Introduction:
Numerous metrics derived from the cardiopulmonary exercise test (CPX) are associated with outcomes among patients with heart failure with reduced ejection fraction (HFrEF). However few studies have examined the independent prognostic value of all variables assessed simultaneously.
Purpose:
Retrospective analysis to describe the relationship between all CPX measures and the composite outcome of mortality, left ventricular assist device (LVAD), or cardiac transplant (CT).
Methods:
Patients (n= 1,201; 33% female; age= 55 ± 13 y) with a CPX between 1997 and 2010 and confirmed HFrEF (ejection fraction [EF] < 40%) were identified. Death data through 2011 was obtained from the National Death Index. The association with the composite endpoint was evaluated separately for 30 CPX measures with adjustment for age, gender, EF, and beta-blocker therapy using Cox regression. Forward stepwise Cox regression was performed to identify which of the CPX variables contribute the most to outcome prediction.
Results:
During a median follow-up of 3.75 years there were 576 (48%) events. When tested separately, nearly all CPX variables (except heart rate reserve/metabolic reserve and peak respiratory exchange ratio) were associated (p<0.05) with the composite endpoint. The top 5 predictors are shown in the Table. Stepwise Cox regression revealed that only % predicted peak oxygen uptake (VO
2
, Wald= 76.1), ventilatory power (peak systolic blood pressure/V
E
-VCO
2
slope, Wald= 58.0), and EF (Wald= 27.0) independently predicted outcomes.
Conclusion:
When considering all variables measured during a CPX test, % predicted peak VO
2
was the variable with the strongest independent association to outcomes in this cohort of patients with HFrEF. The % predicted peak VO
2
may represent a key variable in determining when to consider a patient for an LVAD or CT.
Effects of Dietary Inorganic Nitrate Supplementation on Exercise Performance in Patients With Heart Failure: Protocol for a Randomized, Placebo-Controlled, Cross-Over Trial (Preprint)
