Weiwei Jin, Wei-Min Brian Chiu, Asia Krupa, Christopher Jerdonek, Kendra R Cook, Derek Rinderknecht
Background: B-type natriuretic peptide (BNP) is widely used in the diagnosis, risk stratification, and management of heart failure (HF). However, its diagnostic performance for detecting elevated left ventricular filling pressure, particularly at lower thresholds, remains uncertain. Left ventricular end-diastolic pressure (LVEDP), measured during cardiac catheterization, provides direct evidence of abnormal left ventricular filling pressure and is a well-established indicator for HF. This study evaluated the diagnostic performance of guideline-recommended BNP thresholds for detecting elevated LVEDP. Methods: Data were analyzed from 154 patients referred for left heart catheterization (mean age, 66 ± 10 years; 64% male). LVEDP was measured directly using Millar Mikro-Tip pressure catheters. BNP was measured during the same visit. At a BNP threshold of 35 pg/mL, sensitivities for detecting LVEDP >15, >20, and >25 mmHg were 74%, 88%, and 92%, respectively, with corresponding specificities of 48%, 47%, and 42%. Results: At a BNP threshold of 100 pg/mL, sensitivities were 43%, 63%, and 75%, respectively, with corresponding specificities of 74%, 75%, and 71%. The corresponding areas under the receiver operating characteristic curve were 0.65, 0.75, and 0.81 for LVEDP >15, >20, and >25 mmHg, respectively. Conclusions: BNP demonstrated greater diagnostic accuracy at higher LVEDP thresholds but had limited sensitivity at lower thresholds, which may reduce its ability to detect hemodynamic abnormalities relevant to the evaluation of Stage B HF. BNP should therefore be interpreted in the context of a broader clinical assessment and additional diagnostic findings, particularly when the goal is to identify patients with less pronounced elevations in filling pressure.