Tc 99m labelled leukocytes scintigraphy (LLS) could be useful for the diagnosis of bone and joint infections. The aim of our study was to evaluate its performances specifically in the diagnosis of prosthetic joint infection (PJI).

We conducted a multicenter -7 year- retrospective study including 164 patients with suspected PJI who underwent surgical treatment. In each case, 5 intraoperative samples were taken. Diagnosis of infection was confirmed if two or more samples yielded the same microbial agent. LLS was considered as « positive » if an accumulation of leukocytes was observed in early stage and increased in late stage (24 hours). Among these patients, 123 had also a bone scintigraphy.

A total of 168 PJ were analyzed: 150 by in vitro polymorphonuclear labelled leukocytes scintigraphy (PLLS) and 18 by anti-granulocytes antibodies labelled leukocytes scintigraphy (LeukoScan®). Location of PJ were: hip (n = 63), knee (n = 71), miscellaneous (n = 4). According to microbiological criteria 62 hip prosthesis and 48 knee prosthesis were considered as infected. Sensitivity (Se), Specificity (Sp), Positive Predictive Value (PPV) and Negative Predictive Value of PLLS were: 72%, 60%, 80% and 47%. Se of LLS was higher for knee PJI (87%) than for hip PJI (57%) [p = 0.002]. Although Sp was higher for hip PJI (75%) than for knee PJI (52%) [p = 0.002]. The lowest Se was found for coagulase negative staphylococci (70%) and the highest for streptococci (87.5%). However the difference of Se between bacteria was not significant. Regarding bone scintigraphy, Se, Sp, PPV and NPV were: 94%, 11%, 65% and 50%.

In our study, performances of LLS were rather low and varied according to the location of infection. Differences of LLS Se between bacteria was not significant. Bone scintigraphy has a high Se but lacks Sp.