Prognostic value of NT-proBNP and best cut-offs for risk prediction in obese patients with chronic systolic heart failure

Background N-terminal fraction of pro-B-type natriuretic peptide (NT-proBNP) is a strong predictor of outcome in chronic systolic heart failure (CHF). Although plasma NT-proBNP has been reported to display an inverse correlation with body-mass-index (BMI), its prognostic value in obese patients has been poorly investigated so far. Objectives To evaluate the prognostic value and the best cut-offs for risk prediction of NT-proBNP in obese individuals from a multinational cohort of patients with systolic CHF. Methods We analyzed data from the BIOS (Biomarkers In Heart Failure Outpatient Study) Consortium. Patients with left ventricular ejection fraction (LVEF) ≤50% were selected and classified as nonobese (BMI <30 kg/m2), mildly obese (BMI 30–35 kg/m2), moderately obese (BMI 35–40 kg/m2), or severely obese (BMI ≥40 kg/m2), according to standard nomenclature. Clinical and bio-humoral data, including NT-proBNP testing, were retrieved, and 5-year cardiac and all-cause mortality status were considered as primary and secondary outcome, respectively. The independent prognostic role of NT-proBNP was evaluated through Cox regression analysis, adjusting the model for age, gender, New York Heart Association class, ischaemic aetiology, LVEF, and estimated glomerular filtration rate (eGFR). Results The study population included 11,574 patients (age 65±12 years, LVEF 30±9%, males 78%). Most of patients were nonobese (n=8,937, 77%), while mildly, moderately, and severely obese patients were 1,887 (16%), 499 (4%), and 251 (3%), respectively. Median values of plasma NT-proBNP were progressively lower from nonobese to mildly, moderately, and severely obese patients (1455 ng/L, 903 ng/L, 767 ng/L, 660 ng/L, respectively) and BMI was predictive of NT-proBNP independently from age, LVEF, and eGFR (r=−0.152, p<0.001). The best NT-proBNP cut-offs to predict 5-year cardiac mortality were significantly lower in mildly [1044 ng/L; AUC 0.663 (Sen 66%; Spe 59%), p<0.001] moderately [736 ng/L; AUC 0.670 (Sen 76%; Spe 52%), p<0.001]), and severely [1060 ng/L; AUC 0.635 (Sen 57%; Spe 69%), p=0.021] obese patients compared to nonobese individuals [2034 ng/L; AUC 0.714 (Sens. 65%; Spec. 66%), p<0.001]. Further, NT-proBNP independently predicted 5-year cardiac death in nonobese, mildly and moderately obese patients (all p<0.001), but not in severely obese patients (p=0.457). Similar findings were observed for the secondary endpoint of 5-year all-cause mortality. Conclusions Compared to nonobese CHF patients, obese patients have lower circulating NT-proBNP levels, which retain independent prognostic significance for cardiac and all-cause mortality across most categories of obesity. BMI-adjusted NT-proBNP cut-offs might be considered for prognostic stratification in obese patients with CHF. Funding Acknowledgement Type of funding source: None