Which heart failure patients profit from natriuretic peptide guided therapy? A meta-analysis from individual patient data of randomized trials

Abstract

Aims Previous analyses suggest that heart failure (HF) therapy guided by (N-terminal pro-)brain natriuretic peptide (NT-proBNP) might be dependent on left ventricular ejection fraction, age and co-morbidities, but the reasons remain unclear. Methods and results To determine interactions between (NT-pro)BNP-guided therapy and HF with reduced [ejection fraction (EF) ≤45%; HF with reduced EF (HFrEF), n = 1731] vs. preserved EF [EF > 45%; HF with preserved EF (HFpEF), n = 301] and co-morbidities (hypertension, renal failure, chronic obstructive pulmonary disease, diabetes, cerebrovascular insult, peripheral vascular disease) on outcome, individual patient data (n = 2137) from eight NT-proBNP guidance trials were analysed using Cox-regression with multiplicative interaction terms. Endpoints were mortality and admission because of HF. Whereas in HFrEF patients (NT-pro)BNP-guided compared with symptom-guided therapy resulted in lower mortality [hazard ratio (HR) = 0.78, 95% confidence interval (CI) 0.62-0.97, P = 0.03] and fewer HF admissions (HR = 0.80, 95% CI 0.67-0.97, P = 0.02), no such effect was seen in HFpEF (mortality: HR = 1.22, 95% CI 0.76-1.96, P = 0.41; HF admissions HR = 1.01, 95% CI 0.67-1.53, P = 0.97; interactions P < 0.02). Age (74 ± 11 years) interacted with treatment strategy allocation independently of EF regarding mortality (P = 0.02), but not HF admission (P = 0.54). The interaction of age and mortality was explained by the interaction of treatment strategy allocation with co-morbidities. In HFpEF, renal failure provided strongest interaction (P < 0.01; increased risk of (NT-pro)BNP-guided therapy if renal failure present), whereas in HFrEF patients, the presence of at least two of the following co-morbidities provided strongest interaction (P < 0.01; (NT-pro)BNP-guided therapy beneficial only if none or one of chronic obstructive pulmonary disease, diabetes, cardiovascular insult, or peripheral vascular disease present). (NT-pro)BNP-guided therapy was harmful in HFpEF patients without hypertension (P = 0.02). Conclusion The benefits of therapy guided by (NT-pro)BNP were present in HFrEF only. Co-morbidities seem to influence the response to (NT-pro)BNP-guided therapy and may explain the lower efficacy of this approach in elderly patients.