Determinants of incomplete left ventricular mass regression following aortic valve replacement for aortic stenosis

Objective: Incomplete regression of left ventricular hypertrophy (Abn-LVMI) following AVR for aortic stenosis (AS) may decrease long-term survival. In this prospective study, we identified the predictors of Abn-LVMI. Methods: Between 1990 and 2000, 529 patients undergoing AVR for AS had clinical and hemodynamic data collected prospectively. Preoperative and annual postoperative transthoracic echos were employed to assess left ventricular mass index (LVMI) and hemodynamics. Abn-LVMI was defined as the 75th percentile of the lowest postoperative LVMI (> 128 mg/m(2), n = 133). All other patients were included in the normal regression group (N-LVMI). Univariate and multivariable logistic regression analyses were used to determine the predictors of Abn-LVMI. Results: Preoperative hypertension, diabetes, coronary disease, valve size, mean postoperative gradients, effective orifice area, and patient-prosthesis mismatch (PPM, indexed EOA < 0.60 cm(2)/m(2)) did not predict Abn-LVMI. By logistic regression the most important positive predictor of Abn-LVMI was the extent of preoperative LVMI, with an odds ratio of 37.5 (p < 0.0001). Survival (93.4 +/- 1.8 % vs 94.8 +/- 2.3 %, P = 0.90) and freedom from NYHA III-IV (75.0 +/- 3.7 % vs 76.6 +/- 5.3 %, p = 0.60) were similar for both groups at 7 years. Conclusions: Measures of valve hemodynamics were not important predictors of incomplete regression of hypertrophy. The extent of preoperative hypertrophy was the most important predictor, suggesting that earlier surgical intervention may reduce the extent of hypertrophy postoperatively. Furthermore, the significance of LV hypertrophy to long-term survival must be reassessed, in the absence of scientific evidence.