Prediction of Bronchopulmonary Dysplasia-Associated Pulmonary Hypertension by Combining Clinical and Genetic Data

December 18, 2018
Written by

Identifying patients least likely to develop pulmonary hypertension allows physicians to focus on those with greater risk.

A model combining genetic and clinical data can predict bronchopulmonary dysplasia (BPD)-associated pulmonary hypertension (PH), which significantly increases mortality in preterm infants. The model, developed by a physician-scientist team from Nationwide Children’s Hospital, is published in a new study in the journal Acta Paediatrica.

“Of the roughly 4 million births in the United States every year, about 2 percent of those are at less than 32 weeks, making those infants a high-risk group for BPD. The incidence of BPD in those patients is between 40 and 60 percent, and a significant portion of those develop severe BPD, which puts them at the highest risk of getting pulmonary hypertension,” says Leif Nelin, MD, division chief of Neonatology at Nationwide Children’s and senior author of the study.

The researchers tested models using clinical data, genetic data, or both to develop a model that best predicts PH in patients with BPD.

Patients admitted to the Nationwide Children’s NICU after September 1, 2009, with a diagnosis of BPD, born at less than 35 weeks gestation, were enrolled in the Ohio Perinatal Research Network and Perinatal Research Repository, and had genotype data for two genes with polymorphisms known to have protective effects against PH, ARG1 and DDAH1, were included in the study.

The case-control study had a total of 20 patients with BPD-PH and 59 patients with only BPD.

Of various clinical characteristics, patients with BPD-PH had shorter gestational periods and lower birth weights than patients with BPD (p < 0.05). Thus, these two characteristics were included in the predictive modeling.

While all single genotype tests had low sensitivity and specificity, the researchers found that combining the data from both genes provided a negative predictive value of 91 percent. That is to say, they were able to predict those infants that did not develop PH better than those that did develop PH.

“If we can tell who is not going to get pulmonary hypertension then we can focus on the other group, those infants who are at higher risk, and use targeted preventative therapies,” says Dr. Nelin, who is also a professor at The Ohio State University.

Even though in most cases, genotype data will be unavailable, adds Dr. Nelin, that physicians can take birth weight and gestational age into account when they have a patient with severe BPD and monitor them for the development of PH.

“We’ve been trying to prevent severe BPD for many decades, and it is still a problem. We need to work on improving the care that these patients receive and watch for PH so we can improve their outcomes going forward.”



Trittmann JK, Bartenschlag A, Zmuda EJ, Frick J, Stewart WCL, Nelin LD. Using clinical and genetic data to predict pulmonary hypertension in bronchopulmonary dysplasia. Acta Paediatrica. 2018 Dec;107(12):2158-2164.


Image credit: Nationwide Children’s