Authors: Yu, Elizabeth, MD, Berbari, Elie, MD, Osmon, Douglas, MD, Berry, Daniel, MD, Mayo Clinic, Rochester, MN
Title: Periprosthetic Infection with Propionibacterium acnes: Different from Conventional Arthroplasty Infection
Purpose: Propionibacterium acnes, once thought a contaminant in prosthetic joint infection, is now recognized to cause true prosthetic infections. Aim: To characterize clinical findings and evaluate treatment outcome of P. acnes prosthetic infections.
Methods: A retrospective cohort study of all patients with P. acnes total hip or knee infection diagnosed over a 32 year period at one institution using a strict case definition was conducted. Presentation, treatment and outcomes were analyzed.
Results: Among 2335 prosthetic joint infections treated during the study period 1% (29 of 2335) were due to P. acnes. Median patient age: 62 years (36-80). There were 20 hips and 9 knees. 13 of 29 had previous revisions. Pain was present in 22 of 29. The median preop peripheral WBC was 7800. Median ESR of the 24 patients in whom this test was performed: 35 (range 1-181). In 11 patients ESR was < 25. 20 of 29 (69%) had radiographic periprosthetic lucencies at diagnosis. Intraoperatively, 17 of 29 (59%) had prosthetic loosening but notably only 9 of 29 (39%) had acute or subacute inflammation on histology. 6 patients were treated with resection arthroplasty or debridement. 10 of 29 patients underwent a 2-stage exchange: 7 of these 10 (70%) did not require further surgery. 13 patients underwent partial component revision or single stage exchange because infection was not diagnosed until postop: 6 of these 13 (46%) did not require further surgery. The overall 2- and 5-year survi!val free of reinfection or reoperation for any reason was 89% and 84% respectively. Of 7 patients who became reinfected, P. acnes was the organism in 3.
Discuassion and Conclusion: P. acnes prosthetic joint infection commonly has nonspecific symptoms and radiographic findings at presentation: ESR was < 25 in 46% and intraop histology was negative for acute inflammation in over 60% of patients, leading a substantial number of patients not to be diagnosed as infected until after revision.
