18F-FDG PET/MRI and PET/CT for muscle and malignancy assessment in dermatomyositis

Objectives Dermatomyositis (DM) requires advanced imaging to assess muscle inflammation and timely malignancy screening. Hybrid positron emission tomography (PET) imaging combines metabolic and anatomical data; however, whether it can comprehensively evaluate DM is unclear. We evaluated the ability of 18 F-fluorodeoxyglucose PET/magnetic resonance imaging (MRI) and PET/computed tomography (CT) to detect muscle involvement and cancer-associated myositis (CAM) in DM. Methods This retrospective study evaluated 41 patients, including 39 with classic DM and 2 with clinically amyopathic DM (CADM), who underwent PET/MRI (n=27) or PET/CT (n=14). Whole-body MRI was scored for muscle edema, fatty infiltration and atrophy. FDG uptake was quantified using region-based maximum standardized uptake value (SUV max ) ratios (muscle-to-blood pool) and Deauville scores. Imaging parameters (edema scores, SUV max ratios, Deauville scores) and clinical data (muscle enzymes and myositis-specific autoantibodies) were analyzed. Results The sensitivity of PET/MRI was higher than that of PET/CT for detecting muscle involvement in classic DM (100% [25/25] vs. 78.6% [11/14]; P = 0.04). No muscle involvement was detected in CADM. Total edema scores correlated with creatine kinase (ρ=0.505, 95% CI 0.147–0.747; P<0.01) and lactate dehydrogenase (ρ=0.723, 95% CI 0.466–0.867; P<0.0001). Ten patients fulfilled the temporal definition of CAM (10/41, 24.4%), with nasopharyngeal carcinoma being the most common subtype (4/10, 40.0%). Stratified by modality, 2 malignancies were observed in the PET/MRI cohort (2/27, 7.4%), and 8 malignancies were identified in the PET/CT cohort (8/14, 57.1%). Anti-transcription intermediary factor-1γ antibody grade correlated with the SUV max ratio (ρ=0.452, 95% CI 0.168–0.667; P<0.01). Conclusions PET/MRI detected muscle involvement more frequently than PET/CT in classic DM, while both PET/MRI and PET/CT contributed to malignancy assessment in this cohort. Hybrid PET imaging may aid disease phenotyping and cancer surveillance in DM.