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

Abstract

AimsPrevious 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 resultsTo 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).

Effectiveness and cost-effectiveness of serum B-type natriuretic peptide testing and monitoring in patients with heart failure in primary and secondary care: an evidence synthesis, cohort study and cost-effectiveness model

“…For the subgroup analyses, the HRs and associated 95% confidence intervals (CIs) for the interaction effect were derived from the subgroup-specific HRs with 95% CIs. Stratum-specific treatment effects for all-cause mortality (age and type of HF) and HF-specific admission (type of HF) were available from a previous IPD meta-analysis 33 that included seven RCTs. Interactions were calculated and pooled as an aggregate estimate with additional trials which had contributed IPD for this study.…”

“…We initially asked for the data for the RCTs included in their meta-analysis, allowing us to carry out the analyses; then we proposed that they should do the analyses to our analysis plan; tried to reach a compromise over the proposed analyses (with them doing the analyses); and asked for the pooled estimates from subgroup analyses that had already been carried out. 51 None of these proposals was successful and, therefore, we were unable to obtain IPD, or the results of subgroup analyses, from most RCTs included in the IPD meta-analysis by Troughton et al 25,33 Consequently, we were unable to perform the subgroup analyses specified in the protocol. 35 Other changes to the study from the protocol stage include:…”

“…We aimed to extend previous economic evaluations in two ways. First, we exploited recent IPD meta-analyses, 25,33 including the analyses presented in this report, in estimating the relative effect of BNP-guided therapy. Among the advantages of IPD meta-analysis is the opportunity to investigate the (cost-)effectiveness of BNP-guided therapy in subgroups of patients who are not analysed consistently or not reported in the original RCT publications.…”

New Management Strategies in Heart Failure

“…95 Another meta-analyses of individual patient data sought to investigate the interactions between age, comorbidities, or type of HF (preserved or reduced LVEF) and treatment response in the randomized trials of (NT-pro) BNP-guided therapy in HF. 96 These authors concluded that the benefits of therapy guided by (NT-pro) BNP were present in HFrEF only. However, most investigators agree that the ongoing Guiding Evidence Based Therapy Using Biomarker Intensified Treatment in Heart Failure (GUIDE-IT) trial (Table 2) will help clarify this important question.…”

December 2015 at a glance

“…3 In this issue of the journal, D'Elia et al discuss the role of new biomarkers for phenotyping these patients and differentiate pathogenetic mechanisms. 4 Other studies regard the independent prognostic role of high sensitivity troponin T measurements and their changes in patients with acute HF, 5 the use of biomarkers to select the HF patients at low risk of events, 6 the selection of the patients more likely to benefit from natriuretic peptides guided therapy 7 and a new, and first, marker of muscle wasting. 8 With respect of treatment, a simple treatment of sleep disordered breathing, based on a lateral sleep position, is shown to be effective in patients with HF, above all when obstructive sleep apnea is their main complaint.…”