Sacubitril/Valsartan and B-Type Natriuretic Peptide Value: Recalculating Route

In heart failure (HF), pressure and volume overload results in the release of natriuretic peptides (NP), including B-type natriuretic peptide (BNP), responsible for natriuresis, vasodilation and blockage of the renin–angiotensin–aldosterone system. BNP plays not only a role in diagnostic process, but also has a prognostic value, predicting hospitalization and mortality in chronic HF. Recently, sacubitril/valsartan has been included in the European Society of Cardiology HF therapeutic recommendations for patients with HF with reduced ejection fraction (HFrEF), as it proved to reduce death from cardiovascular causes and the risk of hospitalization for HF. Sacubitril inhibits neprilysin, a protease theoretically responsible for BNP cleavage. Therefore, in this context, the diagnostic and prognostic value of BNP has not been clarified. ProBNP, a 108-amino acid polypeptide, is synthesized and cut into a 76-amino acid non-biologically active peptide [N-terminal proBNP (NT-proBNP)] and a 32-amino acid biologically active peptide (BNP) in cardiomyocytes. Neprilysin was shown to be responsible for NP degradation such as atrial NP and C-type NP,1 but has neither effect on proBNP nor on NT-proBNP.2 As sacubitril acts inhibiting neprilysin activity, it was thought that it would induce an initial rise, followed by a decline, in NP like BNP, while lowering NT-proBNP levels, as seen in the PARADIGM-HF3 and PIONEER-HF trials.4 Thus, BNP use as a prognostic biomarker in patients treated with this drug has been discredited, recommending NT-proBNP. Although it was believed that neprilysin degrades also BNP, recent experimental trials failed to demonstrate neprilysin-mediated BNP degradation (and, therefore, BNP elevation) in human kidney membranes5 (main BNP target). The aim of this study was to evaluate if BNP is still a reliable biomarker. To describe the variation of BNP, patients with HFrEF who initiated sacubitril/valsartan were prospectively assessed. Clinical characteristics, aetiology, functional class, admission and mortality rates were evaluated. The endpoint was to evaluate if BNP levels after sacubitril/valsartan treatment are related to decompensations and HF admissions, maintaining its role as a diagnostic and prognostic biomarker. Abbott Architect BNP assay was used to analyse BNP plasma level before treatment and after 3, 6 and 12 months of treatment, as well as in every in-hospital admission. Significant BNP change was set as 50 pg/mL. Statistical analysis was performed using Statistical Package for Social Sciences V.22.0, statistical software (SPSS, Chicago, IL, USA). Shapiro–Wilk test was performed, proving the sample was not normally distributed. Friedman test was conducted, followed by post-hoc analysis with Wilcoxon signed-rank test. A total of 86 patients started sacubitril/valsartan from October 2016 to June 2018. Patients lacking BNP level during follow-up and those who discontinued sacubitril/valsartan were excluded. Ten patients stopped treatment (due to renal dysfunction in four, hyperkalaemia in one, hypotension in two, and economic difficulties in three). Two patients died during this period. Finally, a total of 70 patients were included. Baseline characteristics of the study patients are detailed in Table 1. At baseline, median BNP concentration was 576 pg/mL [interquartile range (IQR) 258–1058]. There were no significant differences in BNP levels after adjusting for sex, renal function, or New York Heart Association functional class. After 6 months of treatment, median BNP concentration decreased to 367 pg/mL (IQR 149–955; P = 0.001) (Figure 1). Based on BNP kinetics during follow-up, patients were classified into three groups: stable (change ≤ 50 pg/mL), rising and descending BNP level. After 6-month follow-up, 19% showed no changes, 21% presented rising levels, and 60% had lower BNP concentration. Comparing prevalence of readmissions between groups, patients with increasing BNP concentration were more frequently admitted because of HF (50%) than patients with lower concentration (21%). Sixteen patients suffered from acute decompensation during follow-up. There was a significant increase in BNP concentration over baseline (576 vs. 2019 pg/mL; P = 0.002) and a trend toward lower values at discharge, close to baseline values (576 vs. 927 pg/mL; P = 0.674), showing decongestion and decrease in myocardial stretching. Current evidence about BNP degradation is scarce and interaction between neprilysin and BNP levels is not simple. First of all, neither the imprecision of the test used, nor the normal biological variability may be ignored. The complex system of NP leads to an overlapped measurement. Conventional BNP tests detect most of the truncated BNP forms, but also have proBNP cross-reactivity (as they share the C-terminal end); even 53% of the BNP measured might be really proBNP.6 NT-proBNP tests are also inaccurate, as the detection might be underestimated because of sacubitril-mediated glycosylation. In addition, different BNP tests have different sensitivity and specificity, which might lead to inconsistent measurements during treatment with sacubitril/valsartan.7 Biological variability cannot be neglected either. Every stable patient presents a physiologic change in BNP concentration and reference change value reflects the significant minimum percentage change in serial measurements. Reference change values for long-term changes in BNP concentration have been calculated, with 95% confidence interval, at a maximum of 113% (high intraindividual variation of BNP and proBNP in patients with stable chronic HF). On the other hand, the main NP clearance pathway in compensated patients is the one depending on the NP receptor C and not neprilysin.8 In fact, BNP may be quite resistant to cleavage by neprilysin.1 Moreover, during acute decompensations, rising BNP levels induce a reduction in circulating neprilysin activity to a maximum of three times.9 Therefore, neprilysin is not the only degradation mechanism, and its inhibition might not influence that much BNP level. Finally, it has recently been demonstrated that sacubitril/valsartan increases the levels of glycosylated proBNP, which is especially resistant to cleavage into BNP and NT-proBNP.10 Therefore, we would expect lower BNP levels under sacubitril/valsartan treatment,11 although showing cross-reactivity with proBNP, may hide those results. Our study has some limitations that should be acknowledged. Number of patients is limited, and heterogeneity is scarce as it is a cohort followed in a HF unit and, usually, patients are symptomatic with a complicated course. Also, there is no control group to compare BNP evolution under conventional treatment. However, based on previously published data and our own results, we believe BNP rejection as a biomarker in HF patients could be inappropriate. In this sample, mid and long-term BNP level did not significantly increase; quite the opposite, it was described a lowering tendency. Therefore, sacubitril/valsartan does not artefact BNP raise in the acute setting and BNP decrease after improvement, i.e. it preserves its diagnostic value in acute decompensation. Also, persistent rising BNP concentration under sacubitril is related to frequent HF readmission, i.e. BNP keeps its prognostic value. In conclusion, according to our data, BNP preserves its diagnostic and prognostic significance regardless of sacubitril/valsartan treatment. Our findings suggest that BNP is still a reliable biomarker in patients treated with sacubitril/valsartan, and therefore it should continue to be used. Conflict of interest: none declared.