432 - The hydrocortisone-responsive urinary metabolome of premature infants

Poster Number: 432
Publication Number: 432.223

Dara Torgerson, University of California San Francisco, San Francisco, CA, United States; Cheryl Chapin, UCSF Benioff Children's Hospital San Francisco, San Francisco, CA, United States; Peter Her, University of Toronto Temerty Faculty of Medicine, Toronto, ON, Canada; Philip L. Ballard, University of California, San Francisco, School of Medicine, Bodega Bay, CA, United States

Background: Premature infants with lung disease often receive hydrocortisone (HC) to treat hypotension and suspected adrenal insufficiency, or to prevent bronchopulmonary dysplasia. The metabolic effects of corticosteroids in infants are not well defined.

Objective: Identify changes in the urinary global metabolome in response to HC therapy.

Design/Methods: We performed untargeted metabolomics using UHPLC/mass spectrometry (Metabolon Inc) on 302 urine samples collected between 6 and 60 days from 28 intubated TOLSURF infants ≤28 wk gestation. The postnatal time course for abundance of 950 named biochemicals was determined for 14 infants before and after starting HC therapy (recommended regimen 3 mg/kg/day tapering to 0.5) and for 14 matched control infants. There were no significant differences between HC and control infants for gestational age, birth weight, gender, race, Respiratory Severity Score, late surfactant treatment or occurrence of bronchopulmonary dysplasia. Data were assessed using t test, false discovery rate, linear regression and Fisher’s Exact Test.

Results: Excreted cortisol compounds were mean 1.8-1.9-fold higher (p < 0.01) after starting therapy and increased with time (p 0.03) vs controls. Comparing individual metabolite levels at mean 4 days after starting HC (post) to 0-2 days before treatment (pre), there was a significant (p < 0.05) post/pre difference for total 219 chemicals (162 named) compared to 28 for controls. The time course for excreted cortisol and pregnanolone shows changes in abundance of both metabolites by 2 days after starting HC (Fig.). None of the changes in metabolites of HC-treated infants occurred in the control group. With HC therapy, abundance of 82% of regulated metabolites decreased and 18% increased. Major HC responses (p < 0.001, suppression by 24-52%) occurred for excreted derivatives of pregnanolone, DHEA, glutarate, histidine, nicotinate and lactosamine. HC-regulated named metabolites represented all 9 major pathways and 50 of 100 sub pathways, reflecting the known diverse metabolic effects of corticosteroids, and there was enrichment (p < 0.10) for sub pathways of androgenic steroids, DHEA and aminosugar metabolism.Conclusion(s): With the dosing regimen used in TOLSURF, HC treatment of infants altered urinary levels of 18% of metabolites detected by global metabolomics. Despite suppression of adrenal cortex steroidogenesis, excreted cortisol increased nearly 2-fold during therapy, indicating increased systemic corticosteroid exposure consistent with the metabolic perturbations.
Figure. Time course for urinary concentrations of excreted cortisol and pregnanolone in one infant treated with hydrocortisone (HC) starting on day 25.HC Preg 084901 Gr.jpegPregnanolone decreases by 80% and cortisol increases several fold during HC treatment.