Left ventricular performance assessed by radionuclide angiocardiography and echocardiography in patients with previous myocardial infarction.

In 61 patients (77 studies) who had a transmural myocardial infarction, we compared the left ventricular ejection fraction by echocardiography with the ejection fraction determined by a computerized radioisotope technique that makes no assumptions regarding left ventricular geometry. In 31 studies of 26 patients with normal left ventricular wall motion by videotracking and normal left heart size, ejection fraction averaged 0.57 +/- 0.09 (SD) by ultrasound and 0.62 +/- 0.10 by the isotope method. Measurements of ejection fraction by both techniques correlated well (r = 0.86) and there was complete separation between patients with normal and reduced ejection fraction. In 46 studies of 35 patients in whom left ventricular wall motion abnormalities were recorded by videotracking, ejection fraction by the isotope method averaged 0.46 +/- 0.08, while average echo ejection fraction was 0.62 +/- 0.12. The correlation between the ultrasound and isotope methods in these 46 studies was poor (r = 0.33) and in 28 studies measurement of the ejection fraction by the two techniques was discordant. In 26 of the 27 studies where there was a reduced ejection fraction by the isotope method and a normal ejection fraction by echo, the dyssynergy involved the anterolateral left ventricular wall. These data indicate that echocardiographic measurements frequently overestimate left ventricular performance in patients with previous myocardial infarction associated with anterolateral wall motion disorders.