Response to Letter Regarding Article, “Early Surgery Versus Conventional Treatment in Asymptomatic Very Severe Aortic Stenosis”

Abstract

We are grateful to Meurin et al for their interest in our article, 1 and for raising some concerns. They suggested that the results of our study should not help to justify operations for which real benefit is not proved. There were 3 main issues raised in their letter.First, a registry should not be used to compare 2 therapeutic strategies, because too many biases can be involved; patients in the conventional treatment group probably had serious extra cardiac disease more often, as suggested by a higher rate of noncardiac deaths. We have clearly acknowledged nonrandomized assignment of treatment strategies as the major limitation of our study. To minimize selection bias and confounding, we performed propensity score matching, which has been shown to eliminate a greater proportion of baseline differences than stratification or covariate adjustment. For the 57 propensity score-matched pairs, the risk of all-cause mortality was persistently lower in the early surgery group than in the conventional treatment group (hazard ratio 0.135, 95% confidence interval 0.030 to 0.597, Pϭ0.008), and the risk of noncardiac death was similar between the two groups (hazard ratio 0.773, 95% confidence interval, 0.075 to 7.932, Pϭ0.825). Given that no randomized trial data are available, we suggest that the results of our study should be of clinical value to the decision of treatment strategy.Second, our study does not reflect "real life" practice, because baseline characteristics of our study patients were different from those reported in the Euro Heart Survey. 2 Contrary to the Euro Heart Survey, in which 84% of operated patients had preoperative symptoms, patients with any symptoms were excluded in our study, and the mean age of our study patients (63 years) is similar to that of the previous study (60 years), 3 which consecutively included asymptomatic patients only. The prevalence of coronary artery disease and bicuspid valve in clinical studies is also likely to be different according to age, presence or absence of symptoms, and severity of aortic stenosis. We feel strongly that our study patients represent asymptomatic patients with very severe aortic stenosis encountered in real clinical situations.The third concern focused on the absence of exercise testing. The study by Das et al 4 referenced in the letter reported that among 36 patients who developed symptoms, 72% had symptoms during exercise test, and 67% had very severe aortic stenosis. Further studies are needed to explore whether exercise test might have an additional prognostic value in very severe aortic stenosis.We agree that a prospective randomized comparison trial is required to confirm the results of our study, and look forward to working collaboratively for large multicenter prospective trials. DisclosuresNone. Duk