A battery of diagnostic tests is often required to differentiate aseptic loosening from periprosthetic infection since the gold standard remains elusive. We designed a prospective study to determine the accuracy of fluorodeoxyglucose positron emission tomography (FDG-PET) imaging in diagnosing periprosthetic infection in a large multicenter setting. One hundred and thirteen patients with 127 painful hip prostheses were evaluated by FDG-PET. Images were considered positive for infection if PET demonstrated increased FDG activity at the bone-prosthesis interface of the femoral component. A combination of preoperative tests, intraoperative findings, histopathology, and clinical followup constituted the gold standard for diagnosing infection. Among the 35 positive PET scans, 28 hips were confirmed infected according to our criteria for diagnosing periprosthetic infection. Of the 92 hip prostheses with negative FDG-PET findings, 87 were considered aseptic. The sensitivity, specificity, positive and negative predictive values for FDG-PET were 0.85 (28 of 33), 0.93 (87 of 94), 0.80 (28 of 35), and 0.95 (87 of 92), respectively. The overall accuracy of this novel noninvasive imaging modality reached 0.91 (115 of 127). Based on our results, FDG-PET appears a promising and accurate diagnostic tool for distinguishing septic from aseptic painful hip prostheses.
Level of evidence: Level II, diagnostic study. See Guidelines for Authors for a complete description of levels of evidence.