•Introduction and Aim: Candida periprosthetic joint infection (PJI) is usually treated with joint removal, but clinical characteristics, surgical management, and duration of antifungal therapy vary.  We aimed to analyze clinical, microbiological and surgical characteristics of patients diagnosed with Candida PJI of hip or knee.

•Methods: Retrospective study of adults admitted 2017-2024 with Candida PJI of hips and knees.  Patients were identified using ICD-10 codes and cross-referenced with candida cultures from a microbiology database. Charts were reviewed for comorbidities, prior bacterial PJI, prior surgeries, bacterial coinfection, antifungal regimen and surgical management. Outcomes were assessed through 2-year follow-up. Failure defined as recurrent infection requiring additional surgery, amputation or death. Patients were stratified based on isolation of Candida before planned PJI surgery or from operative cultures.

Results:  29 patients met inclusion criteria, median  63 years [IQR 58–71]; 13 (44.8%) male, 15 (51.7%)  knee PJI, 19  prior bacterial PJI (65.5%) and 15 bacterial coinfection (51.7%),  median 3 prior surgeries [IQR 2–6] before fungal PJI diagnosis. C. albicans most common isolate, seen in 18 (62.1%) using bacterial and fungal cultures. No patient had fungemia. Surgery involved spacer placement in 19 (67.9%) and DAIR in 6 (21.4%). Most received fluconazole alone, 25/29 (86.2%) with median antifungal duration of 203 days [IQR 150–300]. Candidal infection identified before planned PJI surgery in 11 (37.9%) and intraoperatively in 18 (62.1%); those diagnosed before surgery were younger, with median age 58 vs. 66.5 years (p=0.03) with higher prior surgical burden; median 6 vs. 2.5 prior surgeries (p=0.05). Despite prolonged therapy, outcomes were poor, with treatment failure in 13 (44.8%) including death in 4 (31%) without difference between groups. (Table 1)

Conclusion: Our patients with candida PJI had multiple prior surgeries and candidal infection often not suspected preoperatively. Bacterial coinfection was common and approach to treatment was variable. Preoperative diagnosis did not impact outcome. More studies are needed to improve diagnosis and treatment.