Chronic PJI remains the most common reason for revision after TJA and the incidence of TJA is projected to increase sharply from 2014 to 2030 [18]. This increase will impose huge medical and economic burdens on public health [19]. Chronic PJI is often caused by microorganisms with low virulence and has considerable delays in diagnosis. It may present with atypical symptoms different from those of acute infection, which are often similar to those of aseptic loosening [20]. Patients may develop a mild systemic response and have normal laboratory markers when PJI is present as a chronic encapsulated infection [5]. For these reasons, diagnosis of chronic PJI is often confusing.
Synovial fluid biomarkers that can be used to predict PJI have been reported previously and IL-1β was included in these studies [6-7, 15-16]. In fact, IL-1β is a multifunctional and highly potent pro-inflammatory cytokine [21] that was confirmed to be associated with bone resorption and osteoporosis in some inflammatory diseases [22]. Nicolas et al. [23] found that IL-1β played an important role in early control of the bacterial burden in post-surgical joints. Deirmengian et al. [6] evaluated 29 PJI cases and 66 aseptic joint cases, and described that the AUC for IL-1β was 0.966, with specificity of 95% (95% CI, 87%–99%) and sensitivity of 96% (95% CI, 82%,–00%). Frangiamore et al. [24] reported that the AUC, sensitivity, and specificity of IL-1β were 0.92, 90.3% (95% CI, 74%–98%), and 87% (95% CI, 76%–95%), respectively, with a decreasing trend in IL-1β between first-stage explantation and second-stage reimplantation. Gollwitzer et al. [16] found that IL-1β had sensitivity of 67% and specificity of 95% to distinguish aseptic loosening from staphylococcal infection. However, their study only focused on the diagnostic value of IL-1β for predicting PJI and did not explore the role of IL-1β in different types of PJI (acute or chronic). However, the inflammatory response in chronic PJI is very different from that in acute PJI [13]. In the present study on chronic PJI, synovial fluid IL-1β had sensitivity of 97.3% (95% CI, 85.8%–99.9%), specificity of 94.64% (95% CI, 85.1%–98.9%), and AUC of 0.991 (95% CI, 0.945–1.000) when a cutoff value of 312.7 pg/mL was used.
Elevated synovial fluid PMN% has been identified as a useful marker for diagnosis of PJI in previous studies. Trampuz et al. [25] showed that when the threshold of PMN% was >65%, the sensitivity and specificity for the diagnosis of PJI were 97% and 98%, respectively. Due to the synovial fluid composition changes with increasing postoperative time, the optimal cutoff values of PMN% for the diagnosis of acute and chronic PJI are different. During the 2013 International Consensus Meeting, the recommended cutoff value for diagnosis of acute PJI (<6 weeks after surgery) was above 90% and that for chronic PJI (>6 weeks after surgery) was above 80% [26]. Higuera et al. [27] demonstrated that the sensitivity, specificity, PPV, and NPV of PMN% above 80% for chronic hip PJI were 92.1%, 85.8%, 59.3%, and 98.0%, respectively. While in this study, when the optimal cutoff value of synovial fluid PMN% was 69.96%, the AUC, sensitivity, and specificity were 0.981 (95% CI, 0.928–0.998), 94.59% (95% CI, 81.8%–99.3%), and 92.86% (95% CI, 82.7%–98.0%), respectively.
Serum ESR and CRP have been recommended for first-line diagnostic evaluation in patients with suspected PJI by the American Academy of Orthopaedic Surgeons and the International Consensus Meeting [28]. Alijanipour et al. [29] showed that early postoperative and late chronic PJI had different thresholds of ESR and CRP, with the late chronic PJI values being higher than the early postoperative PJI values, in a retrospective review of 1962 patients who underwent revision arthroplasty for aseptic prosthetic failure (n=1689) or first onset of PJI (n=273) between 2000 and 2009. ESR >30 mm/h and CRP >10 mg/L were recommended as the optimal thresholds for diagnosis of chronic PJI in the 2013 International Consensus Meeting on Surgical Site and Periprosthetic Joint Infection [30].. In our study on chronic PJI, similar results were obtained again, the optimal cutoff values of serum ESR were 34 mm/h with specificity of 78.57% and sensitivity of 54.05% indicating the limited diagnostic value for chronic PJI. In the same way, the sensitivity and specificity of serum CRP were 89.19% (95% CI, 74.6%–97.0%) and 50% (95% CI, 36.3%–63.7%) for the diagnose of chronic PJI with a cutoff value of 13 mg/dL.
A single indicator cannot provide 100% diagnostic accuracy and in the presence of high clinical suspicion, a combination of tests should be used to refute or confirm the possibility of infection [31]. The present study explored diagnosis of chronic PJI using combinations of synovial fluid IL-1β and serum CRP or synovial fluid PMN%. When synovial fluid IL-1β was combined with serum CRP or synovial fluid PMN%, the specificity was 96.43% or 100%, respectively, while the sensitivity was decreased. This trade-off is typical for many tests. After comparing the two combinations, we found that the combination of synovial fluid IL-1β and PMN% had the highest accuracy for detecting chronic PJI. Furthermore, when both synovial fluid IL-1β and PMN% were above their thresholds of 312.7 pg/mL and 69.96%, respectively, the specificity, PPV, and accuracy reached 100%, 100%, and 96.77%, respectively. These results indicate that the combination is more accurate than each single index alone.