Limitations in the interpretation of rest-exercise ejection fraction changes after early thrombolytic therapy during acute myocardial infarction.