Pediatric Pharmacogenomics from Whole-Exome Sequencing: Developmentally Appropriate Interpretation in 1159 Russian Children and Newborns

Background/Objectives: The application of pharmacogenomics (PGx) in pediatrics is limited by the lack of age-oriented interpretation, as adult algorithms do not account for ontogenetic changes in drug-metabolizing enzymes and transport proteins. We aimed to evaluate the clinical applicability of PGx data in Russian children, to assess directional interaction between genotype-based recommendations and enzyme ontogeny, and develop recommendations for age-oriented PGx reporting. Methods: We analyzed whole-exome sequencing data from 524 pediatric patients and 635 newborns. Pharmacogenomic annotations were filtered according to PharmGKB/ClinPGx evidence levels (1A–2B) and the “Pediatrics” tag. In 100 pediatric patients, medical records were retrospectively reviewed for pharmacotherapy and adverse drug reactions (ADRs). An allele-frequency-weighted ADR cost index (AFW-ADR) was calculated for 25 gene–variant–drug–ADR associations (in pediatric cohort), with cost anchors graded by proxy quality. A utilization-weighted variant (AFW-ADR × fdrug) was additionally calculated. Results: Clinically relevant annotations accounted for 5% of initial annotations in both cohorts; 67.6% and 67.2% of these in pediatric and neonatal cohorts, respectively, involved risk alleles with altered function. ADRs occurred in 21% of pediatric patients; only 2/21 cases were explained by high-evidence annotations. Genotype-based recommendations and neonatal enzyme ontogeny were directionally aligned in 5 of 11 assessable gene–drug pairs (45.5%). In the AFW-ADR ranking, the highest values were MTHFR rs1801133–methotrexate toxicity ($22,368), CYP2C9*2/*3–phenytoin SJS/DRESS/TEN ($10,020 and $9540), UGT1A1*28–irinotecan neutropenia ($4658), and HLA-A*31:01–carbamazepine SJS/DRESS/TEN ($3720); however, MTHFR is not clinically actionable. After weighting by pediatric drug utilization, methotrexate remained first, carbamazepine rose to the top ranks, followed by phenytoin, while UGT1A1*28–irinotecan fell to the lowest ranks. This index does not incorporate penetrance and should be interpreted as an exploratory prioritization framework. Conclusions: Age is critical for pediatric PGx interpretation because current reporting approaches do not adequately incorporate ontogeny. We propose an age-oriented PGx model incorporating patient age, ontogenetic adjustment, evidence-level stratification, and multidisciplinary clinical assessment, requiring prospective validation.