Comparison of [18 F]FDG PET/CT and MRI in the diagnosis of active osteomyelitis

OBJECTIVE: In diagnosing osteomyelitis (OM) both MRI and [18 F]FDG PET-CT proved to be accurate modalities. In anticipation of the advent of hybrid PET/MRI scanners we analyzed our patient group to give direction to future imaging strategies in patients with suspected OM. MATERIALS AND METHODS: In this retrospective study all patients of a tertiary referral center who underwent both an MRI and a PET for the diagnosis of OM were included. The results of those scans were evaluated using patient's histology, microbiological findings, and clinical/radiological follow-up. Additionally, ROC curve analysis of the SUVmax and the SUVmax ratio on the PET scans was performed. Two imaging strategies were simulated: first MRI followed by PET, or vice versa. RESULTS: Twenty-seven localizations in 26 patients were included. Both MRI and PET were shown to be accurate in our patients for the qualitative detection of OM. A cut-off value for the SUVmax of 3 gave optimal results (a specificity of 90 % with a sensitivity of 88 %). The SUVmax ratio gave a worse performance. The two simulated imaging strategies showed no difference in the final diagnosis in 20 out of 27 cases. Remarkably, 6 equivocal cases were all correctly diagnosed by the second modality, i.e., PET or MRI. CONCLUSION: Both MRI and [18 F]FDG PET were accurate in diagnosing OM in a tertiary referral hospital population. Simulation of imaging strategies showed that a combined sequential strategy was optimal. It seems preferable to use MRI as a primary imaging tool for uncomplicated unifocal cases, whereas in cases with (possible) multifocal disease or a contraindication for MRI, PET is preferred. This combined sequential strategy looks promising, but needs to be confirmed in a larger prospective study.