Point-of-care NT-proBNP for the screening of PREGnancy-related Heart Failure – the PREG-HF study

Introduction

Cardiovascular disease is an important cause of maternal mortality worldwide. However, diagnosing heart failure (HF) during pregnancy remains challenging. Patients with HF present with symptoms that are often attributed to the physiological changes of pregnancy. Although the measurement of natriuretic peptides (such as NT-proBNP) has been recommended as a cost-effective screening test for HF, its value in predicting underlying structural heart disease on echocardiography during pregnancy is unclear.

Purpose

To evaluate the accuracy of point-of-care (POC) NT-proBNP to predict echocardiographic evidence of structural heart disease in pregnant women.

Methods

All consecutive consenting pregnant women with symptoms of HF (New York Heart Association functional class [NYHA FC] II–IV), who underwent echocardiography at Groote Schuur Hospital (a tertiary hospital in South Africa) between 1 March 2021 and 15 March 2022 were recruited and compared with asymptomatic pregnant women (NYHA FC I). Demographic and obstetric data were collected, as well as clinical and echocardiographic parameters. POC NT-proBNP was measured; a receiver operating characteristic (ROC) curve was used to determine the level of NT-proBNP that would have the best predictive value for detecting structural heart disease on echocardiography.

Results

We included 121 women with a median age of 31.3 years (IQR 24.9–36.4) and a median gravidity of 3 (2–4), mostly in their third trimester of pregnancy (75.2%). Symptomatic women (66.1%) presented mainly with dyspnoea (92.5%) and fatigue (73.8%). Most common signs of HF were peripheral oedema (46.8%), jugular venous distension (17.7%) and pulmonary crackles (17.7%). Overall, the median POC NT-proBNP was 98pg/ml (60–506) but was not statistically different between symptomatic and asymptomatic participants. However, NT-proBNP levels were significantly elevated in those with left ventricular (LV) dilatation (376 [86–744] vs 65 [60–191], p=0.001), left atrial enlargement (395 [65–744] vs 60 [59–108], p<0.001), LV systolic dysfunction (510 [113–1668] vs 65 [60–210], p<0.001), diastolic dysfunction (300 [77–1450] vs 78 [60–257], p=0.031), mitral regurgitation (167 [60–672] vs 78 [60–243], p=0.038) and pericardial effusion (440 [81–1031] vs 71 [60–243], p=0.011) (Figure 1). An NT-proBNP of <200pg/ml had the highest negative predictive value (83%) to rule out structural heart disease (AUC 0.68 [0.59–0.77]). Patients with NT-proBNP of ≥200pg/ml were at increased risk of having echocardiographic evidence of structural heart disease (diagnostic odds ratio 4.9 [95% confidence interval 2.0–11.7]).

Conclusion

In this cohort of pregnant women with symptoms suggestive of HF, POC NT-proBNP identified those with structural heart disease with acceptable discrimination. POC NT-proBNP testing might be particularly useful as a screening test in settings where pregnant women do not readily have access to echocardiography.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): World Heart Federation Salim Yusuf Emerging Leaders Programme