PREDICTIVE VALUE OF PROSTHESIS SONICATION ON SURVIVAL OF TWO-STAGE REVISION FOR THE TREATMENT OF PERIPROSTHETIC JOINT INFECTION

Introduction & aim Chronic periprosthetic joint infection (PJI) is often treated with two-stage revision. The success rate of two-stage revision arthroplasty ranges from 65-93%. Failure to eradicate infection often necessitates further operation with associated morbidity. While preoperative factors predictive of failure have been established, current literature identifying risk factors acutely following the first stage, resection arthroplasty, is scarce. The use of colony forming units (CFU) from sonication of prostheses has been shown to correlate with the diagnosis of PJI and serves as an adjunct in identifying PJI when Musculoskeletal Infection Society (MSIS) criteria are incomplete or confounded by methodological errors. The utility of CFU from sonicated prostheses in predicting failure of two-stage revision has not been explored. We sought to identify a CFU value predictive of two-stage revision arthroplasty failure. Methods Receiver operating characteristic (ROC) analysis identified CFU threshold most predictive of treatment failure. Patients were stratified into cohorts based on CFU results. Baseline characteristics were compared using one-way ANOVA and chi-squared tests. Failure-free survival was compared using Kaplan-Meier analysis. Failure was defined as the need for revision arthroplasty related to PJI or patient death related to infection within two years of the index operation. Results A total of 204 patients underwent 2-Stage revision for treatment of PJI and had a documented numeric CFU. A threshold of 57.5 CFU was most predictive of treatment failure. There were no significant differences in baseline characteristics between cohorts. The CFU 0-50 cohort (14.75%) had a significantly lower failure rate compared to CFU 50-100 (37.5%) and CFU >100 cohort (24.75%) (p = 0.046). However, failure-free survival did not demonstrate a significant difference between cohorts (p = 0.08). Cox proportional hazard model revealed CFU cohort as a significant predictor of failure-free survival (HR = 5.63, p < 0.01). Conclusions We identified a threshold of bacterial CFU derived from sonication of explanted prostheses predictive of treatment failure following the first stage of two-stage revision arthroplasty. Further, cases with CFU 0-50 were less likely to fail two-stage revision arthroplasty than those with higher CFU values. This data is available acutely postoperatively and thus allows for consideration of earlier intervention. Adjustments to standard treatment including antibiotic duration or antibiotic spacer exchange prior to reimplantation may improve success of two-stage revision arthroplasty for PJI with sonication culture results > 57 CFU. Higher powered analyses may one day delineate a CFU threshold predictive of two-stage revision arthroplasty failure.