Background: Large vessel vasculitis (LVV), including giant cell arteritis (GCA) and Takayasu arteritis (TAK), is the most common primary vasculitis. Accurately determining disease activity in LVV is challenging, particularly once treatment has started, and can lead to both under-treatment, with resultant disease complications, and over-treatment, leading to adverse effects of toxic therapies. Hybrid PET/MR imaging utilises ~20% of the radiation dose associated with PET/CT and has potential as a diagnostic tool in LVV.1 Here, we evaluated, for the first time, the potential of PET/MR for tracking LVV disease activity and monitoring treatment-response. Methods: Patients aged ≥18 years with active LVV were recruited to this Scotland-wide prospective, observational study. The study consisted of 2 visits ≥6 months apart; at each visit clinical evaluation was followed by 18F-fluorodeoxyglucose (FDG) PET/MR scanning (Siemens 3T Biograph mMR). Images were obtained from the Circle of Willis to iliac arteries, and included gadolinium-enhanced MR angiography. PET/MR images were assessed for disease activity qualitatively and semi-quantitatively, and compared with clinical assessment of disease activity. Results: Thirty-nine PET/MR scans were performed in 24 patients with LVV (14 GCA, 6 TAK, 4 unspecified). Mean age was 61±15 years and 17 (71%) were female. Interpretation of PET/MR scans by a radiologist blinded to clinical details demonstrated a sensitivity of 77% and specificity of 88% for distinguishing active from inactive LVV. Investigator quantification of FDG uptake was then performed for each of 9 arterial segments on a scale of 0-3, and a cumulative PET Vasculitis Activity Score (PETVAS) calculated (0-27; higher scores indicate greater burden of disease).2 PETVAS was higher in active versus inactive disease (15.6±7.0 versus 8.8±4.2, P=0.001), and higher in GCA versus TAK (14.9±7.0 versus 8.6±3.4, P=0.01). ROC analysis of PETVAS performance yielded an AUC of 0.78 and a suggested cut-off score of 12 for distinguishing active from inactive disease (sensitivity 73%, specificity 76%). In those with >1 scan, PETVAS fell significantly from baseline to follow-up (18.2±6.4 versus 9.1±4.5, P=0.0001), reflecting a reduction in disease activity over time (Figure – baseline (A) and follow-up (B) PET/MR imaging). MR metrics, including mural enhancement and mural signal, also differed between active and inactive disease. We observed correlations between PETVAS and acute phase reactants, but not with patient-reported fatigue scores. Conclusions: PET/MR may be useful in tracking disease activity and assessing treatment response in patients with LVV. Based on our findings larger, prospective, multi-centre trials assessing the utility of PET/MR in LVV are now warranted. Disclosures: none