Spondylodiscitis is a rare infectious entity that requires multimodal diagnostic procedures. We evaluated the diagnostic performance of 18F-FDG PET on suspected spondylodiscitis based on published literature.
We searched the PubMed and EMBASE for pertinent studies up to July 2013. We implemented a patient-based meta-analysis of diagnostic data for FDG PET (the index test) against clinical, laboratory, and/or radiologic evidence of disease (the reference standard). A bivariate analysis was implemented to account for variability beyond the threshold effect. The individual patient data analysis was used to assess confounding factors that moderate diagnostic performance.
Twelve studies provided the diagnostic data on FDG PET and spondylodiscitis, comprising 224 patients. The combined sensitivity across studies was 0.97 [95% confidence interval (CI), 0.83–1.00], the specificity was 0.88 (95% CI, 0.74–0.95), and the area under the curve was 0.98 (95% CI, 0.96–0.99). For prior probabilities greater than 0.50, the corresponding positive predictive value was 0.96 (0.93–0.98), and the negative predictive value was 0.85 (0.82–0.88). In the individual patient data analysis, metallic implants, dual PET/CT scanners and the addition of other imaging modalities to confirm disease were significant outcome moderators; only PET/CT remained significant in the adjusted analysis. PET/CT scanners improved the diagnostic performance, as opposed to the clinical data (age, sex, lesion site), which did not alter outcome.
FDG PET is a robust diagnostic test when spondylodiscitis is suspected and is excellent for exclusion of infectious spondylodiscitis given its low likelihood ratio negative (<0.1). Importantly, this diagnostic test is unaffected by other confounders, including the presence of implants, when PET/CT is used.
From the *Division of Infectious Diseases, Rhode Island Hospital and Warren Alpert Medical School of Brown University; †Department of Medicine, Warren Alpert Medical School of Brown University, Providence, RI; and ‡Department of Medical Imaging and Interventional Radiology, General Hospital of Chest Diseases “Sotiria,” Athens, Greece.
Received for publication September 6, 2013; revision accepted November 19, 2013.
Drs Ziakas and Prodromou have contributed equally to the study.
Conflicts of interest and sources of funding: none declared.
Reprints: Panayiotis D. Ziakas, MD, PhD, Division of Infectious Diseases, Rhode Island Hospital, 593 Eddy St, POB, 3rd Floor, Suite 328/330, Providence, RI 02903. E-mail: email@example.com.