Journal of Orthopaedic Surgery and Research (Nov 2019)
Does serum interleukin-6 guide the diagnosis of persistent infection in two-stage hip revision for periprosthetic joint infection?
Abstract
Abstract Purpose The diagnosis of persistent infection at reimplantation of two-stage revision arthroplasty for periprosthetic joint infection (PJI) remains challenging. Several studies have shown the benefit of serum interleukin-6 (IL-6) in diagnosing periprosthetic joint infection (PJI). Recent data indicated serum IL-6 could be promising in differentiating persistent infection. The purpose of this study was to validate the efficacy of serum IL-6 in diagnosing persistent infection at reimplantation. Methods A retrospective review of 86 PJI patients with a two-stage exchanged hip arthroplasty from 2013 to 2017 was conducted. Persistent infection was defined using the modified Musculoskeletal Infection Society (MSIS) criteria combined with follow-up results. Serum IL-6 at reimplantation were collected and compared among patients with or without persistent infection. Receiver operating characteristic (ROC) curves were generated to evaluate the diagnostic performance and optimal cut-off value of serum IL-6 at reimplantation. Results Sixteen cases were diagnosed as persistent infection at reimplantation. There was no significant difference in serum IL-6 levels between cases with persistent infection and controls (7.89 pg/ml vs. 5.56 pg/ml; P = 0.179). The area under the ROC curve (AUC) for serum IL-6 in diagnosing persistent infection at reimplantation was 0.59 (95% confidential interval [CI] 0.40–0.77). With the calculated threshold set at 8.12 pg/ml, the corresponding sensitivity, specificity, positive predictive value, and negative predictive values were 38%, 88%, 38%, and 87%, respectively. Conclusion Serum IL-6 is inadequate in diagnosing persistent infection at reimplantation for two-stage revision arthroplasty. With the serum IL-6 threshold set at 8.12 pg/ml, the specificity to rule out persistent infection is high, but the sensitivity to predict persistent infection is not satisfactory.
Keywords