Predicting cardiac mortality after uncomplicated myocardial infarction by exercise radionuclide ventriculography and exercise-induced ST segment elevation.

In 183 consecutive patients with recent, uncomplicated myocardial infarction, the following variables were associated with 4-year cardiac death: haemodynamic decompensation with exercise (P = 0.01), left ventricular ejection fraction at rest (P = 0.004) and at peak exercise (P = 0.003), persistent ST segment elevation at rest in the area of infarction = (P = 0.004), exercise-induced ST segment elevation (P = 0.02), and late aneurysmal evolution (P = 0.01). Exercise left ventricular ejection fraction was the sole variable selected by Cox regression analysis as an independent predictor of cardiac death. In 40 patients with ST segment elevation at rest, left ventricular ejection fraction was 42 +/- 17% at rest and 40 +/- 18% at peak exercise, versus 52 +/- 12% and 52 +/- 14% in the remaining patients (both P less than 0.01). Among these 40, 16 (all with anterior infarction) also had exercise-induced ST segment elevation; their ejection fraction was 32 +/- 13% at rest, 30 +/- 13% during exercise, versus 53 +/- 15% and 53 +/- 15% in 129 patients with no ST segment elevation either at rest, or during exercise (both P less than 0.01). The 4-year risk of death was 20% in the former 40 patients, 36% in the latter 16, while in the complete absence of ST segment elevation, such risk was 3%. All 14 patients with ST segment elevation only during exercise were alive after 4 years: their left ventricular ejection fraction was 47 +/- 12% at rest, 45 +/- 13% with exercise. ST segment elevation was associated with late aneurysmal evolution but not with exercise-induced ischaemia.(ABSTRACT TRUNCATED AT 250 WORDS)