revision for PJI of the hip. The risk of all-cause implant revision (using a competing-risk estimator, except for dislocation) was 20% (95% CI 12% to 33%) after 5 years. Twenty-three percent (16 of 70) of patients underwent revision surgery for reinfection and 3% (two of 70) of patients underwent stem exchange for a traumatic periprosthetic fracture. No patients underwent revision for aseptic loosening. We found no differences in patient-related and procedure-related factors or acetabular component positioning for patients with dislocation with the numbers available; however, patients with total femoral replacements had a higher likelihood of dislocation (subhazard ratio 3.9 [95% CI 1.1 to 13.3]; p = 0.03) and revision for a dislocation (subhazard ratio 4.4 [95% CI 1 to 18.5]; p = 0.04) than those who received PFR. Although dual-mobility bearings might be an intuitive potential choice to reduce the dislocation risk in revision THA, there is a considerable dislocation risk for PFR after two-stage surgery for PJI, particularly in patients with total femoral replacements. Although the use of an additional constraint might appear tempting, published results vary tremendously, and future studies should compare the performance of tripolar constrained implants to that of unconstrained dual-mobility cups in patients with PFR to reduce the risk of instability. Level III, therapeutic study.