Author(s): Timothy Chryssikos*, Elie Ghanem, Javad Parvizi, Andrew Newberg, Hongming Zhuang, Abass Alavi; Rothman Institute,Philadelphia, PA
Title: FDG-PET imaging in the differentiation of infected from non-infected painful hip prosthesis
Purpose: The accurate differentiation of aseptic loosening from periprosthetic infection in the painful hip prosthesis is a major clinical challenge. This prospective study was designed to determine the efficacy of FDGPET imaging for this purpose.
Methods: One hundred and thirteen patients with 127 painful hip prostheses were evaluated by FDGPET.Approximately 60 minutes after the intravenous administration of FDG images of the lower extremities were acquired using a dedicated PET machine. FDG-PET images were interpreted by experienced nuclear medicine physicians. Images were considered positive for infection if PET demonstrated increased FDG activity at the bone-prosthesis interface of the femoral component of the prosthesis. Surgical findings, histopathology, and clinical follow-up served as the gold standard.
Results: FDG-PET was positive for infection in 35 hips and negative in 92 hips. Among 35 positive PET studies, 28 were proven to be infected by surgical and histopathology findings as well as follow up tests. In 92 hip prostheses with negative FDG-PET findings, 87 were proven to be aseptic. The sensitivity, specificity, positive and negative predictive values for FDG-PET were 0.85 (28/33), 0.93 (87/94), 0.80 (28/35), and 0.95 (87/92), respectively. The overall accuracy of FDG-PET in this clinical setting was 90.5% (115/127).
Discussion: The results demonstrate that FDG-PET is a highly accurate diagnostic test for differentiating infected from non-infected painful hip prosthesis. Therefore, FDG-PET imaging is considered the study of choice in the evaluation of patients with suspected hip prosthesis infection.