Abstract
Extremely preterm infants often require prolonged respiratory support due to lung immaturity and inflammation, placing them at high risk of lung injury and development of bronchopulmonary dysplasia (BPD). In many of these infants, systemic postnatal corticosteroids are used to reduce lung inflammation, facilitate mechanical ventilation (MV) weaning and extubation, and improve short-term pulmonary outcomes. However, despite decades of clinical use, substantial variation persists in timing, choice of agent and dosing. These inconsistencies reflect a lack of strong evidence and a limited understanding of the systemic and organ-specific effects of therapy for a highly heterogenous population usually exposed to this medication. This narrative review addresses these gaps by integrating current knowledge of the inflammatory and respiratory effects of postnatal corticosteroids in extremely preterm infants. We explore how corticosteroids modulate pulmonary inflammation, their effects on lung development, and how they affect key clinical outcomes such as extubation success and BPD severity. We also examine evolving approaches to corticosteroid administration and dosing, highlighting the importance of individualized strategies informed by developmental and disease-specific considerations. Comparative data from randomized controlled trials are reviewed, including the efficacy and side-effect profiles of commonly used regimens. Current evidence supports judicious use of late low-dose dexamethasone, while early prophylaxis with inhaled or intratracheal steroids remains experimental and is not routinely advised. In line with a physiology-driven approach, we also discuss emerging domain-specific monitoring tools that may enhance patient selection and optimize timing of intervention. By synthesizing mechanistic insights with clinical evidence, this review supports a more nuanced, individualized approach to postnatal corticosteroid therapy in extremely preterm infants, balancing therapeutic benefits with potential systemic trade-offs.