Recent advances in understanding and management of bronchopulmonary dysplasia

Does an ideal definition of bronchopulmonary dysplasia exist?

Various definitions exist for diagnosing BPD; however, most of them suffer from the primary limitation of utilizing a treatment modality as the primary factor defining the disease rather than the disease pathology or severity²². One of the commonly used definitions of BPD is the NIH consensus definition by Jobe et al., which recommends diagnosing infants with BPD based on their oxygen exposure and determines severity by assessing their respiratory support at 36 weeks postmenstrual age (PMA)²³. A major limitation of the available definitions of BPD is their inability to diagnose BPD in infants dying from severe respiratory failure before 36 weeks PMA. These infants comprise a subset of extremely preterm infants who are at high risk for developing severe BPD if they survive. An additional challenge comes from the increasing use of high-flow nasal cannula for oxygen therapy that provides a variable degree of positive end-expiratory pressure, thus creating the conundrum of whether these infants should be classified on the basis of supplemental oxygen concentration alone or by including the positive pressure that they receive. On the other end of the spectrum, we have infants on 100% oxygen via very low nasal cannula flows who may not be appropriately categorized with the available definitions. Another group of infants difficult to classify are those dependent on positive-pressure ventilation for airway abnormalities or abnormal respiratory control with no or minimal parenchymal lung disease. These infants should not be classified as having BPD, but we do not have the adequate means to re-categorize them currently²⁴.

In the summary from the workshop on BPD organized by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), the expert panel proposed an updated definition of BPD²⁵. In this new definition, the panel addresses the use of contemporary modes of ventilation and stage BPD based on the precise respiratory support and fraction of inspired oxygen (FiO₂) being used. They also exclude the need for supplemental O₂ exposure for 28 days and include “radiographic confirmation of parenchymal disease”²⁶. The use of this definition accurately predicted death or serious respiratory morbidity at 18–26 months corrected age in 81% of infants in the study²⁷. An ideal definition of BPD should be objective, take into account the pathophysiology, and be able to predict meaningful long-term outcomes^28,29. Use of long-term biomarkers for assessing lung injury³⁰, imaging studies like functional magnetic resonance imaging to assess lung structure³¹, and novel ways to assess pulmonary function will be required to create a new classification system for BPD that is more centered on lung function evaluation and pathophysiology²⁸.

The “omics” in bronchopulmonary dysplasia

In order to accurately assess the lung pathology in BPD, several traditional biomarkers have been utilized, but, owing to variable phenotypes seen in BPD, these biomarkers have low predictive value³⁰. In the past decade, newer technologies have allowed the detailed analysis of “big data sets” through bioinformatics to help in delineating the pathogenesis, mechanisms, and diagnosis of diseases. Novel systems biology-based “omic” approaches can be utilized to define the multiple cellular and humoral interactions that regulate the dysregulated lung development and injury response in BPD.

Metabolomics has been utilized to develop a panel of biomarkers to identify neonates at risk of developing BPD. Metabolomics consists of quantitative analysis of many low-molecular-mass metabolites found in a specific cell, organ, or organism involving substrates or products of a defined metabolic pathway. The changes in metabolite composition reflect the interaction among a genetic predisposition, a specific pathophysiological state, and environmental stimuli and can provide a unique fingerprint of those pathophysiological changes that predict the predisposition to the disease³². La Frano et al., studied the metabolomics in umbilical cord blood from 10 neonates who later developed BPD and compared them to 10 controls and found an alteration in lipid metabolism that may be responsible for an immaturity of lipid biosynthesis in these infants³³. In another study evaluating tracheal aspirate fluid from 68 preterm neonates on the first day of life, the authors demonstrated that neonates who develop BPD later in life have higher levels of serine, alanine, taurine, and citrulline compared to neonates who do not develop BPD³⁴.

As mentioned earlier in this review, various studies have demonstrated associations between dysbiosis of the airway microbiome and BPD; however, the direction of causality between airway injury during lung development and respiratory colonization with microorganisms is unclear. A systematic review of six studies investigating the airway microbiome of preterm infants showed that early microbiome contains Firmicutes and Proteobacteria as the predominant phyla and Staphylococcus and Ureaplasma as the predominant genera¹⁹. Infants who eventually developed severe BPD demonstrated an abundance of Ureaplasma and less Staphylococcus in their tracheal aspirates initially after birth¹⁸. BPD progression was associated with increased microbial community turnover, change in relative abundance of Firmicutes and Proteobacteria, and decrease in Lactobacilli^19,20. Metabolomics and microbiomics can be integrated to help guide us to the pathways that are activated in newborns who develop BPD based on the microbiome they have in their airways/lungs³². Lal et al. investigated this relationship between airway microbiome and metabolome at birth. They found that in infants with BPD, there is an increase in Proteobacteria, a reduction in Lactobacilli, and a decrease in the ratio of acetyl-CoA/propionyl-CoA carboxylase, indicating a curtailed fatty acid β-oxidation pathway. These changes suggest that the metabolic activity of the airway microbiome may in fact modulate the metabolome. The authors further postulated that the lipid metabolites from the bacteria may cause increased airway inflammation, leading to BPD³⁵.

Proteomics is the large-scale study of the structure and function of proteins which may help evaluate the networks of proteins that can specify the real-time status of disease state and protein function modulation³⁶. Analysis of bronchoalveolar lavage fluid in preterm babies using a proteomics-based approach showed that calcium signaling-related proteins differed with BPD severity³⁷. In a study evaluating plasma proteins using proteomics in preterm infants, it was noted that there was decreased gelsolin, afamin, and carboxypeptidase N subunit 2 levels in cord blood³⁸. In addition, there was decreased hemoglobin subunit gamma-1 and galectin-3 binding protein levels, along with increased serotransferrin in plasma at 36 weeks PMA³⁸. In a recent study by Förster et al., proteomic screening was utilized in plasma samples from preterm infants to identify novel biomarkers of BPD that could help in the early identification of such infants. They reported that increased levels of sialic acid-binding immunoglobulin-type lectin (SIGLEC)-14 and basal cell adhesion molecule (BCAM) along with decreased levels of angiopoietin-like 3 (ANGPTL3) in the first week of life can predict the outcome of BPD with high specificity and sensitivity³⁹.

In the field of genomics, various approaches including integrated genomic analysis⁴⁰ and gene expression profiling⁴¹ have led to the identification of novel pathways of lung development and repair (the cell-surface glycoprotein CD44, phosphorus oxygen lyase activity) and novel molecules and pathways (adenosine deaminase, targets of microRNA [miR]) that are associated with the genetic origins of BPD⁴². In addition to whole exome sequencing and whole genome sequencing to identify non-common variants in patients with BPD, RNA sequencing methods have been developed to evaluate protein-coding mRNA, transcripts of pseudogenes, long non-coding RNA (lncRNA), and miRs⁴³. In the last few years, miRs have been described as important mediators of normal growth, development, and disease. In a recent study, Syed et al. showed that miR-34a levels are significantly increased in the lungs of neonatal mice exposed to hyperoxia; deletion or inhibition of miR-34a improved the pulmonary phenotype and BPD-associated pulmonary arterial hypertension (PAH) in mouse models of BPD⁴⁴. Pharmacologic inhibition of miR-34a can thus be used as a potential therapeutic option in neonates to prevent hyperoxia-induced lung injury⁴⁵. In another study, increased expression of miR-451 was noted in animal models of BPD and inhibition of miR-451 was shown to improve the cardiopulmonary phenotype⁴⁶. Similarly, potential roles for miR-17∼92⁴⁷, miR-29b⁴⁸, miR-876-3p⁴⁹, miR-199a-5p⁵⁰, and miR-489⁵¹ have been described in the pathogenesis of BPD.

Stem cell-based therapies in bronchopulmonary dysplasia

MSCs play a vital role in lung development, and the therapeutic use of MSCs to treat and prevent BPD has emerged as a promising field. In preclinical studies, exogenous stem cells protect and repair hyperoxia-induced lung injury and improve survival. The precise mechanism by which they act is not completely known; however, it is believed that MSCs act primarily via a paracrine effect⁶³. In a systematic review evaluating the various cell‐based therapies in preclinical BPD, MSCs were found to be the most efficacious among all interventions, although they were exclusively investigated in the rodent hyperoxia-exposed model of BPD⁶⁴. Human umbilical cord tissue and cord blood have been identified as a potential source of MSCs and have been further explored as a feasible source of these cells and exosomes^65,66. At least two phase I clinical trials of human cord blood-derived MSCs have been successfully conducted in South Korea and the United States^67,68. Another source for similar cells is from placental membranes, human amnion epithelial cells (hAECs), which have shown considerable promise in preclinical models of BPD. The first human clinical trial of hAECs in neonates with BPD has been completed to assess the safety of these cells⁶⁹.

Insulin-like growth factor-1

IGF-1 is a significant regulator of lung angiogenesis, fetal growth, and development. Lower levels of IGF-1 have been independently associated with increased risk of BPD and retinopathy of prematurity. In neonatal rodent models of hyperoxia-induced lung injury, the administration of IGF-1 seems to reduce lung injury⁷⁰. In a phase II clinical trial, treatment with recombinant human IGF along with its binding protein (rhIGF-1/rhIGFBP-3) reduced the incidence of severe BPD as a secondary outcome⁷¹.