Background and aim of the study: The potential for left ventricular hypertrophy regression and associated functional improvements may well be the underlying mechanism of results in general after valve replacement for aortic stenosis. The study aim was to investigate preoperative predictors and the time course of such ventricular changes.
Methods: Forty-six patients (mean age 61 years; range: 24-82 years) with aortic stenosis were prospectively followed with serial investigations (Doppler echocardiography, radionuclide ventriculography) at eight days (n = 43), three months (n = 42) and 18 months (n = 39) after valve replacement with a mechanical valve (19-29 mm). The postoperative course of left ventricular ejection fraction (EF), fast filling fraction, mass index and end-diastolic volume index (EDVi) was analyzed in an independent increments statistical model for repeated measurements.
Results: EF rose marginally, from 59+/-15% preoperatively to 64+/-16% (p <0.05) at 18 months, independently related to preoperative EF (p = 0.0001) and fast filling fraction (p = 0.0001). Changes in fast filling fraction were similarly predicted by the preoperative starting point (p = 0.003) and by preoperative left ventricular systolic radius:wall thickness ratio (p = 0.0002) with an inverse relation (the larger the chamber and the poorer its contractility, the lower the postoperative fast filling fraction). Mass index was independently related to the time point of postoperative measurement, indicating continuing regression of hypertrophy, from 200+/-66 g/m2 preoperatively to 148+/-49 g/m2 at 18 months (p <0.0001), when only 18% of the patients had normal mass index. Mass index was also independently related to preoperative left ventricular end-systolic dimension index (p = 0.0008) with a constant influence, and systolic wall stress (p = 0.0009) which was modified by time: the influence of wall stress was significant at eight days, weak at three months, and pronounced at 18 months. Left ventricular EDVi after surgery was inversely related to preoperative EF modified by time. Associated coronary artery disease, size of the prosthetic valve, and peak gradient across the valve (mean 15 mmHg; range: 7-26 mmHg at 18 months) did not influence any of the four target variables.
Conclusion: We conclude that left ventricular response to valve replacement for aortic stenosis is predictable. A significant reduction in hypertrophy occurs during the first 18 postoperative months, but to a normal ventricular mass in only a minority of patients. Insufficient regression of hypertrophy was related to indices of irreversible myocardial disease, which also prevented functional ventricular improvement despite successful valve replacement and a hemodynamically well functioning valve.