Longitudinal cardiac magnetic resonance imaging following clinical response to rilonacept and prior to recurrence upon treatment suspension: a RHAPSODY subgroup analysis.

Abstract

In the phase 3 trial, RHAPSODY, rilonacept effectively resolved active pericarditis recurrences, and long-term treatment led to sustained pericarditis recurrence risk reduction. Prior analysis suggested association between higher late gadolinium enhancement (LGE) at baseline and more rapid recurrence upon rilonacept suspension after 12 weeks of treatment. This subgroup analysis assessed the utility of longitudinal serial cardiac magnetic resonance (CMR) imaging for tracking clinical improvement and predicting post-treatment-cessation outcomes to help guide clinical decision making. At an 18-month decision milestone (18MDM) in the RHAPSODY long-term extension, investigators decided if patients would continue rilonacept, suspend rilonacept for off-treatment observation, or discontinue the study. Pericardial thickness, pericardial edema (T2-STIR), and LGE were determined at baseline and 18MDM by an imaging core lab blinded to clinical data, and pericarditis recurrence was investigator-assessed. CMR results in patients with data at both baseline and 18MDM (n=13) showed that pericardial thickness, T2-STIR, and LGE were reduced during rilonacept treatment. Among patients with CMR data who suspended rilonacept at the 18MDM (n=7), 5 (71%) had a pericarditis recurrence within 1-4 months of rilonacept suspension, despite all having had none/trace LGE (n=7) and negative T2-STIR (n=7) at the 18MDM and 2 having received prophylactic colchicine. Continued clinical improvement during prolonged rilonacept treatment corresponded with improvement on CMR, including reduced pericardial thickness, resolution of pericardial edema, and resolution of LGE. However, none/trace LGE at 18MDM while on treatment did not predict absence of pericarditis recurrence upon subsequent rilonacept suspension in this size-limited subgroup.