Role of NT-proBNP and lung ultrasound in diagnosing and classifying heart failure in a hospitalized oldest-old population: a cross-sectional study

Abstract

Abstract Aim Diagnosing and classifying heart failure (HF) in the oldest-old patients has technical and interpretation issues, especially in the acute setting. We assessed the usefulness of both N-terminal pro-brain natriuretic peptide (NT-proBNP) and lung ultrasound (LUS) for confirming HF diagnosis and predicting, among hospitalized HF patients, those with reduced ejection fraction (HFrEF). Methods We performed a cross-sectional study on 148 consecutive patients aged ≥ 80 years admitted to our Internal Medicine and Geriatrics ward with at least one symptom/sign compatible with HF and NT-proBNP ≥ 125 pg/mL. We measured serum NT-proBNP levels and performed LUS and transthoracic echocardiography (TTE) on admission before diuretic therapy. We divided our cohort into three subgroups according to the left ventricular ejection fraction (LVEF): reduced (LVEF ≤ 40%), mildly-reduced (LVEF = 41-49%) and preserved (LVEF ≥ 50%). Results The mean age was 88±5 years. Male prevalence was 42%. Patients with HFrEF were 19%. Clinical features and laboratory parameters did not differ between the three subgroups, except for higher NT-proBNP in HFrEF patients, which also had a higher number of total B-lines and intercostal spaces of pleural effusion at LUS. Overall, NT-proBNP showed an inverse correlation with LVEF (r = -0.22, p = 0.007) and a direct correlation with age, total pulmonary B-lines, and intercostal spaces of pleural effusion. According to the ROCs, NT-proBNP levels, pulmonary B-lines and pleural effusion extension were poorly predictive for HFrEF. The best-performing cut-offs were 9531 pg/mL for NT-proBNP (SP 0.70, SE 0.50), 13 for total B-lines (SP 0.69, SE 0.85) and one intercostal space for pleural effusion (SP 0.55, SE 0.89). Patients with admission NT-proBNP ≥ 9531 pg/mL had a 2-fold higher risk for HFrEF (OR 2.5, 95% CI 1.3-4.9), while we did not find any association for total B-lines ≥ 13 or pleural effusion ≥ 1 intercostal space with HFrEF. A significant association with HFrEF emerged for the combination of NT-proBNP ≥ 9531 pg/mL, total B-lines ≥ 13 and intercostal spaces of pleural effusion ≥ 1 (adjusted OR 4.3, 95% CI 1.5-12.9). Conclusions Although NT-proBNP and LUS help diagnose HF, their accuracy in discriminating HFrEF from non-HFrEF was poor in our real-life clinical study on oldest-old hospitalized patients, making the use of TTE still necessary to distinguish HF phenotypes in this peculiar setting. These data require confirmation in more extensive and longer prospective studies.