One-Stage Versus Two-Stage Exchange Arthroplasty for Periprosthetic Joint Infection

BACKGROUND: A 2-stage approach is most commonly used to treat periprosthetic joint infection (PJI). Some successful studies of the 1-stage approach were underpowered, lacked a 2-stage comparative group, and excluded patients with draining sinuses, comorbidities, and/or antibiotic-resistant organisms. Given the morbidity and expense associated with 2-stage treatment, we conducted a prospective, multicenter, randomized trial to compare the results of 1- and 2-stage PJI treatment, specifically including patients with draining sinuses, comorbidities, and resistant organisms. METHODS: Patients presenting for surgical treatment of a chronic PJI with a known organism following primary total hip or knee arthroplasty were included (with infection defined by Musculoskeletal Infection Society [MSIS] criteria). Patients with prior revision, culture-negative infection, or fungal infection, or who were immunosuppressed or had soft-tissue involvement precluding wound closure, were excluded. Patients were classified according to the McPherson host staging system. Clinical success was defined as (1) no clinical failure or reinfection with the same or new organism; (2) no reoperation for PJI; and (3) no PJI-related death. A double-instrument setup was used for all patients, as were similar irrigation and antibiotic protocols. A total of 323 patients (166 one-stage; 157 two-stage) were randomized. Groups were similar with respect to demographics and host classification. After excluding patients who died or were lost to follow-up, 258 of the 323 patients had 2-year follow-up (135 one-stage and 123 two-stage). The rate of patient loss to follow-up was similar between the treatment groups. RESULTS: Sixteen patients in the 1-stage group and 9 patients in the 2-stage group died prior to 2-year follow-up. Overall, the 2-year success rate of 1-stage treatment was 97% (131 of 135), while the success of 2-stage treatment was 91% (112 of 123) (p = 0.04). Compared with the 2-stage group, the 1-stage group had 3-times the odds of overall success in a regression analysis (unadjusted odds ratio = 3.22 [95% confidence interval = 1.0 to 10.38]). After adjusting for specific variables (McPherson host grade, resistant organism, and draining sinuses), 1-stage treatment also had 3-times the odds of success. CONCLUSIONS: The results of this prospective randomized trial indicated that 1-stage treatment (97% success) was statistically noninferior to 2-stage treatment (91% success) when treating chronic PJI following primary total hip or knee arthroplasty, provided the protocols described here are explicitly followed. Extrapolation to other patient cohorts and clinical situations should be avoided. LEVEL OF EVIDENCE: Therapeutic Level I. See Instructions for Authors for a complete description of levels of evidence.