Can we rely on histopathological results for the diagnosis of prosthetic joint infection?

ABSTRACT

BACKGROUND: Histopathology is one of the diagnostic criteria for prosthetic joint infection (PJI) proposed by all academic societies. The aim of this study was to compare histopathological and microbiological results from samples taken intraoperatively at the same site in patients with suspected or proven PJI. PATIENTS AND METHODS: We conducted a monocenter retrospective study including all patients having undergone surgery from 2007 to 2015 with suspected or proven PJI. During surgery, both histopathological and microbiological samples were taken. Patients with a history of antimicrobial treatment 2 weeks prior to surgery were excluded. We considered as major criteria and gold standard for PJI diagnosis the presence of a sinus tract communication and/or the same microorganism in at least two cultures. RESULTS: Finally, 181 patients who underwent 309 surgeries were included. The median number of samples per surgery was 4 (interquartile range (IQR) = 3-5) for histopathology and 5 (IQR = 4-6) for microbiology. Major criteria were observed in 177 patients (57.3%), while positive histology in at least one intraoperative sample was present in 119 (38.5%). The concordance was 74%. The sensitivity and specificity of histopathology were 61% and 92% respectively. Available "histopathology-culture" sample pairs numbered 1247. Among them, positive histopathology was found in 292 samples (23%) and culture in 563 (45%). Concordance was 64%. The highest correlation was observed for very early infection (<1 month) (OR 9.1, 95% CI 3.6-23) and for virulent microorganisms, such as Staphylococcus aureus (OR 7.8, 95% CI 5.2-11.8), Streptococci (OR7.8; 95% CI 4-15.2) or Enterobacterales (OR 7.4; 95% CI 4.2-13.1). CONCLUSION: Histopathologic examination is a valuable criterion for PJI diagnosis, but it may lack sensitivity for chronic infections or due to low-virulence pathogens.