Fifty consecutive patients undergoing aortic valve replacement for isolated aortic regurgitation were studied prospectively by echocardiography, electrocardiography and cardiac catheterization. Good quality echocardiograms were obtained in 49 of the 50 patients. Left ventricular (LV) dilatation was present in all 49 patients. LV systolic function, as assessed by echocardiographic percent fractional shortening, was normal in many patients but was moderately to severely reduced (less than 25%) in 14 patients (29%). Echocardiographic studies 6 months postoperatively revealed significant reductions in LV end-diastolic dimension (73.8 mm vs 58.7 mm; p less than 0.01), and serial echocardiographic studies early and late after operation revealed that the decrease in LV size had occurred by the time of the early study (8-22 days postoperatively), with little additional change thereafter. Operative deaths occurred in three of the 49 patients (6%). Eight of the 49 patients (16%) died of congestive heart failure (CHF) after hospital discharge at times ranging from 5-43 months after operation. Preoperative echocardiographic measurements of the LV end-systolic dimension and percent fractional shortening were strongly associated (p less than 0.01) with these late CHF deaths. Preoperative LV end-systolic dimension greater than 55 mm and fractional shortening less than 25% identified the high-risk group: nine of 13 patients (69%) in this group died either at operation or subsequently from CHF. In contrast, of 32 patients with LV end-systolic dimension less than 55 mm, only one died at operation and one died late from CHF. Thus, the population at high risk of late death from CHF was identified before operation by echocardiography.

In most patients with aortic regurgitation, valve replacement results in reduction in left ventricular dilatation and an increase in ejection fraction. To determine the relation between serial changes in ventricular dilatation and changes in ejection fraction, we studied 61 patients with chronic severe aortic regurgitation by echocardiography and radionuclide angiography before, 6-8 months after, and 3-7 years after aortic valve replacement. Between preoperative and early postoperative studies, left ventricular end-diastolic dimension decreased (from 75 +/- 6 to 56 +/- 9 mm, p less than 0.001), peak systolic wall stress decreased (from 247 +/- 50 to 163 +/- 42 dynes x 10(3)/cm2), and ejection fraction increased (from 43 +/- 9% to 51 +/- 16%, p less than 0.001). Between early and late postoperative studies, diastolic dimension and peak systolic wall stress did not change, but ejection fraction increased further (to 56 +/- 19%, p less than 0.001). The increase in ejection fraction correlated with magnitude of reduction in diastolic dimension between preoperative and early postoperative studies (r = 0.63), between early and late postoperative studies (r = 0.54), and between preoperative and late postoperative studies (r = 0.69). Late increases in ejection fraction usually represented the continuation of an initial increase occurring early after operation. Thus, short-term and long-term improvement in left ventricular systolic function after operation is related significantly to the early reduction in left ventricular dilatation arising from correction of left ventricular volume overload. Moreover, late improvement in ejection fraction occurs commonly in patients with an early increase in ejection fraction after valve replacement but is unlikely to occur in patients with no change in ejection fraction during the first 6 months after operation.

Our results indicate that preoperative left ventricular systolic function continues to influence postoperative prognosis and suggest that the discordant conclusions of previous studies probably resulted from interpretive differences and differences in patient selection.

No abstract

No abstract