Search

Endpoint selection is a critically important step in clinical trial design. It poses major challenges for investigators, regulators, and study sponsors, and it also has important clinical and practical implications for physicians and patients. Clinical outcomes of interest in heart failure trials include all-cause mortality, cause-specific mortality, relevant non-fatal morbidity (e.g., all-cause and cause-specific hospitalization), composites capturing both morbidity and mortality, safety, symptoms, functional capacity, and patient-reported outcomes. Each of these endpoints has strengths and weaknesses that create controversies regarding which is most appropriate in terms of clinical importance, sensitivity, reliability, and consistency. Not surprisingly, a lack of consensus exists within the scientific community regarding the optimal endpoint(s) for both acute and chronic heart failure trials. In an effort to address these issues, the Heart Failure Association of the European Society of Cardiology (HFA-ESC) convened a group of expert heart failure clinical investigators, biostatisticians, regulators, and pharmaceutical industry scientists (Nice, France, 12-13 February 2012) to evaluate the challenges of defining heart failure endpoints in clinical trials and to develop a consensus framework. This report summarizes the group's recommendations for achieving common views on heart failure endpoints in clinical trials.

We evaluated the effect of (dual) renin-angiotensin-aldosterone system (RAAS) blockade with valsartan and an ACE inhibitor [92.7% of patients were treated with an ACE inhibitor in the Valsartan in Heart Failure Trial (Val-HeFT)] in patients with NYHA class II-IV heart failure (HF) and reduced EF on cardiovascular (CV) death and HF hospitalization by subgroups and by presence of early worsening of renal function (EWRF) and according to baseline estimated glomerular filtration rate (eGFR).We analysed the data from 5010 patients enrolled in the Val-HeFT study. A total of 2346 (46.8%) patients had baseline renal impairment (i.e. baseline eGFR60 mL/min/1.73 m(2)). Further, 425 patients (8.6%) had EWRF (i.e. eGFR decrease20% within 1 month after randomization), whereas 4503 patients (91.4%) had ≤20% decline in eGFR. Overall, the difference between valsartan and placebo on the composite endpoint of CV death and HF hospitalization was significant [P = 0.0005; hazard ratio (HR) 0.83, 95% confidence interval (CI) 0.75-0.92)]. In patients with baseline renal impairment, the difference between the treatment groups was also significant (P = 0.0002; HR 0.76, 95% CI 0.66-0.88). Patients with EWRF had higher risk of CV death and HF hospitalization vs. those without ERWF (P0.0001; HR 1.44, 95% CI 1.21-1.71), and within the EWRF group a significant difference was also observed between valsartan and placebo (P = 0.0086; HR 0.63, 95% CI 0.45-0.89). However, the interaction between treatment and eGFR at Month 1 was not significant (P = 0.1160).Benefits were maintained in patients with renal dysfunction at baseline and those who experienced EWRF.