Abstract Rationale Clinical response to postnatal corticosteroids varies among extremely premature infants, and identifying biomarkers of response and understanding the drivers of this variability could improve therapeutic outcomes. A genetic variant in the intron of corticotropin-releasing hormone receptor 1 (CRHR1, rs7225082) has previously been associated with a change in respiratory severity score (RSS = MAP x Fi02) following treatment with systemic corticosteroids. We aimed to validate the rs7225082 association with dexamethasone (DEX) response in a prospective study of premature infants and assess other potentially functional variants in CRHR1. Methods The DExamethasone for NeonAtal Lung Illness (DENALI) study included 101 infants from five U.S. NICUs (2016-2023). IRB approval was obtained at each site. Eligible infants were ≤32 weeks gestational age and 14-175 days postnatal age (PNA) at the start of DEX treatment. Infants were treated with a 7 day tapering course of DEX (cumulative dose 0.59 - 2.85 mg/kg). The primary outcome was change in lung function, assessed by RSS within 24 hours before treatment (pre-RSS) and 7 days after DEX initiation (post-RSS), with clinical response defined as ΔRSS = post-RSS - pre-RSS. Genotyping of ∼1.8M single nucleotide polymorphisms was performed on the Illumina Genomic Diversity Array followed by imputation using the TOPMed imputation server. Associations between CRHR1 variants and ΔRSS were tested using linear regression, adjusting for sex, pre-RSS, PNA, postmenstrual age, cumulative steroid dose by day 7, genetic relatedness, genetic population structure, and maternal race/ethnicity. Results The distribution of ΔRSS ranged from −12.56 to + 4.18 with a mean of − 3.47 (SD 3.11). No clinical or demographic characteristics were correlated with ΔRSS in single variate analyses. A total of 11 infants’ RSS worsened after receiving DEX, 9 of whom carried at least one T allele at rs7225082. This allele was also associated with a smaller improvement in RSS overall (ΔRSS, β = −0.74, p = 0.04, frequency 61.4%), consistent with a reduced clinical response to corticosteroids as previously reported. Additional CRHR1 variants were in linkage disequilibrium with rs7225082, and were similarly associated with ΔRSS at p 0.05 (min p-value=1.37x10-2). This included rs1040238305, which overlaps regulatory 5′ UTR regions, and rs4076452, an intronic variant previously linked to mediating hypothalamic-pituitary-adrenal axis stimulation. Conclusion We provide independent validation that rs7225082 is associated with clinical response to corticosteroids in premature infants, and identify additional CRHR1 variants as candidates for future investigations into biomarker development and functional studies to elucidate causal roles. This abstract is funded by: 1R21HD101111, R01HL164570
Garcia et al. (2026) studied this question.