Timing of antenatal corticosteroids and survival without neurologic disabilities at 5½ years in children born before 35 weeks of gestation

• Published in: American journal of obstetrics and gynecology 2023-12, Vol.229 (6), p.675.e1-675.e18
• Main Authors: Guerini, Claire, Goffinet, François, Marchand-Martin, Laetitia, Delorme, Pierre, Pierrat, Véronique, Ancel, Pierre-Yves, Schmitz, Thomas
• Format: Article
• Language: English
• Subjects: Adrenal Cortex Hormones - therapeutic use; antenatal corticosteroids; Child; Female; Gestational Age; Humans; Infant, Newborn; Infant, Newborn, Diseases; Infant, Premature; neurologic disabilities; Pregnancy; Premature Birth - drug therapy; preterm birth; survival; time interval
• ISSN: 0002-9378; 1097-6868
• DOI: 10.1016/j.ajog.2023.06.047

The efficacy of antenatal corticosteroids for neonatal preterm complications wanes beyond 7 days after treatment. The neurodevelopmental effects of longer treatment-to-birth intervals have not been adequately evaluated. This study aimed to assess the impact of antenatal corticosteroid timing on survival without moderate or severe neurologic disabilities at 5½ years. This was a secondary analysis of the EPIPAGE-2 study, a national population-based cohort (France) that recruited neonates in 2011 and followed them up at 5½ years (results first reported in 2021). Participants were children born alive between 24+0 and 34+6 weeks, with a complete corticosteroid course, delivery >48 hours after the first injection, and neither limitation of care decided before birth nor severe congenital malformation. The study included 2613 children, 2427 of whom were alive at 5½ years; 71.9% (1739/2427) had a neurologic assessment at this age; 1537 had a clinical examination (complete for 1532), and 202 were assessed with a postal questionnaire. Exposure was defined as the interval between the first injection of the last antenatal corticosteroid course and delivery in days, studied in 2 categories (days 3–7 and after day 7), in 4 categories (days 3–7, 8–14, 15–21, and after day 21), and continuously in days. The main outcome was survival at 5½ years without moderate/severe neurologic disabilities, defined as moderate/severe cerebral palsy, or unilateral or bilateral blindness or deafness, or Full-Scale Intelligence Quotient 2 standard deviations below the mean. A multivariate analysis with a generalized estimated equation logistic regression model assessed the statistical association between the main outcomes and the interval from the first corticosteroid injection of the last course to birth. Multivariate analyses were adjusted for potential confounders, defined with a directed acyclic graph: gestational age in days, number of corticosteroid courses, multiple pregnancy, and cause of prematurity in 5 categories. Because neurologic follow-up was complete in only 63.2% of cases (1532/2427), the analyses used imputed data. Among 2613 children, 186 died between birth and 5½ years. Overall survival was 96.6% (95% confidence interval, 95.9–97.0), and survival without moderate or severe neurologic disabilities was 86.0% (95% confidence interval, 84.7–87.0). Survival without moderate or severe neurologic disabilities was lower after day 7 (85.0%) than during the interval from day 3 to