Differences in the prognostic relevance of myocardial ischaemia and scar by cardiac magnetic resonance in patients with and without diabetes mellitus.

AIMS To evaluate the prognostic significance of myocardial ischaemia and scar in patients with and without diabetes mellitus (DM) who undergo dobutamine stress cardiac magnetic resonance (DCMR) and late gadolinium enhancement (LGE) imaging for known and suspected coronary artery diseases (CADs). METHODS AND RESULTS A total of 1969 consecutive patients [age 63 ± 12 years, 29% female, left ventricular ejection fraction = 59 ± 12%] referred for a cardiac magnetic resonance (CMR) examination including DCMR and LGE with the suspicion of CAD or progression of CAD in three tertiary cardiac centres were analysed. Cardiac death and nonfatal myocardial infarction (MI) were registered as hard cardiac events. Patients with a revascularization procedure within the first 3 months after CMR were censored at the time of 'early' revascularization. Patients were followed for 3.2 ± 1.5 years (median 2.9, interquartile range 2-4.3 years). In total, 90 (4.6%) cardiac deaths and MI were registered. Among them, 328 patients (16.6%) had diabetes. The proportion of dobutamine-induced wall motion abnormalities (DWMA) and LGE was higher in patients with DM when compared with those without DM (27 vs. 19% and 53.6 vs. 41.2%, respectively, P < 0.001 for both for proportions). Both DWMA and LGE were independent predictors of cardiac death and MI in patients without DM (HR for DWMA 8, CI 4.5-14.3, HR for LGE 2.1, CI 1.1-4.1) and with DM (HR for DWMA 8.6, CI 3.5-21, HR for LGE 4.5, CI 1.5-13.1). Tests for interaction showed that LGE more strongly influences prognosis in patients with than in those without DM (P = 0.03 for interaction), whereas the presence of DWMA is related to similarly poor outcomes in patients with and without DM (P = NS). CONCLUSION Myocardial scar by LGE is a hallmark of markedly poorer outcome in patients with DM, while the presence of inducible myocardial ischaemia seems to be predictive both in patients with and without DM. Both markers surpass the predictive value of conventional atherogenic risk factors both in patients with and without DM.