Exploring clinical, echocardiographic and molecular biomarkers to predict bronchopulmonary dysplasia

Abstract

INTRODUCTION: Bronchopulmonary dysplasia (BPD) is the most common chronic lung disease in childhood, related to prematurity, and the most common cause of pulmonary hypertension (PH) secondary to pulmonary disease in children. Moderate and severe BPD have a worse outcome and relate more frequently with PH. The prediction of moderate or severe BPD development in extremely premature newborns is vital to implement preventive strategies. Starting with the hypothesis that molecular biomarkers were better than clinical and echocardiographic factors, this study aims to explore the ability of clinical, echocardiographic and analytical variables to predict moderate or severe BPD in a cohort of extremely preterm infants. PATIENTS AND METHODS: We designed a prospective longitudinal study, in which we followed a cohort of preterm newborns (gestational age <28 weeks and weight </= 1250 grams). In these newborns we recorded weekly clinical and echocardiographic variables as well as blood and tracheal aspirate samples, to analyze molecular biomarkers (IL-6, IL-1, IP10, uric acid, HGF, endothelin-1, VEGF, CCL5). Variables and samples were collected since birth up to week 36 (postmenstrual age), time-point at which the diagnosis of BPD is established. RESULTS: We included 50 patients with a median gestational age of 26 weeks (IQR 25-27) and weight of 871 g (SD 161,0) (range 590-1200g). Three patients were excluded due to an early death. Thirty-five patients (74.5%) developed BPD (mild n = 14, moderate n = 15, severe n = 6). We performed a logistic regression in order to identify risk factors for moderate or severe BPD. We compared two predictive models, one with two variables (mechanical ventilation and inter-ventricular septum flattening), and another-one with an additional molecular biomarker (ET-1). CONCLUSIONS: The combination of clinical and echocardiographic variables is a valuable tool for determining the risk of BPD. We find the two variable model (mechanical ventilation and echocardiographic signs of PH) more practical for clinical and research purposes. Future research on BPD prediction should be oriented to explore the potential role of ET-1.