Neonatal outcomes after maternal biomarker-guided preterm birth intervention: the AVERT PRETERM trial

Abstract

1. Background Vaginal progesterone, low dose aspirin and care management (comprising increased outreach and education) has been shown to reduce the rate of prematurity in select populations, but identifying at-risk pregnancies has been problematic. 2. Objective(s) Test the hypothesis that screening singleton, non-anomalous pregnancies lacking traditional clinical risk factors with a validated blood test for preterm birth risk prediction, then targeting those with elevated risk for preventive treatment, would improve neonatal outcomes as compared to a large historical population. 3. Study Design The study took place from June 2018-September 2020 at ChristianaCare Hospital (Newark, DE). Singleton non-anomalous pregnancies with no history of preterm birth were enrolled in a prospective study arm and followed through neonatal hospital discharge. Participants were screened using a serum proteomic test for spontaneous preterm birth risk during a gestational age window spanning 19 1/7-20 6/7 weeks. Pregnancies identified by the test to be at elevated risk for preterm birth (≥16.0%, approximately twice the U.S. population risk) were offered aspirin 81 mg daily, open-label vaginal progesterone 200 mg daily and care management. We compared outcomes for women who screened either low-risk or higher-risk accepting treatment with those in a historical study arm of 10,000 pregnancies. Our co-primary outcomes were neonatal hospital length of stay and an ordinal neonatal morbidity index score based on the occurrence of grade III/IV interventricular hemorrhage, necrotizing enterocolitis (Bell stage II/III), respiratory distress syndrome, bronchopulmonary dysplasia, retinopathy of prematurity (stage III), sepsis, and neonatal death or neonatal intensive care unit length of stay. Cox proportional hazards survival analysis and ordinal logistic regression were utilized to evaluate outcomes and control for population differences. 4. Results A total of 1463 women were screened and tested before research operations were ceased due to the advent of the COVID-19 pandemic, and three women were subsequently deemed ineligible after screening. Of these, 34.72% (507/1460) were deemed high-risk, with 56.41% (286/507) accepting intervention and 43.59% (221/507) forgoing intervention. The remaining 65.3% (953/1460) were designated as low-risk. Women in the prospective arm were older, more obese, more likely to have hypertension and smoke, and less likely to use opioids compared to women in the historical arm. The primary analyses found that neonates in the prospective arm were discharged from the hospital earlier (P=0.01; hazard ratio, 1.35; 95% confidence interval, 1.08-1.70) and had lower neonatal morbidity index scores (P=0.031; odds ratio, 0.81; 95% confidence interval, 0.67-0.98). Average neonatal hospital length of stay decreased by 21%, and severe neonatal morbidity (neonatal morbidity index ≥3) was reduced on average by 18%. 5. Conclusions Identifying singleton, non-anomalous pregnancies lacking traditional risk factors with a validated proteomic blood test for preterm birth risk and providing treatment to those at risk resulted in improved neonatal outcomes compared to controls in a racially diverse cohort. This test-and-treatment strategy shows promise for ameliorating the impacts of premature birth among individuals in this previously unidentified patient population.