EBNEO Commentary: Does the effect of hydrocortisone in preterm infants depend on the baseline risk of BPD?

The findings published by Gentle et al. must be considered within the context of a subpopulation of extremely preterm infants with high baseline illness severity and may not generalise to broader preterm populations. The current data do not examine whether the baseline risk for BPD or death modifies the effect of hydrocortisone in all extremely preterm infants as infants in the NICHD NRN Hydrocortisone Trial carried a high risk for developing BPD based on the physiologic definition⁵–83.4% of infants in the hydrocortisone group and 86.4% of infants in the placebo group developed the composite outcome of death or BPD.⁶Prior studies investigating postnatal hydrocortisone found that 49% and 71% of infants receiving placebo died or developed BPD, with lower rates of death or BPD among hydrocortisone exposed infants.^7, 8 Infants in those studies were exposed to hydrocortisone within 24 h of birth and at a median of 10 days, respectively. Therefore, infants in the NICHD NRN Hydrocortisone Trial represent a subset of extremely preterm infants with higher overall risk for death or BPD and were exposed to hydrocortisone at an older age.

While no significant interaction of hydrocortisone treatment and baseline risk for BPD or death with the primary efficacy outcome was noted, close investigation suggests a trend towards a lower comparative relative risk for the development of grade 2 or 3 BPD in infants in the highest quartile of BPD or death risk (RR 0.94; 95% CI 0.81, 1.09) compared to infants in the lowest quartile of risk (RR 1.13; 95% CI 0.82, 1.55). The risk difference estimates for these quartiles highlight the clinical relevance of these trends: an estimate of a 5% higherabsolute risk for BPD or death among infants exposed to hydrocortisone in the lowest risk quartile and a 5% lower absolute risk among infants exposed to hydrocortisone in the highest risk quartile. These data suggest that infants with higher risks of BPD or death may comparatively benefit from postnatal hydrocortisone as administered in this trial, but with too much uncertainty.

In conclusion, among extremely preterm infants enrolled in the NICHD NRN Hydrocortisone Trial, the effect of hydrocortisone was not modified by baseline risk for developing grades 2 or 3 BPD or death. Further research in larger and more representative cohorts of extremely preterm infants would aid in understanding potential differential effects of hydrocortisone exposure among infants with baseline differences in lung disease severity."