Mitral systolic velocity at peak exercise predicts impaired exercise capacity in patients with heart failure with preserved ejection fraction

Background: Nearly half of patients with heart failure have normal left ventricular ejection fraction (LVEF), but their prognosis is no better than those with reduced LVEF. Although peak oxygen consumption (VO2) is an independent predictor of mortality in heart failure, it is unclear how cardiac function during exercise contributes to peak VO2. Therefore, we explored the useful parameters measured by exercise stress echocardiography to predict peak VO2 in patients with heart failure with preserved LVEF (HFpEF). Methods and Results: We assessed 80 patients being investigated for effort intolerance or dyspnea and finally analyzed 50 patients who satisfied the HFpEF criteria. Mean peak VO2 was 16.4 +/- 2.8 mL/kg/min. Twenty-three patients (46.0%) achieved a peak VO2 < 16.0 mL/kg/min (Weber class C or D). There was a significant relationship between mitral systolic velocity (S') and cardiac output (CO) at rest (R=.55, P <.0001) and peak exercise (R=.64, P <.0001). The absolute increase in S' from rest to peak exercise also correlated with the absolute increase in CO (R=.32, P=.02). Multivariate logistic regression analysis showed that S' at peak exercise independently predicted peak VO2. Receiver-operator characteristic curve analysis identified that an S' at peak exercise of <= 8.13 cm/s predicted a peak VO2 < 16.0 mL/kg/min (sensitivity 95.7%, specificity 44.4%, area under curve 0.70, 95% confidence interval 0.55-0.84, P=.004). Conclusions: Mitral systolic velocity at peak exercise accurately reflects peak VO2 and may facilitate stratification of risk in patients with HFpEF.