Background: Diagnostic reliability of indexations of peak exercise ST-segment depression (deltaST) for heart rate reserve (HRi) or chronotropic reserve (CR) to identify significant coronary artery disease (CAD) by bicycle exercise testing has not been evaluated previously.
Methods: Upright bicycle exercise testing (25 W increment every 3 min) was performed in consecutive patients in primary prevention with at least one of the following criteria: history of exercise-induced chest discomfort and cardiovascular risk factors; overt peripheral arterial disease; type 2 diabetes associated with two or more additional cardiovascular risk factors. Coronary angiography was performed to define significant CAD (stenosis > or = 70% of the main coronary arteries or of their major branches, or isolated left main stenosis > or = 50%, or two or more stenoses 50-69%). Duke angina index was used to grade exercise-induced chest pain; deltaST, ST/HRi and ST/CR were calculated at peak exercise; three different criteria for the definition of inducible myocardial ischemia were tested versus significant CAD: peak deltaST > or =100 microV, ST/HRi > 1.69 microV/b/min or ST/CR > 1.76 microV/%.
Results: Of the study sample (n = 46), 40% had typical angina; during stress test 80% showed deltaST > or = 100 microV; 76% had ST/HRi > 1.69 microV/b/min; 62% had ST/CR >1.76 microV/%. Diagnostic accuracy of deltaST > or = 100 microV, of ST/HRi > 1.69 micro5V/b/min, and of ST/CR > 1.76 microV/% were 78%, 72%, and 89% respectively (p < 0.001 for the difference in diagnostic performance). ST/CR > 1.76 microV/% showed the highest diagnostic accuracy both in patients with submaximal exercise (96%) and in women (92%). Similarly, ST/CR >1.76 microV/% was associated with the highest diagnostic accuracy both in patients with maximal exercise (78%) and in men (88%). Analyses of the ROC curve revealed that ST/CR was associated with the greatest area under the curve, and a population-specific cut-off of 1.77 microV/% was associated with a sensitivity of 88% and a specificity of 90%.
Conclusions: Our pilot study suggests that in patients undergoing bicycle stress testing for differential diagnosis or screening of significant CAD, and with moderate-to-high pre-test probability, the use of ST/CR > 1.76 microV/% may provide elevated sensitivity and specificity, and the best diagnostic accuracy, which was consistent in patients with submaximal exercise test and in women.