Large vessel giant cell arteritis in remission: Is imaging useful as predictor for relapses?

OBJECTIVE: To evaluate whether vasculitic findings (VF) on [¹⁸F] fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT) and/or magnetic resonance imaging (MRI) at treatment stop (TS) are associated with relapse in patients with large vessel giant cell arteritis (LV-GCA), and to explore the association between PET/CT-based polymyalgia rheumatica (PMR) findings and relapse. METHODS: LV-GCA patients included in our local GCA cohort BARK who discontinued treatment in remission and underwent PET/CT and/or MRI within ±4 months of TS were included. VF were defined using standardized criteria. Periarticular inflammation in PET/CT was scored by the Leuven score and a cut-off ≥12 was considered positive. Patients were followed for relapse. Statistics used Kaplan-Meier and Cox regression. RESULTS: VF at TS were observed in 54% of patients (25/46), including 47% of MRIs (15/32), 50% of PET/CTs (14/28), and 4/14 positive on both modalities. During follow-up, 48% (22/46) of patients relapsed in median 4.7 months after TS (82% minor, 18% major relapses). Overall, VF at TS was not associated with relapse (p = 0.1). PET/CT findings, including SUV segment/liver ratio and a modified PETVAS were not predictive for relapse (p ≥ 0.6). Absence of VF in MRI at TS was associated with relapse (p = 0.005). Six patients had positive Leuven score at TS and all relapsed (HR 5.35, 95% CI 1.70-16.83, p = 0.004). CONCLUSION: Vascular imaging at TS did not predict relapse in our cohort. There is a signal pointing towards a potential association of periarticular inflammation in PET/CT at TS to subsequent relapses, that warrants further validation.