Diagnostic value of cardiac magnetic resonance imaging for myocardial infarction with non-obstructive coronary arteries: The CRIMINAL prospective registry.

Background: Numerous studies have investigated the use of cardiac magnetic resonance (CMR) imaging in patients with of myocardial infarction with non-obstructive coronary arteries (MINOCA).

Aims: To assess the diagnostic and prognostic yield of CMR in patients presenting with a working diagnosis of MINOCA-like condition.

Methods: Between May 2015 and August 2018, 457 consecutive patients with MINOCA-like condition were prospectively enrolled in an observational cohort study in 15 centres affiliated with the Collège National des Cardiologues des Hôpitaux (CNCH). Prior to CMR, the cardiologist established one provisional underlying mechanism. Using CMR findings, the cardiologist reassigned the patient to a final diagnosis.

Results: Overall, 336 of 457 patients underwent CMR. Guided by CMR, the cardiologist provided a diagnosis in 90.5% (95% confidence interval [95% CI]: 87.3-93.6%) of cases and changed the underlying cause in 35.7% (95% CI: 30.6-40.8%) of patients. Specifically, 33.7% (95% CI: 23.7-43.7%) of type 1 myocardial infarctions (MIs) were misdiagnosed prior to CMR. A type 1 MI occurred for three patients (at 3, 4 and 6months of follow-up) with a CMR diagnosis of non-ischaemic underlying cause (1.2%), which is reassuring for therapeutic management guided by CMR. Patients with an inconclusive CMR had a longer median time to complete the CMR (17 vs. 7days; P<0.01) and lower conventional and high-sensitivity troponin values (1.1 vs. 3.6μg/L; P=0.002 and 433 vs. 707ng/L; P=0.007, respectively).

Conclusion: The systematic use of CMR plays a pivotal role in identifying the underlying cause of myocardial injury, impacting patients, health professionals and society at large.