Improved diagnostic accuracy with the classification tree method for diagnosing low-grade periprosthetic joint infections by quantitative measurement of synovial fluid alpha-defensin and C-reactive protein.

Abstract

The diagnosis of low-grade periprosthetic joint infections (PJIs) is challenging, because patients may present with unspecific symptoms, false-negative cultures, or marginally elevated values of serum biomarkers like C-reactive protein (CRP). This may lead to the unintended implantation of a revision prosthesis into an infected surgical site with a repeat risk of short-term failure. Conversely, false diagnosis of joint infection may result in multistage revision procedures, which expose the patient to unnecessary surgical procedures and inappropriate antibiotic treatment. Here, we investigated whether synovial biomarkers can preoperatively distinguish between aseptic prosthesis loosening and low-grade joint infection and the most accurate biomarker combinations. Inclusion criteria for the study were indication for revision arthroplasty due to aseptic implant failure, acute high-grade infection, or (suspected) low-grade infection. We prospectively collected synovial fluid of patients undergoing revision arthroplasty for quantitative measurement of alpha defensin, CRP, interleukin (IL-6), IL-10, and lipopolysaccharide binding protein (LBP). The classification tree method revealed alpha defensin and CRP as the most suitable biomarker combination to distinguish between aseptic loosening and low-grade joint infection. The combination of CRP > 2.0mg/L and alpha defensin > 90.000pg/mL correctly identified nine of 11 patients with low-grade infection. Alpha defensin plus CRP seems to be the most helpful combination for pre-operative discrimination of aseptic loosening vs. low-grade joint infection.