Exercise testing in patients with variant angina: results, correlation with clinical and angiographic features and prognostic significance.

Eighty-two patients with variant angina underwent a treadmill exercise test using 14 ECG leads, and 67 also underwent exercise thallium-201 scans. The test induced ST elevation in 25 patients (30%), ST depression in 21 (26%) and no ST-segment abnormality in 36 (44%). ST elevation during exercise occurred in the same ECG leads as during spontaneous attacks at rest, and was always associated with a large perfusion defect on the exercise thallium scan. In contrast, exercise-induced ST depression often did not occur in the leads that exhibited ST elevation during episodes at rest. The ST-segment response to exercise did not accurately predict coronary anatomy: Coronary stenoses greater than or equal to 70% were present in 14 of 25 patients (56%) with ST elevation, in 13 of 21 (62%) with ST depression and in 14 of 36 (39%) with no ST-segment abnormality (NS). However, the degree of disease activity did correlate with the result of the exercise test: ST elevation occurred during exercise in 11 of 14 patients who had an average of more than two spontaneous attacks per day, in 12 of 24 who had between two attacks per day and two per week, and in only two of 31 who had fewer than two attacks per week (p less than 0.005). St elevation during exercise was reproducible in five of five patients retested during an active phase of their disease, but not in three of three patients who had been angina-free for at least 1 month before the repeat test. Twelve patients wih exercise-induced ST elevation were retested during treatment with calcium antagonist drugs; in 10 of 12, ST elevation did not occur with the second test. During a mean follow-up of 20.3 +/- 14.5 months, death or myocardial infarction occurred in three of the 25 patients with ST elevation during exercise, none of 21 with ST depression and two of 36 with no ST abnormality. We conclude that in variant angina patients, the results of an exercise test correlate well with the degree of disease activity but not with coronary anatomy, and do not define a high-risk subgroup.