Caffeine, Lorazepam, and Morphine Use in Neonatal Intensive Care Unit Population and Potential Relevance of Pharmacogenomics

Abstract

Abstract Medications are critical for managing acutely ill infants in neonatal intensive care unit (NICU). Tailoring therapies to individual differences in response and dose requirements might help improve outcomes. However, tools for optimizing pharmacotherapy remain limited. In this retrospective study, we explored the potential utility of pharmacogenetics for commonly prescribed NICU medications. All patients admitted to our local NICU March-December, 2022 were eligible for inclusion in the study. In infants who received at least three doses of one or more medications, we extracted DNA from residual EDTA blood collected for routine clinical care according to IRB-approved study protocols. Demographic, pharmacy, and clinical information was recorded retrospectively. Targeted pharmacogenetics testing was performed with a custom TaqMan OpenArray and a QuantStudio 12K Flex (ThermoFisher Scientific) with DNA normalized to 50 ng/mL. Results were analyzed using Genotyper Software version 1.3 and allele calls were made based on the combination of variants detected and consensus nomenclature. For drugs administered to at least 30 NICU patients, pharmacogenes were evaluated for variant allele frequency and potential clinical relevance. Of the 132 patients enrolled, 81 were male. Average length of stay was 47 days. Of the 64 different medications prescribed to at least one infant, for 15 of them are described clinical annotations with specific pharmacogenes by the PharmGKB and/or Clinical Pharmacogenetics Implementation Consortium. Of those 15 medications, three were administered at least three times to at least 30 infants: caffeine (n=102), morphine (n=56), and lorazepam (n=31). The associated pharmacogenes were CYP1A2 (caffeine); CYP3A4, OPRM1, COMT, ABCB1 (morphine); and UGT2B15 (lorazepam). Relevant to the rate of clearance of caffeine, rs2069514 (CYP1A2) was heterozygous in 18% of patients, and homozygous in 5%. Relevant to the rate of clearance of morphine, CYP3A4*22 was heterozygous in 5%. The common variants associated with response to morphine, c.472G>A (COMT) and c.118A>G (OPRM1) were heterozygous in 43% and 26%, and were homozygous in 22% and 8%, respectively. The transporter variant ABCB1*2 was heterozygous in 6% and homozygous in 11%. Relevant to the rate of clearance of lorazepam, UGT2B15*2 (rs1902023) was heterozygous in 45%, and homozygous in 31%. Correlation of pharmacogenetic results to specific clinical phenotypes is currently in progress. Pharmacogenetic variants relevant to commonly prescribed medications in the NICU are prevalent. Further clinical correlation of pharmacogenetic results may help inform selection and dosing of medications for critically ill NICU infants. This research may also facilitate detection and definition of new drug-gene associations unique to this patient population.