Abstract
BACKGROUND: Two-stage revisions for chronic periprosthetic joint infections (PJIs) often include antibiotic-loaded cement spacers to control for infection and preserve function. While studies have reported on complications (dislocations, readmissions, and reoperations) after static versus articulating spacer types, there is a paucity of evidence about whether the degree of spacer constraint in articulating spacers affects these complications. This study aims to address a key gap in understanding as to whether the level of spacer constraint affects complications in two-stage revision THA utilizing articulating spacers. QUESTIONS/PURPOSES: (1) Among patients receiving nonconstrained versus constrained articulating antibiotic spacers during first-stage revision THA for PJI, are there differences in major complications, such as dislocation, loosening, periprosthetic fracture, reinfection, and unplanned revisions? (2) After second-stage reimplantation, do patients who received a nonconstrained versus constrained liner during the first stage show differences in the risk of complications, reoperations, and readmissions? METHODS: This is a retrospective review of 539 patients who underwent two-stage revision THA for PJI at a single-specialty, urban academic referral center between July 2011 and March 2023. Of these 539 patients, 72% (388) were excluded for undergoing a full component revision (femoral or acetabular) for any reason before their first stage, 3% (15) for receiving static spacers, and 6% (35) for receiving prefabricated femoral mono-block stems as part of their first stage. Those who underwent only liner exchange were not excluded. The remaining 19% (101 of 539) of patients were included in the final analysis and categorized by degree of liner constraint: 32 were in the nonconstrained group and 69 were in the constrained group. All surgeons included in this study specialize in adult reconstruction and are fellowship trained, and the selected level of constraint was solely based on their routine practice for articulating spacer construct. Baseline characteristics and clinical data, including age, self-reported gender, race, BMI, American Society of Anesthesiologists score, smoking status, surgical history, and perioperative details, were collected. There were no differences in baseline characteristics between the groups except for smoking status. A priori power analysis determined that 150 patients (75 per group) would be needed to detect a statistical difference in the risk of dislocation between groups, assuming a 20% dislocation risk for the constrained group, at a 0.05 alpha level, and 80% power. RESULTS: Between patients receiving nonconstrained versus constrained liners, there were no differences in complications after the first stage of revision. Three percent (1 of 32) of the nonconstrained liners developed dislocations compared with 3% (2 of 69) in the constrained group (relative risk [RR] 1.1 [95% confidence interval (CI) 0.09 to 12.3]; p > 0.99). Three percent (1 of 32) of the nonconstrained group developed periprosthetic fractures compared with 7% (5 of 69) in the constrained group (RR 0.4 [95% CI 0.05 to 3.69]; p = 0.72). Similarly, 3% (1 of 32) versus 7% (5 of 69) had persistent infection (RR 0.4 [95% CI 0.05 to 3.69]; p = 0.72). One incident of loosening occurred in the constrained group. There were also no differences in spacer revision incidence: 10% (3 of 32) of the nonconstrained group and 10% (7 of 69) of the constrained group underwent an unplanned revision after the first stage (RR 1.0 [95% CI 0.29 to 3.91]; p = 0.91). For the second stage, dislocation was 14% (3 of 21) in the nonconstrained group and 10% (5 of 52) in the constrained group (RR 1.1 [95% CI 0.2 to 5.9]; p > 0.99). When comparing periprosthetic fractures, 10% (2 of 21) of the nonconstrained group developed periprosthetic fractures compared with 4% (2 of 52) in the constrained group (RR 2.2 [95% CI 0.3 to 16.6]; p = 0.78). Nineteen percent (4 of 21) in the nonconstrained group had persistent infection compared with 12% (6 of 52) in the constrained group (RR 1.5 [95% CI 0.39 to 5.74]; p = 0.81). The occurrence of readmission after the second stage was 19% (4 of 21) in the nonconstrained group compared with 15% (8 of 52) in the constrained group (RR 1.1 [95% CI 0.3 to 3.9]; p > 0.99). Twenty-four percent (5 of 21) of patients in the nonconstrained group required a surgery-related emergency department visit compared with 13% (7 of 52) in the constrained group (RR 1.6 [95% CI 0.4 to 5.6]; p = 0.64). The incidence of reoperation was 14% (3 of 21) in the nonconstrained group and 13% (7 of 52) in the constrained group (RR 0.9 [95% CI 0.2 to 3.8]; p > 0.99). CONCLUSION: Our results indicated no differences in the risk of dislocations, reinfections, reoperations, and readmissions between patients undergoing constrained versus nonconstrained articulating spacers for two-stage revision THA. Because constrained liners are typically preferred in patients at higher risk of instability, our findings suggest that their use does not necessarily increase the risk of complications. However, because of the small sample size, larger studies are needed to demonstrate whether there is superiority of liner constraint in this patient population. LEVEL OF EVIDENCE: Level III, therapeutic study.