We have read the editorial by Baez-Ferrer et al.1 in Revista Española de Cardiología on cardiopulmonary exercise testing (CPET) in asymptomatic severe aortic stenosis (AS). The authors highlight the value of unmasking symptoms in patients with apparently asymptomatic severe AS during CPET by using parameters such as peak oxygen consumption (pVO2)2 and the slope comprising minute ventilation to CO2 production ratio (VE/VCO2 slope)3 abnormalities. This approach would lead to aortic valve replacement (AVR) and a better prognosis for patients with severe asymptomatic AS.2–4 However, our experience differs in this respect.5

We studied a cohort of 113 patients with AS undergoing CPET.5 The classic established criteria were applied to determine whether the test was positive.6 Parameters related to gas analysis were determined, but were not criteria for stopping the test. During CPET, the presence of symptoms differed by group, which had been established according to the severity of AS (P <.05); however, no differences were found in pVO2 or in the VE/VCO2 slope. According to pVO2, all patients were in Weber et al.7 class I or II. There were no differences in pVO2 regarding the presence or absence of symptoms during CPET or according to the type of AS, with values well above the pVO2 ≤ 14mL/kg/min4 cutoff used to indicate disease. According to the VE/VCO2 slope, all patients were in ventilatory class I. The foregoing aspects led us to think that the worst functional class would not only depend on the severity of the valvular heart disease, but—as the authors themselves recognize—on comorbidities such as aging, frailty, obesity, or others.2,4

Studies supporting the prognostic value of pVO2 and VE/VCO2 slope have been based on small population samples with few events, which makes it difficult to adequately control the covariables in the survival analysis and to ignore other prognostic variables. In previous series, it is striking that a predicted VO2 value of <80% in patients with asymptomatic severe AS had a stronger relationship with the presence of coronary artery disease or atrial fibrillation than with aortic valve area.2 Other authors have suggested that blood BNP level is a stronger predictor of revealed symptoms than pVO2 during CPET.8 The widespread adoption of a therapeutic strategy of AVR in patients with severe asymptomatic AS based on the results of CPET could increase the number of patients treated with this strategy. Such an approach, in the setting of an increasing prevalence of degenerative AS, would lead to an increase in both the burden of care and that of health care costs. On the other hand, the latter aspect would not necessarily lead to better health, as other authors have already suggested.9 We believe that studies with a larger population of patients with severe asymptomatic AS are needed to guarantee the robustness of the prognostic variables described in the setting of CPET.2–4