Prospective comparison of whole-body (18)F-FDG PET/CT and MRI of the spine in the diagnosis of haematogenous spondylodiscitis

Abstract

Purpose: To prospectively compare (18)F-FDG PET/CT and MRI in the diagnosis of haematogenous spondylodiscitis

Methods: The study included 26 patients (12 women, 14 men; mean age 59 ± 17 years) with clinical symptoms of infection of the spine. Patients who had had prior spinal surgery or any type of antibiotic therapy in the previous 3 months were excluded from the study. Whole-body PET/CT 60 min after injection of 4.07 MBq/kg of (18)F-FDG and an MRI scan of the spine was performed in all patients. SUVmax in an area surrounding the lesions with the suspicion of infection as well as a background SUVmean in a preserved area of the spine were calculated for quantification. Infection was diagnosed by microbiological documentation in cultures of image-guided spinal puncture fluid or blood. Infection was excluded if symptoms were absent without antimicrobial therapy during a follow-up of at least 6 months.

Results: Spondylodiscitis was confirmed in 18 of the 26 patients. Staphylococcus aureus was found in 8 patients, Mycobacterium tuberculosis in 4, Escherichia coli in 2 and other pathogens in 4. Of the remaining 8 patients, the diagnoses were degenerative spondyloarthropathy in 5 and vertebral fracture in 3. The sensitivity, specificity, and positive and negative predictive value were 83%, 88%, 94% and 70% for (18)F-FDG PET/CT, and 94%, 38%, 77% and 75% for MRI, respectively. The accuracies of (18)F-FDG PET/CT and MRI were similar (84% and 81%, respectively). The combination of (18)F-FDG PET/CT and MRI detected the infection in 100% of the patients with spondylodiscitis. (18)F-FDG uptake, quantified in terms of SUVmax corrected by the background SUVmean, was significantly higher in patients with spondylodiscitis than in those without infection (p < 0.001).

Conclusion: Due to its high specificity, (18)F-FDG PET/CT should be considered as a first-line imaging procedure in the diagnosis of spondylodiscitis. Quantification of uptake in terms of SUVmax was able to discriminate infection of the spine from other processes in this series of patients.