ABSTRACT
Aims: To determine the utility of serial cardiac magnetic resonance (CMR) imaging for guidance of therapy management in patients treated with anakinra due to recurrent pericarditis (RP), compared with C-reactive protein (CRP) assay alone. Methods and results: In 2018-21, we enrolled 18 (14.5 ± 1.8 years old, 72% males) consecutive RP patients treated with anakinra (100 mg/day in patients ≥ 18 years old; 2 mg/kg/day < 18 years old) due to RP corticosteroid-dependent or not responsive to colchicine or non-steroidal anti-inflammatory drugs. After hospitalization, they were 1:1 randomized to CMR [no pericardial oedema and/or late gadolinium enchantment (LGE)] or CRP (<0.6 mg/dL). Tests were repeated every 3 months until negative to halve the anakinra dosage and cessation. The idiopathic aetiology was the most prevalent (n = 8, 44%), followed by post-pericardiotomy (n = 6, 33%). After a median treatment period of 8.7 ± 3.6 months, CRP-guided RP patients experienced more recurrences than CMR-guided ones (6 vs. 1, P = 0.016), with the worst prognosis in terms of recurrences (log-rank, P = 0.025) and significantly increased time of treatment (12.7 ± 2 vs. 16.1 ± 3.4 months, P = 0.019). In a multivariable exploratory Cox regression model, the number of previous recurrences and the idiopathic aetiology were independent predictors of RP during the anakinra treatment. New recurrences were subsequently directed to CMR imaging, and therapy was modified according to the LGE/oedema trend. After 1-year follow-up, no further recurrence was detected. Conclusion: Among patients with RP and treated with anakinra, serial CMR imaging of the pericardium can be utilized as an imaging biomarker, more informative for therapy duration than the solely CRP assessment. ClinicalTrialsgov Identifier: NCT06071156.