Diagnosing Periprosthetic Joint Infection in Inflammatory Arthritis: Assumption is the Enemy of True Understanding

Background: Despite concern over the interpretation of serum and synovial fluid tests to screen and diagnose chronic periprosthetic joint infection (PJI) in patients with inflammatory arthritis, only a single study has specifically investigated this area. Our aim was to assess diagnostic accuracy of clinical and laboratory markers for diagnosis of PJI in the context of underlying inflammatory arthritis within a large multicenter study of patients undergoing revision surgery.

Methods: This multicenter study was conducted on patients undergoing revision total joint arthroplasty (TJA) at three different centers between January 2001 and July 2016. PJI was defined based on Musculoskeletal Infection Society (MSIS) major criteria. Acute PJI cases were excluded. Aseptic revisions were patients operated for a diagnosis other than infection, who did not subsequently fail at one year follow-up. Serum C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), synovial white blood-cell (WBC) and differential, as well as alpha-defensin and results of frozen section were documented. Only patients with rheumatoid and psoriatic arthritis were included in the inflammatory arthritis group.

Results: Overall, 1220 patients undergoing revision TJA (567 PJI and 653 aseptic revisions) were included in the study. Fifty five patients (9.7%) in the septic and 61 patients (9.3%) in the aseptic group had inflammatory arthritis. While mean levels of serum CRP and synovial WBC in inflammatory arthritis patients were significantly higher compared to patients without inflammatory arthritis undergoing aseptic revisions (1.6 mg/dL and 1.9×10³ cells compared to 0.83 mg/dL and 1.1×10³ cells; p=0.006 and p=0.04), no significant difference were seen in patients with PJI. The thresholds associated with increased risk for PJI in patients with and without inflammatory arthritis were similar and closely resembled traditional cut-points; 32.0 mm/h and 31.5 mm/h for ESR, 1.3 mg/dL and 1.1 mg/dL for CRP, 2533 cells/µL and 2683 cells/µL for synovial fluid WBC, and 73% and 72% for synovial fluid PMN%. There were no significant differences in the sensitivity and specificity of the different tests using the new and traditional thresholds.

Conclusions: We demonstrate higher baseline immune upregulation in aseptic revision cases with inflammatory arthritis, but no significant differences are seen for PJI diagnosis. Conventional PJI thresholds for serum and synovial diagnostic markers should be adhered to. Assumptions about inflammatory arthritis patients needing differential diagnostic protocols should be avoided.