Against the Norm: Do Not Rely on Serum C-Reactive Protein and White Blood Cell Count Only When Assessing Eradication of Periprosthetic Joint Infection.

Abstract

Introduction: Periprosthetic joint infections (PJI) following primary arthroplasty continue to be a serious complication, despite advances in diagnostics and treatment. Two-stage revision arthroplasty has been commonly used as the gold standard for the treatment of PJI. However, much discussion persists regarding the interim of the two-stage procedure and the optimal timing of reimplantation. Serology markers have been proposed as defining parameters for a successful reimplantation. The objective of this matched-pair analysis was to assess the role of serum C-reactive protein (CRP) and white blood cell count (WBC) in determining infection eradication and proper timing of reimplantation. We investigated the delta (∆) change in CRP and WBC values prior to both stages of two-stage revision arthroplasty as a useful marker of infection eradication. Methods: We analyzed 39 patients and 39 controls, matched by propensity score matching (BMI, age, ASA-classification), with a minimum follow-up of 24 months and treated with a two-stage revision THA or TKA in our institution. Data of serum CRP and WBC values were gathered at two selected time points: prior to the explantation of the implant (preexplantation) and following the completion of antibiotic treatment regimen, both systemic and with a drug-eluting cement spacer (prereimplantation). Patient records were reviewed electronically for preexisting comorbidities, overall health status, synovial fluid cultures, inflammatory serologies, revision surgeries, and recurrent or persistent infection based on the modified Musculoskeletal Infection Society criteria. Patient demographics, ∆CRP, ∆WBC, and time interval to reimplantation were statistically analyzed using receiver operator curves (ROC), Pearson’s correlation coefficient, Levene’s test, and Student’s t-test. Results: Infection-free patients exhibited higher mean CRP and WBC than did patients who were reinfected at both time points. When comparing preexplantation with prereimplantation values, the median ∆CRP was 9.48 mg/L (interquartile range (IQR) 2.3–36.6 mg/L) for patients who did not develop a reinfection versus 2.74 mg/L (IQR 1.4–14.2 mg/L) for patients who developed reinfection (p = 0.069). The median ∆WBC was 1.5 × 109/L (IQR 0.6–4.0 × 109/L) for patients who remained infection-free versus 1.2 × 109/L (IQR 0.8–2.2 109/L) for patients who developed reinfection (p = 0.072). Analysis of areas under the curve (AUC) using ROC demonstrated poor prediction of persistent infection by ∆CRP (AUC = 0.654) and ∆WBC (AUC = 0.573). Although a highly significant correlation was found between the interim interval and infection persistence (r = 0.655, p < 0.01), analysis using ROC failed to result in a specific threshold time to reimplantation above which patients are at significantly higher risk for reinfection (AUC = 0.507). Conclusion: No association could be determined between the delta change in serum CRP and WBC before and after two-stage revision arthroplasty for PJI and reinfection risk. Even though inflammatory serologies demonstrate a downtrending pattern prior to reimplantation, the role of CRP and WBC in determining the optimal timing of reimplantation seems to be dispensable. Planning a second-stage reimplantation requires assessing multiple variables rather than relying on specific numeric changes in these inflammatory marker values.