To assess the added value of serial 2-deoxy-2-[18F]fluoro-D-glucose (FDG) uptake analysis in predicting clinical response to treatment in infectious spondylodiscitis (IS). We sought to analyze changes in quantitative FDG-PET/CT parameters among patients with clinical response or treatment failure and to compare the sensitivity and specificity of serial FDG-PET/CT and MRI in predicting treatment response in IS.
Materials and methods
This retrospective study consisted of 68 FDG-PET/CT examinations in 34 patients performed before and after at least 2 weeks of antibiotic treatment. Serial MRI scans were available in 32 (94%) patients before and after treatment. FDG-avid lesions were quantified as maximum standardized uptake value (SUVmax), partial-volume corrected lesion metabolic volume (LMV), and partial-volume corrected lesion metabolic activity (LMA).
All FDG-PET/CT parameters significantly decreased in patients with clinical improvement (31/34, 91%, P < 0.001), while patients with disease progression did not show FDG-PET/CT improvement. FDG uptake decrease was similar between patients undergoing early assessment (< 6 weeks) compared with those performing FDG-PET/CT after 6 weeks of treatment. SUVmax, LMV, and LMA decrease over time was 39.0%, 97.4%, and 97.1%, respectively. In predicting clinical responses, SUVmax reduction > 15% and > 25% showed 94% and 89% sensitivity and 67% and 100% specificity compared with 37% and 50% of MRI, respectively. Low degree of agreement with clinical response was shown for MRI compared with FDG-PET/CT parameters using the Cohen kappa coefficient.
FDG-PET/CT monitoring is a valuable tool to predict clinical response to treatment in IS and has greater sensitivity and specificity compared with MRI.