Prognostic role of early cardiac magnetic resonance in myocardial infarction with non-obstructive coronary arteries

Abstract Background Myocardial infarction with non-obstructive coronary arteries (MINOCA) represents a significant proportion of acute myocardial infarction (AMI) population. MINOCA is a working diagnosis and an accurate investigation of the underlying causes should always be performed. In this setting, cardiac magnetic resonance (CMR) imaging plays a pivotal diagnostic role. However, a prognostic stratification based on the CMR findings in ischemic MINOCA is still unavailable. Purpose To evaluate the potential prognostic impact of specific CMR findings - especially ischemic late gadolinium enhancement (LGE) patterns - in order to look for measurable parameters that may guide the management of this still troubled clinical entity. Methods We assessed 461 MINOCA from January 2016 to June 2021. We excluded acute myocarditis, Tako-tsubo syndromes, cardiomyopathies, or non-pathological CMR. According to CMR findings, MINOCA were classified in two phenotypes: LGE-positive (an ischemic subendocardial or transmural LGE pattern) or LGE-negative (cases without LGE but exhibiting regional myocardial injury defined by myocardial edema in a coronary territory with a typically ischemic “wave-front” and/or regional wall motion abnormality consistent with coronary distribution). All-cause mortality, re-infarction, stroke, heart failure (HF) and the composite endpoint (MACE) were evaluated. Extended LGE was considered as the presence of >2 segments with transmural LGE. The mean follow-up was 36.1±15.2 months and CMR was performed at a mean of 6±2.9 days from the acute presentation. Results The final cohort included 175 MINOCA with a likely-ischemic etiology: 121 (69.1%) constituted the LGE-positive group. The mean age of the study population was 62.3±12.9 years and more than 61% were females. During follow-up, HF (15.7% vs 1.9%, p=0.008) and MACE (20.7% vs 7.4%, p=0.029) occurred more frequently in MINOCA “LGE-positive” compared to the “LGE-negative” ones. Extended LGE was significantly more prevalent in patients with versus without subsequent HF. On multivariable Cox regression, extended LGE was an independent predictor of HF occurrence (HR 18.49, 95% CI 4.65–73.61, p<0.001) and MACE (HR 14.64, 95% CI 3.91–54.86, p<0.001). Conclusions Our data suggest that in MINOCA patients the detection of LGE is correlated with the incidence of major cardiovascular events and heart failure during long-term follow-up. In fact, LGE extension was identified as the strongest predictor of cardiac adverse events. The early execution of CMR is useful in the prognostic stratification of MINOCA and this could guide the subsequent clinical and therapeutic management. Funding Acknowledgement Type of funding sources: None.