Poor utility of serum interleukin-6 levels to predict indolent periprosthetic shoulder infections.

Background Infection after shoulder arthroplasty can present a diagnostic challenge. The purpose of this study was to evaluate the utility of serum interleukin-6 (IL-6) levels in diagnosis of periprosthetic infection in patients undergoing revision shoulder arthroplasty. Methods We prospectively enrolled 69 patients who underwent revision shoulder arthroplasty at one institution. All patients underwent a standard preoperative and intraoperative workup for infection, which included shoulder aspirate culture, erythrocyte sedimentation rate, C-reactive protein level, tissue culture, and frozen section analysis. In addition, serum levels of IL-6 were measured preoperatively in all patients. Infection classification was divided into 4 groups, (1) definite, (2) probable, (3) possible, and (4) no infection, on the basis of previously reported criteria using intraoperative cultures and preoperative and intraoperative findings of infections. Results Of the 69 patients, 24 were classified as having a definite or probable infection. Propionibacterium acnes was the offending organism for the majority of these cases (20 of 24, 83%). IL-6 was not a sensitive marker of infection for these patients (sensitivity: 3 of 24, 12%; specificity: 3 of 45, 93%). The sensitivity of serum IL-6 was lower compared with erythrocyte sedimentation rate (sensitivity: 10 of 24, 42%; specificity: 37 of 45, 82%) and C-reactive protein level (sensitivity: 11 of 24, 46%; specificity: 42 of 45, 93%). For the non-P. acnes cases (1 Staphylococcus aureus, 1 Enterobacter cloacae, 2 coagulase-negative Staphylococcus species), the sensitivity of IL-6 was 25% (1 of 4). Conclusion Serum IL-6 is not an effective marker for diagnosis of infection in shoulder arthroplasty. On the basis of this large prospective study, we do not recommend its use as a preoperative diagnostic test in patients undergoing revision shoulder arthroplasty. [ABSTRACT FROM AUTHOR]