IS IT POSSIBILE TO IDENTIFY EARLY “CULTURE-NEGATIVE” PERIPROSTHETIC JOINT INFECTIONS?

Aim Pathogen identification in periprosthetic joint infection (PJI) is a key factor for their treatment. According to literature, traditional diagnostic techniques fail in this goal in up to one third of cases. New molecular approaches, usually more expansive and time-consuming, have been proposed to improve microbiological diagnosis. Aim of this study is to identify peri-operative factors predictive of “culture-negative” (CN) PJI to select those patients in which modern techniques should be used from the beginning. Method We retrospectively evaluated the medical records of patients who had undergone revision surgery for septic failure from 2014 to July 2024. Demographic data, pre- and intra-operative results have been examined. In particular comorbidities and data from intra-operative synovial fluid (SF) examination have been considered. Continuous variables were described using the median and interquartile range (IQR), and categorical variables were described using counts and frequencies. The p values for the comparison of different subgroups were derived from t-test or Mann–Whitney test, according to data distribution. Statistical significance was set for a two-tailed p value ≤ 005. The receiver operating characteristic (ROC) curve was used to assess the best cut off for selected variables to discriminate between CN-PJI and PJI with microbiological confirmation. Results Data from 178 patients, of which 51 CN (29%), were recorded and analyzed. Among comorbidities, rheumatologic conditions showed a stronger correlation with CN PJI (p=0.018). On the other side, intra-operative synovial fluid examination highlighted higher level of glucose (p=0.003), lower level of LDH (p<0.001), white cell count (p=0.003) and polymorphonuclear neutrophils (PMN) (p<0.001) for CN-PJI. ROC allowed to identify an optimal threshold to discriminate between CN-PJI and PJI with microbiological confirmation of 2576.5 UI/l for SF-LDH, 29 mg/dl for SF-glucose and 89% for SF-PMN, all with an area under curve (AUC) above 0.7. Conclusions Patients suffering from rheumatologic conditions or with specific peculiarities in their intra-operative synovial fluid should be considered at higher risk for failure in microbiological identification with traditional methods, thus deserving further molecular techniques to improve the diagnostic sensitivity to facilitate the selection of a targeted antibiotic therapy.