Background: Peak oxygen consumption (V̇o2) and minute ventilation (V̇e)/carbon dioxide production (V̇co2) slope have been widely demonstrated to have strong prognostic value in patients with heart failure (HF). In the present study, we investigated the effect of HF etiology on the prognostic applications of peak V̇o2 and V̇e/V̇co2 slope.
Methods: Two hundred sixty-eight subjects underwent symptom-limited cardiopulmonary exercise testing (CPX). The population was divided into ischemic (115 men and 22 women) and nonischemic (108 men and 23 women) subgroups. The occurrence of cardiac-related events over the year following CPX was compared between groups using receiver operating characteristic curve (ROC) analysis
Results: Mean age ± SD was significantly higher (61.0 ± 10.0 years vs 50.3 ± 16.2 years) while mean peak V̇o2 was significantly lower (15.0 ± 5.2 mL/kg/min vs 17.5 ± 6.7 mL/kg/min) in the ischemic HF group (p < 0.05). ROC curve analysis demonstrated that both peak V̇o2 and V̇e/V̇co2 slope were significant predictors of cardiac events in both the ischemic group (peak V̇o2, 0.74; V̇e/V̇co2 slope, 0.76; p < 0.001) and the nonischemic group (peak V̇o2, 0.75; V̇e/V̇co2 slope, 0.86; p < 0.001). Optimal prognostic threshold values for peak V̇o2 were 14.1 mL/kg/min and 14.6 mL/kg/min in the ischemic and nonischemic groups, respectively. Optimal prognostic threshold values for the V̇e/V̇co2 slope were 34.2 and 34.5 in the ischemic and nonischemic groups, respectively.
Conclusions: Baseline and exercise characteristics were different between ischemic and nonischemic patients with HF. However, the prognostic power of the major CPX variables was strikingly similar. Different prognostic classification schemes based on HF etiology may therefore not be necessary when analyzing CPX responses in clinical practice.