Improving the prediction of bronchopulmonary dysplasia by identifying kidney-based risk factors - Approximately 50,000 infants are born less than 28 weeks’ gestation in the United States each year, and the development of severe bronchopulmonary dysplasia (BPD) in these infants is common. However, difficulties with early identification of at-risk infants and incomplete knowledge of modifiable risk factors have hampered improvements in BPD care. Acute kidney injury (AKI) and fluid overload are emerging risk factors for severe BPD in premature infants. A more complete understanding of how AKI and fluid overload in infants impacts the development of BPD could lead to better-targeted interventions in the early neonatal period as AKI and fluid overload are both preventable and treatable conditions. My long-term career goal is to develop a portfolio of research focusing on the relationship between lung and kidney disease in premature infants and decreasing the incidence of BPD in premature infants by identifying and mitigating kidney-based risk factors. This K23 proposal summarizes a 5-year program of mentored professional development tied to a research project to establish critical kidney-based candidates for early detection of severe BPD in premature infants and develop a clinical prediction model to identify premature infants at risk for severe BPD. My central hypothesis is that acute kidney injury (AKI) and fluid overload contribute to severe BPD development and are linked by angiogenesis. Specific aims for the K23 are to: (1) Develop a clinical prediction model that uses markers of kidney disease to estimate severe BPD risk in premature infants, and (2) Identify longitudinal patterns of angiogenesis biomarkers in premature infants that predict BPD. Aim 1 will test the hypothesis that a clinical prediction model which incorporates both AKI and fluid overload will more accurately identify BPD in premature infants than an existing model. Aim 2 will test the hypothesis that premature infants with AKI and subsequent severe BPD will have altered patterns of angiogenesis during their neonatal hospitalization. Through my career development plan and guidance from my mentors, I will develop expertise in a) bronchopulmonary dysplasia and angiogenesis, b) machine learning and health informatics, and c) clinical trial design and biostatistical principles. Together, the research and educational aims of this proposal will provide me with the necessary groundwork to compete for additional funding as an independent investigator. Specifically, I will seek R01 grant funding from the NHLBI in year 4 of my K23 award to develop individualized fluid management and AKI detection/treatment protocols for testing in a clinical trial. I will also apply for an R03 in year 3 to further augmented the developed clinical prediction models with biomarkers. Overall, my proposed line of research is the first step in a career focused on the detection, monitoring and treatment of kidney-based risk factors for BPD in premature infants. I will accomplish this work under the mentorship of a multidisciplinary team of faculty with expertise in health services research, clinical trials, machine learning, bronchopulmonary dysplasia, and kidney disease.