Coronary Artery Calcium and Coronary Flow Reserve by Hybrid Rb-82 PET/CT Imaging in Non-diabetic and Diabetic Patients With Normal Myocardial Perfusion

Objectives We assessed the relationship between coronary artery calcium (CAC) and coronary flow reserve (CFR) in a propensity score-matched cohort of diabetic and non-diabetic patients with normal stress myocardial perfusion imaging.

Methods We evaluated 620 consecutive patients (470 non-diabetic and 150 diabetic patients) without a documented history of CAD and normal myocardial perfusion imaging at vasodilator stress Rb-82 PET/CT imaging. To account for differences in baseline characteristics between diabetics and non-diabetics patients, we created a propensity score-matched cohort considering clinical variables. After matching, clinical characteristics were comparable in 150 diabetic and 150 non-diabetic patients. The CAC score was measured according to the Agatston method. Resting and hyperemic myocardial blood flow (MBF) and CFR were automatically quantified. Patients were stratified into two groups based on their CAC score (0-400 and >400). The Spearman correlation coefficient was assessed between continuous CAC score and MBF or CFR values. Univariable and multivariable logistic regression analyses were used to determine the variables associated with reduced (<2) CFR.

Results In both non-diabetic and diabetic patients, CAC score was inversely correlated with hyperemic MBF (r=-0.21 and r=-0.25, respectively, p<0.01) and CFR (r -0.21 and r=-0.28, respectively, p<0.01). On the contrary, there was no correlation between CAC and baseline MBF in both non-diabetic and diabetic patients. Among non-diabetic patients, 150 (80%) had CAC score <400 and 30 (20%) CAC score >400, while in diabetic patients 110 (73%) had CAC <400 and 40 (27%) >400 (both p<0.05). CFR was reduced in 22 (18%) non-diabetic patients and in 38 (35%) diabetic patients with CAC <400 (p<0.01). In the overall population, at univariable analysis age (p<0.01), diabetes (p<0.001) and CAC score >400 (p<0.001) displayed were significant predictors of reduced CFR; at multivariable analysis only diabetes (p<0.001) and CAC score (p<0.05) were independently associated with reduced CFR. In non-diabetic patients, univariable analysis demonstrated that male gender and CAC score >400 (p<0.05) were significant predictors of reduced CFR; at multivariable analysis both male gender and CAC score >400 were independently related with reduced CFR (p<0.05). In diabetic patients, age (p<0.05) and CAC score (p<0.05) displayed an inverse relationship with CFR. However, at multivariate analysis neither age nor CAC score >400 resulted as independent predictor of reduced CFR in diabetic patients.

Conclusions After balancing clinical characteristics by propensity score analysis, diabetic patients with normal myocardial perfusion imaging, showed a higher prevalence of reduced CFR compared to non-diabetic patients, even in presence of CAC score values <400. In non-diabetic patients there is an inverse correlation between CAC score and both hyperemic MBF and CFR that remain also after adjusting for age and conventional cardiovascular risk factors. In diabetic patients this association disappeared after adjusting for conventional risk factors. A reduced CFR in diabetic patients without known CAD, could help to identify early structural alterations of the vessels.