Electrocardiographic predictors of coronary microvascular dysfunction in patients with non-obstructive coronary artery disease: Utility of a novel T wave analysis program

Background: Coronary microvascular dysfunction (CMD) is linked to adverse cardiovascular events. Definitive diagnosis of CMD requires invasive provocative testing during angiography. We developed and tested a novel computerized T wave analysis tool to identify electrocardiographic signatures of CMD. Methods: 1552 patients underwent an invasive assessment of coronary microvascular function. Patients with interpretable pre-procedural ECGs were divided into 2 age and sex matched groups (n = 261 in each group, 75% female): normal microvascular function, CFR > 2.5 (CFR+), and abnormal microvascular function, CFR <= 2.5 (CFR-). ECGs were evaluated using a novel T wave program that quantified subtle changes in T wave morphology. Results: T wave repolarization parameters were significantly different between patients with normal and abnormal microvascular function. The top 3 features in males comprised of T wave area in V6 (CFR+: 10091.4 mV(2) vs. CFR-: 8152.3 mV(2), p < 0.05); T1 Y-center of gravity in lead II (CFR+: 17.8 mV vs. CFR-: 22.4, p < 0.005) and T Peak-T End in lead II (CFR+: 97.6 msec vs. CFR-: 91.1 msec, p < 0.05). These could identify the presence of an abnormal CFR with 74 +/- 0.2% accuracy. In females, the top 3 features were T wave right slope lead V6 (CFR+: -2489.1 mV/msec vs. CFR-: -2352.3 mV/msec, p < 0.005); Amplitude in V6 (CFR+: 190.4 mV vs. 172.7 mV, p = 0.05) and Y-center of gravity in lead V1 (CFR+: 33.3 vs. CFR-: 40.0, p = 0.001). These features could identify the presence of an abnormal CFR with 67 +/- 0.3% accuracy. Conclusion: Our data demonstrates that a computer-based repolarization measurement tool may identify electrocardiographic signatures of CMD. (C) 2015 Elsevier Ireland Ltd. All rights reserved.