Pharmacogenomic (PGx) data from Thailand have been limited, and population-level genotype surveys often do not quantify how many individuals with actionable genotypes are actually exposed to relevant medications. This study profiled 4,662 Thai adults using an SNP-array optimized for Asian populations together with a prespecified PGx panel of 11 genes and 26 markers. The primary objective was to assess genotyping callability and the prevalence of CPIC level A/B actionable phenotypes, and to link those results to hospital electronic medical record (EMR) prescription and dispensation data to quantify real-world, or "realized," actionability.
Participants were genotyped using an SNP-array platform and a pre-specified pharmacogenetic panel consisting of 11 genes and 26 genetic markers. A hybrid required/optional calling policy guided diplotype and phenotype assignment from the array data. The array was described as Asian-optimized, enabling scalable PGx phenotyping in this Thai adult cohort.
Overall callability across reported gene-results was 98.62%, indicating that the array and calling policy produced interpretable genotypes for the large majority of participants. Most genes had callability exceeding 99%, with two exceptions noted in the report: CYP2C19 had callability of 95.99% and NUDT15 had callability of 90.28%.
Among nine genes for which phenotype coding was reported, 95.99% of participants carried at least one CPIC-actionable result. The median number of actionable results per individual was 2 (interquartile range 2–3). The most prevalent actionable phenotypes in the cohort were CYP3A5 (58.54%) and CYP2C19 (56.67%), followed by ABCG2 (45.10%) and UGT1A1 (27.37%). These prevalence estimates reflect the distribution of actionable pharmacogenetic phenotypes in this Thai adult sample as determined by the array and panel used.
Genotype results were linked to hospital EMR prescription and dispensation records to quantify how many participants with actionable genotypes had actual exposure to relevant medications. EMR linkage identified 1,529 participants, representing 32.58% of the cohort, who had been exposed to at least one medication included in the study’s medication list. This linkage allowed the investigators to move beyond genotype-only prevalence and estimate the subset of individuals for whom a genotype-informed prescribing decision could have been relevant in routine care.
The most commonly recorded study medications in EMR data were the proton pump inhibitor omeprazole (n = 658) and statins, with atorvastatin (n = 606) and simvastatin (n = 603) both frequent. Among users of these medications, a substantial proportion had actionable phenotypes: for CYP2C19–omeprazole, 55.02% of omeprazole users carried an actionable CYP2C19 phenotype; for SLCO1B1–statins, actionable SLCO1B1 phenotypes were present in 21.95–23.05% of statin users.
By combining population genotype frequencies with EMR-derived medication exposure, the study highlights specific high-yield targets for pre-emptive PGx implementation in Thailand. The findings indicate particular priority for CYP2C19 testing in the context of proton pump inhibitor prescribing and for SLCO1B1 testing related to statin therapy. The Asian-optimized SNP array demonstrated feasibility for scalable phenotyping that can be linked to EMR systems to identify such priority targets.
The investigators conclude that an Asian-optimized SNP array supports large-scale pharmacogenomic phenotyping in Thai adults and that EMR linkage provides a practical measure of realized actionability. The combined approach both quantifies the burden of CPIC-actionable phenotypes and identifies common, real-world medication exposures where genotype-informed prescribing could have immediate impact, notably proton pump inhibitors and statins.
The authors declared that no competing interests exist. The abstract notes that the article is open access under a Creative Commons Attribution License. Specific methodological details beyond the abstract and full data availability statements were not reported in the abstract text provided here.