Use of imaging and clinical data to screen for cardiovascular disease in asymptomatic diabetics

Carlos Henrique Reis Esselin Rassi, Timothy W. Churchill, Carlos A. Fernandes Tavares, Mateus Guimaraes Fahel, Fabricia P. O. Rassi, Augusto H. Uchida, Bernardo L. Wajchenberg, Antonio C. Lerario, Edward Hulten, Khurram Nasir, Márcio S. Bittencourt, Carlos Eduardo Rochitte, Ron Blankstein

There is increasing evidence to suggest that not all individuals with type 2 diabetes mellitus (T2DM) have equal risk for developing cardiovascular disease. We sought to compare the yield of testing for pre-clinical atherosclerosis with various approaches. 98 asymptomatic individuals with T2DM without known coronary artery disease (CAD) were enrolled in a prospective study and underwent carotid ultrasound, exercise treadmill testing (ETT), coronary artery calcium (CAC) scoring, and coronary computed tomography angiography (CTA). Of 98 subjects (average age 55 ± 6, 64 % female), 43 (44 %) had coronary plaque detectable on CTA, and 38 (39 %) had CAC score >0. By CTA, 16 (16 %) had coronary stenosis ≥50 %, including three subjects with CAC = 0. Subjects with coronary plaque had greater prevalence of carotid plaque (58 % vs. 38 %, p = 0.01) and greater carotid intima media thickness (0.80 ± 0.20 mm vs. 0.70 ± 0.11 mm, p = 0.02). Notably, 18 of 55 subjects (33 %) with normal CTA had carotid plaque. Eight subjects had a positive ETT, of whom five had ≥ 50 % coronary stenosis, two had <50 % stenosis, and one had no CAD. Among these tests, CAC scoring had the highest sensitivity and specificity for prediction of CAD. Among asymptomatic subjects with T2DM, a majority (56 %) had no CAD by CTA. When compared to CTA, CAC was the most accurate screening modality for detection of CAD, while ETT and carotid ultrasound were less sensitive and specific. However, 33 % of subjects with normal coronary CTA had carotid plaque, suggesting that screening for carotid plaque might better characterize stroke risk in such patients.