Association between Baseline Cortisol Serum Concentrations and the Effect of Prophylactic Hydrocortisone in Extremely Preterm Infants.

Abstract

To define nomograms of serum cortisol values before 24 hours of postnatal life for extremely preterm infants and determine whether baseline cortisol values affect the benefit/risk ratio of prophylactic hydrocortisone to improve survival without bronchopulmonary dysplasia (BPD). We performed a predefined secondary analysis of the multicenter randomized controlled PREMILOC trial that included inborn infants delivered before 28 weeks of gestation. Nomograms of baseline serum cortisol values measured in 325 enrolled patients were determined for male and female neonates and correlated to perinatal events. BPD-free survival and severe adverse events were analyzed in placebo and hydrocortisone groups according to the cortisol z score in multivariate logistic regression models. Increased cortisol levels measured before 24 hours following birth were associated with a significantly higher chance of BPD-free survival only in placebo-treated infants (aOR [95% CI] 1.57 [1.08-2.27], P = .02) based on sex-specific nomograms for baseline cortisol levels. The cortisol z score for infants treated with prophylactic hydrocortisone predicted a risk of high-grade intraventricular hemorrhage (aOR [95% CI] 1.82 [1.06-3.15], P = .03) and spontaneous intestinal perforation (aOR [95% CI] 4.81 [1.34-17.22], P = .02). We found no predictive value of baseline cortisol levels for BPD-free survival in infants born extremely preterm treated with hydrocortisone. However, high cortisol levels early after birth were associated with a greater risk of severe intraventricular hemorrhage and spontaneous intestinal perforation in infants treated with hydrocortisone and, therefore, a lower benefit/risk ratio for the treatment. EudraCT 2007-002041-20, ClinicalTrial.gov: NCT00623740.