Sex differences in the diagnostic yield of CMR timing in patients with acute coronary syndromes and non-obstructive coronary arteries

Abstract Background Cardiovascular magnetic resonance imaging holds a Class I indication for the evaluation of suspected myocardial infarction with non-obstructive coronary arteries (MINOCA). Although the timing of CMR significantly influences its diagnostic yield, the extent to which this relationship is modulated by sex remains unclear. Current guidelines provide universal timing recommendations without considering potential sex-specific effects. Purpose To evaluate sex-specific timing interactions on CMR diagnostic probability in suspected MINOCA patients. Methods A total of 888 consecutive patients presenting with acute coronary syndrome and non-obstructed coronary arteries underwent comprehensive 1.5T CMR (51% female, mean age 57 ± 15.9 years), at a median of 20 days (IQR 6, 55 days). Based on CMR findings, patients were categorised into five diagnostic groups: myocardial infarction, myocarditis, cardiomyopathy, Takotsubo syndrome, and normal CMR. Logistic regression incorporating restricted cubic splines was used to model nonlinear interactions between sex and CMR timing in relation to diagnostic probability. Model performance was evaluated using the Akaike Information Criterion (AIC) and the concordance index (C-index). Results CMR identified underlying causes in 75.7% of patients (27.4% myocardial infarction, 24.4% myocarditis, 13.0% Takotsubo, 10.9% cardiomyopathy). Sex-specific diagnostic patterns were observed: males exhibited a significantly higher prevalence of myocarditis and cardiomyopathy compared to females (37.2% vs. 12.1%, respectively 15.4% vs. 6.6%), whereas females demonstrated significantly greater rates of Takotsubo syndrome and normal CMR findings (23.6% vs. 1.8%, respectively 29.1% vs. 19.3%, all p 0.001). No significant sex-based difference was found for myocardial infarction, a time-independent diagnosis (p =0.4). Time to CMR was 25 (IQR 6, 56) days in females and 17 (IQR 5, 54) days in males (p = 0.2). Adding sex as an interaction variable to the model explaining diagnostic yield as a function of time to CMR increased the model’s performance (AIC improvement: 29.1, C-index: 0.714) with statistically significant interaction (p = 0.022). Conclusions This study demonstrated that the prevalence of myocardial infarction in a large patients’ cohort presenting with MINOCA is similar between males and females. However, the prevalence of myocarditis and cardiomyopathies was higher in males, whilst the prevalence of Takotsubo syndrome and normal CMR was higher in females. Due to the higher relative prevalence of Takotsubo, prolonged wait times disadvantage females more than males, with the diagnostic yield of CMR decreasing steeply in the first twenty days from presentation. Tailoring timing strategies by sex may enhance diagnostic yield and optimise resource allocation in suspected MINOCA patients.