Accurate Prediction of Bronchopulmonary Dysplasia: Are We There Yet?

See related article, p 113370 See related article, p 113370 Modern clinical care practices have led to improved survival of infants born at <28 weeks of gestation, with a concurrent increase in the number of infants with bronchopulmonary dysplasia (BPD).2Stoll B.J. Hansen N.I. Bell E.F. Walsh M.C. Carlo W.A. Shankaran S. et al.Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993-2012.JAMA. 2015; 314: 1039-1051Crossref PubMed Scopus (1776) Google Scholar Despite a decrease in the incidence of other neonatal morbidities, the incidence of BPD remains steady between 30% and 60%.2Stoll B.J. Hansen N.I. Bell E.F. Walsh M.C. Carlo W.A. Shankaran S. et al.Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993-2012.JAMA. 2015; 314: 1039-1051Crossref PubMed Scopus (1776) Google Scholar,3Poindexter B.B. Feng R. Schmidt B. Aschner J.L. Ballard R.A. Hamvas A. et al.Comparisons and limitations of current definitions of bronchopulmonary dysplasia for the prematurity and respiratory outcomes program.Ann Am Thorac Soc. 2015; 12: 1822-1830Crossref PubMed Scopus (193) Google Scholar BPD is the most common chronic complication of premature birth and is associated with an increased risk of mortality, as well long-term neurodevelopmental and respiratory morbidities.2Stoll B.J. Hansen N.I. Bell E.F. Walsh M.C. Carlo W.A. Shankaran S. et al.Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993-2012.JAMA. 2015; 314: 1039-1051Crossref PubMed Scopus (1776) Google Scholar,4Schmidt B. Roberts R.S. Davis P.G. Doyle L.W. Asztalos E.V. Opie G. et al.Prediction of late death or disability at age 5 years using a count of 3 neonatal morbidities in very low birth weight infants.J Pediatr. 2015; 167: 982-986.e2Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar,5Keller R.L. Feng R. DeMauro S.B. Ferkol T. Hardie W. Rogers E.E. et al.Bronchopulmonary dysplasia and perinatal characteristics predict 1-year respiratory outcomes in newborns born at extremely low gestational age: a prospective cohort study.J Pediatr. 2017; 187: 89-97.e3Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar For many years, extensive research has been conducted to identify therapies that prevent the development of BPD.6Abiramalatha T. Ramaswamy V.V. Bandyopadhyay T. Somanath S.H. Shaik N.B. Pullattayil A.K. et al.Interventions to prevent bronchopulmonary dysplasia in preterm neonates: an umbrella review of systematic reviews and meta-analyses.JAMA Pediatr. 2022; 176: 502-516Crossref PubMed Scopus (15) Google Scholar Postnatal corticosteroids are one of the few proven therapies used for BPD prevention.7Doyle L.W. Halliday H.L. Ehrenkranz R.A. Davis P.G. Sinclair J.C. An update on the impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk of bronchopulmonary dysplasia.J Pediatr. 2014; 165: 1258-1260Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar,8Ramaswamy V.V. Bandyopadhyay T. Nanda D. Bandiya P. Ahmed J. Garg A. et al.Assessment of postnatal corticosteroids for the prevention of bronchopulmonary dysplasia in preterm neonates: a systematic review and network meta-analysis.JAMA Pediatr. 2021; 175: e206826Crossref PubMed Scopus (39) Google Scholar Although they are associated with decreased risk of BPD and improvement in lung function, their use has also been associated with cerebral palsy, neurodevelopmental impairment, and other serious adverse events.7Doyle L.W. Halliday H.L. Ehrenkranz R.A. Davis P.G. Sinclair J.C. An update on the impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk of bronchopulmonary dysplasia.J Pediatr. 2014; 165: 1258-1260Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar, 8Ramaswamy V.V. Bandyopadhyay T. Nanda D. Bandiya P. Ahmed J. Garg A. et al.Assessment of postnatal corticosteroids for the prevention of bronchopulmonary dysplasia in preterm neonates: a systematic review and network meta-analysis.JAMA Pediatr. 2021; 175: e206826Crossref PubMed Scopus (39) Google Scholar, 9Zeng L. Tian J. Song F. Li W. Jiang L. Gui G. et al.Corticosteroids for the prevention of bronchopulmonary dysplasia in preterm infants: a network meta-analysis.Arch Dis Child Fetal Neonatal Ed. 2018; 103: F506-F511Crossref PubMed Scopus (43) Google Scholar, 10Harmon H.M. Jensen E.A. Tan S. Chaudhary A.S. Slaughter J.L. Bell E.F. et al.Timing of postnatal steroids for bronchopulmonary dysplasia: association with pulmonary and neurodevelopmental outcomes.J Perinatol. 2020; 40: 616-627Crossref PubMed Scopus (22) Google Scholar For this reason, the recently updated American Academy of Pediatrics guidelines emphasize the importance of appropriate patient selection to avoid use of postnatal corticosteroids for BPD prevention in infants who are unlikely to benefit from this therapy.11Cummings J.J. Pramanik A.K. COMMITTEE ON FETUS AND NEWBORNPostnatal corticosteroids to prevent or treat chronic lung disease following preterm birth.Pediatrics. 2022; e2022057530https://doi.org/10.1542/peds.2022-057530Crossref Google Scholar In addition, there is evidence showing that postnatal corticosteroids are most beneficial in infants who are at the highest risk of developing BPD, further highlighting the importance of individual patient consideration.7Doyle L.W. Halliday H.L. Ehrenkranz R.A. Davis P.G. Sinclair J.C. An update on the impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk of bronchopulmonary dysplasia.J Pediatr. 2014; 165: 1258-1260Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar However, predicting an infant's risk of developing BPD is challenging, especially in the first few postnatal weeks. BPD prediction models (estimators) are valuable to clinicians particularly during this time because not only could they theoretically help guide clinical therapy decisions such as the administration of postnatal steroids, but they could also aid with counseling families regarding individual infant risk and prognosis. In this volume of The Journal, Romijn et al performed a systematic review and meta-analysis of multivariable prediction models for BPD.12Romijn M. Dhiman P. Finken M.J.J. van Kaam A.H. Katz T.A. Rotteveel J. et al.Prediction models for bronchopulmonary dysplasia in preterm infants: a systematic review and meta-analysis.J Pediatr. 2023; 258113370Abstract Full Text Full Text PDF Scopus (2) Google Scholar They included 65 studies that were designed to develop or validate a BPD prediction model in preterm infants at 36 weeks of postmenstrual age (PMA). The authors pooled the data from these studies to generate a pooled c-statistic with corresponding 95% CI. Among 57 externally validated models included in the meta-analysis, the pooled c-statistic was 0.78 (95% CI 0.74-0.82). The median c-statistic for development models was 0.84, and that of validation models was 0.77. For validated models, c-statistics increased after the first week of age. Even though all validated models demonstrated good performance based on c-statistic values, they all scored high for risk of bias or applicability concerns. This study had multiple limitations which were appropriately highlighted by the authors, including selection bias and variability in predictors used to generate the models. Chief among the limitations of this type of analysis is that BPD is a clinical diagnosis with variable definitions in the literature. The difference in outcome definitions likely contributed to the variability in c-statistics. Since the initial description of BPD, the diagnostic criteria for BPD have evolved.3Poindexter B.B. Feng R. Schmidt B. Aschner J.L. Ballard R.A. Hamvas A. et al.Comparisons and limitations of current definitions of bronchopulmonary dysplasia for the prematurity and respiratory outcomes program.Ann Am Thorac Soc. 2015; 12: 1822-1830Crossref PubMed Scopus (193) Google Scholar The most commonly used traditional definition by Shennan et al defines BPD as the need for oxygen at 36 weeks of PMA.3Poindexter B.B. Feng R. Schmidt B. Aschner J.L. Ballard R.A. Hamvas A. et al.Comparisons and limitations of current definitions of bronchopulmonary dysplasia for the prematurity and respiratory outcomes program.Ann Am Thorac Soc. 2015; 12: 1822-1830Crossref PubMed Scopus (193) Google Scholar,13Bancalari E. Abdenour G.E. Feller R. Gannon J. Bronchopulmonary dysplasia: clinical presentation.J Pediatr. 1979; 95: 819-823Abstract Full Text PDF PubMed Scopus (424) Google Scholar In 2000, the National Institutes of Health categorized BPD as mild, moderate, or severe based on respiratory support required at 36 weeks of PMA, in infants who also had required supplemental oxygen for at least 28 days.14Jobe A.H. Bancalari E. Bronchopulmonary dysplasia.Am J Respir Crit Care Med. 2001; 163: 1723-1729Crossref PubMed Scopus (3739) Google Scholar The most recent definition by Jensen et al categorizes BPD into 3 grades depending on the level of respiratory support at 36 weeks of PMA, regardless of oxygen therapy.15Jensen E.A. Dysart K. Gantz M.G. McDonald S. Bamat N.A. Keszler M. et al.The diagnosis of bronchopulmonary dysplasia in very preterm infants: an evidence-based approach.Am J Respir Crit Care Med. 2019; 200: 751-759Crossref PubMed Scopus (402) Google Scholar Ultimately, all these options tether the diagnostic definition to treatment given for the diagnosis, which is problematic because individual physician practices are an important factor determining the use of oxygen and/or respiratory support at any given time. Some consistency in supplemental oxygen use has been achieved following the Neonatal Oxygen Prospective Meta-Analysis studies and other trials that have informed changes in clinical care practices surrounding oxygen saturation goals.16Askie L.M. Darlow B.A. Finer N. Schmidt B. Stenson B. Tarnow-Mordi W. et al.Association between oxygen saturation targeting and death or disability in extremely preterm infants in the neonatal oxygenation prospective meta-analysis collaboration.JAMA. 2018; 319: 2190-2201Crossref PubMed Scopus (253) Google Scholar, 17Manja V. Lakshminrusimha S. Cook D.J. Oxygen saturation target range for extremely preterm infants: a systematic review and meta-analysis.JAMA Pediatr. 2015; 169: 332-340Crossref PubMed Scopus (123) Google Scholar, 18Huizing M.J. Villamor-Martinez E. Vento M. Villamor E. Pulse oximeter saturation target limits for preterm infants: a survey among European neonatal intensive care units.Eur J Pediatr. 2017; 176: 51-56Crossref PubMed Scopus (15) Google Scholar However, the need for respiratory support may be multifactorial and difficult to distinguish whether needed for parenchymal lung disease or other unrelated conditions.19Coste F. Ferkol T. Hamvas A. Cleveland C. Linneman L. Hoffman J. et al.Ventilatory control and supplemental oxygen in premature infants with apparent chronic lung disease.Arch Dis Child Fetal Neonatal Ed. 2015; 100: F233-F237Crossref PubMed Scopus (15) Google Scholar At the time of BPD determination, premature infants may be on respiratory support for multiple reasons such as immature ventilatory control, apnea of prematurity, chest wall dyssynchrony, and airway obstruction.19Coste F. Ferkol T. Hamvas A. Cleveland C. Linneman L. Hoffman J. et al.Ventilatory control and supplemental oxygen in premature infants with apparent chronic lung disease.Arch Dis Child Fetal Neonatal Ed. 2015; 100: F233-F237Crossref PubMed Scopus (15) Google Scholar Because of this, current BPD definitions also capture infants who are otherwise well, and may not require respiratory support for true parenchymal lung disease.20Jobe A.H. The new bronchopulmonary dysplasia.Curr Opin Pediatr. 2011; 23: 167-172Crossref PubMed Scopus (423) Google Scholar Predicting an outcome that continues to be inconsistently defined is an exercise fraught with challenge if not futility, and the variability and limitations in performance of BPD estimators is certainly evidence of this. The authors have also published recent evidence suggesting the addition of more biologic/physiologic measures along with available estimator equations may strengthen the predictive value, and, potentially lead to better understanding of BPD phenotype variability.21Romijn M. van Kaam A.H. Fenn D. Bos L.D. van den Akker C.H.P. Finken M.J.J. et al.Exhaled volatile organic compounds for early prediction of bronchopulmonary dysplasia in infants born preterm.J Pediatr. 2023; 257113368Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar Even if it were possible to arrive at a perfect and unified definition of BPD, another factor to consider in the applicability of BPD estimators is individual site variability. Differences in care practices at individual sites lead to varying rates in BPD, and these differences are important to take into account when interpreting results of BPD estimators.22Walsh M. Laptook A. Kazzi S.N. Engle W.A. Yao Q. Rasmussen M. et al.A cluster-randomized trial of benchmarking and multimodal quality improvement to improve rates of survival free of bronchopulmonary dysplasia for infants with birth weights of less than 1250 grams.Pediatrics. 2007; 119: 876-890Crossref PubMed Scopus (107) Google Scholar The Neonatal Research Network BPD estimator, for example, did not include site as a covariate in the model to allow for generalizability across all centers.1Greenberg R.G. McDonald S.A. Laughon M.M. Tanaka D. Jensen E. Van Meurs K. et al.Online clinical tool to estimate risk of bronchopulmonary dysplasia in extremely preterm infants.Arch Dis Child Fetal Neonatal Ed. 2022; https://doi.org/10.1136/archdischild-2021-323573Crossref PubMed Scopus (13) Google Scholar Therefore, that estimator is likely to perform better at some sites than others. In one study examining the validity of the Neonatal Research Network BPD estimator at a single center, use of the estimator resulted in an overestimation of BPD and higher predicted postnatal steroid use compared with actual use.15Jensen E.A. Dysart K. Gantz M.G. McDonald S. Bamat N.A. Keszler M. et al.The diagnosis of bronchopulmonary dysplasia in very preterm infants: an evidence-based approach.Am J Respir Crit Care Med. 2019; 200: 751-759Crossref PubMed Scopus (402) Google Scholar,23Srivatsa B. Srivatsa K.R. Clark R.H. Assessment of validity and utility of a bronchopulmonary dysplasia outcome estimator.Pediatr Pulmonol. 2023; 58: 788-793Crossref PubMed Scopus (1) Google Scholar The question remains–how reliable or accurate does a BPD estimator need to be in order for it to be useful? The answer likely depends on the purpose for which it is used. When a BPD estimator is used for giving anticipatory guidance to families, some amount of uncertainty in predictive power might be acceptable, as long as this uncertainty is appropriately communicated to families. It is tempting to use BPD estimators as objective tools to aid in clinical decision-making, such as whether and when to initiate postnatal corticosteroids. Further studies of existing and ongoing cohorts are needed to identify and validate cut-offs that lead to improved outcomes, and inaccuracy of current estimators could limit the ability to accomplish this. The greatest immediate value of BPD estimators might lie in their integration into clinical trials so that investigators can better understand in real time the effects of trial interventions. For all potential uses of BPD estimators, increased uptake of use will likely correlate with accuracy of prediction; the article by Romijn et al takes an important step forward in summarizing the degree of reliability for estimators currently available to neonatal clinicians and investigators. Prediction Models for Bronchopulmonary Dysplasia in Preterm Infants: A Systematic Review and Meta-AnalysisThe Journal of PediatricsVol. 258PreviewTo review systematically and assess the accuracy of prediction models for bronchopulmonary dysplasia (BPD) at 36 weeks of postmenstrual age. Full-Text PDF Open Access