Objectives Myocardial perfusion reserve (MPR) is the ratio of myocardial blood flow (MBF) at stress to rest. A reduced MPR has been associated with a poor prognosis in quantitative Positron Emission Tomography studies. A likely mechanism is microvascular disease. Patients with diabetes mellitus often have microvascular disease and may have reduced MPR. We used automated in-line perfusion mapping, to quantify MBF at a pixel level in order to assess the MPR in patients with diabetes and other patients referred for clinical perfusion CMR.
Method Over 7 months, stress perfusion CMR with perfusion mapping was performed on 1201 clinically referred patients. Of these, we identified 121 who had also had angiography (invasive or CT) within 6 months (mean 6.4 weeks). Patients with unobstructed epicardial coronary arteries (<50% stenosis) were used in the final analysis (n=45). Global LV MPR was averaged across 3 short axis LV slice perfusion maps. The MPR of patients with diabetes (n=10) was compared to those without. Patient age, sex, body surface area (BSA), LV end-diastolic volume (EDV), ejection fraction (EF) and the presence or absence of hypertension and late gadolinium enhancement (LGE) were recorded. A multivariable analysis was performed to determine the contributions of these factors to the MPR.
Results Global LV MPR was: 3.07 across all patients, 2.33 for those with diabetes and 3.27 in those without diabetes (p=0.009). Multivariable analysis indicated that diabetes and age were negatively associated with MPR even after adjustment for sex, BSA, LGE, hypertension, LV EF and EDV (p<0.05 for each group).