Echocardiographic prediction of clinical outcome in medically treated patients with aortic stenosis

Background The onset of symptoms is crucial in the natural history of sever e aortic stenosis. In contrast, the impact of the degree of valve obstructi on and left ventricular dysfunction on clinical outcome in terms of progres sion of symptoms and mortality is undefined. Methods and Results Between April 1989 and June 1996, 108 patients (75% mal e, aged 68.7 +/- 10.3 years) with pure and isolated aortic stenosis of at l east moderate degree had a complete Doppler echocardiography. Left ventricu lar end-diastolic and end-systolic diameters, thickness of ventricular sept um, mass and election fraction, and maxima[ instantaneous aortic gradient w ere measured. Patients were followed up through March 1999. Sixty-five pati ents who underwent aortic valve replacement were censored at the time of su rgery. The end point was considered to be death or symptomatic progression (onset of new symptoms or worsening of symptoms). At the time of diagnosis 64 (59%) were in New York Heart Association (NYHA) class I-Il and 44 (41%) in NYHA class Ill-IV. After a mean follow-up of 46 +/- 21 months 6 patients died and 45 had worsening of symptoms. Univariate predictors of clinical o utcome (death and worsening of symptoms) included left ventricular end-dias tolic diameter (hazard ratio 1.03, P = .08), left ventricular end-systolic diameter (HR 1.04, P = .012), and left ventricular septum thickness (HR 1.1 4, P = .009) but not the degree of aortic obstruction. Multivariate predict ors of clinical outcome were left ventricular septum thickness (P = .016) a nd left ventricular end-systolic diameter (P = .008). Conclusion In patients with various degrees of aortic stenosis the rate of clinical outcome is predicted by left ventricular function and septum thick ness. Therefore both the left ventricular and aortic valve gradients should be taken into account when choosing the timing of intervention.