A research team has identified three genes that could help prioritize pharmacogenomic testing for older adults in China who receive multiple medications. An analysis of more than 3.3 million annual patient records found that approximately one in five included exposure to at least two drugs with established genetic prescribing guidance. Three genes—CYP2C19, SLCO1B1, and CYP2C9—covered medication exposure in at least 98% of these records. The findings provide a basis for prospectively evaluating targeted genetic testing to support more personalized medication management.
Older adults often receive several medications to manage chronic conditions, increasing the complexity of prescribing and the risk of adverse drug reactions. Pharmacogenomics examines how genetic differences influence drug response and can help clinicians interpret genetic results when selecting medicines or doses. Because inherited genetic profiles remain stable, testing results may inform prescribing decisions across different episodes of care. However, designing useful testing panels requires evidence about which relevant medicines patients actually receive. In China, limited evidence on exposure to multiple drugs with actionable genetic guidance has complicated efforts to prioritize genes and assess practical testing strategies for older populations.
A study (DOI: 10.48130/targetome-0026-0043 ) published in Targetome on 10 September 2026 by Xiaoyan Nie's team, Peking University Health Science Center, identifies frequent annual exposure to genetically relevant medications and a small core gene set that warrants further evaluation for targeted pharmacogenomic testing.
The researchers conducted a retrospective, cross-sectional analysis of independently sampled China Health Insurance Research Association claims from 2015 through 2017, including adults aged 65 years or older. They standardized medication names into active ingredients and matched them against Clinical Pharmacogenetics Implementation Consortium Level A drug–gene guidance updated in July 2026. The analysis retained 53 ingredients mapped to 19 pharmacogenes, with 18 genes represented in the multiple-exposure groups. Researchers defined the primary exposure as receiving at least two distinct relevant drugs within one calendar year and examined a stricter threshold of five or more. Repeated prescriptions for the same ingredient counted once annually. They compared medication patterns across years and economic regions, then ranked genes by exposure prevalence and calculated the additional patient-record coverage obtained as genes were added. Finally, they compared listed testing charges using an illustrative capped, per-locus fee schedule from Zhejiang Province. Among 3,309,025 patient-year records, 43.4% included at least one relevant drug, 20.2% included at least two, and 1.9% included five or more. Atorvastatin was the most frequent medication associated with the two-drug threshold, while pantoprazole led at the five-drug threshold. Statins and proton pump inhibitors were prominent therapeutic classes, with broadly consistent patterns across study years and sampled regions. CYP2C19 and SLCO1B1 together covered approximately 96% of records meeting the two-drug threshold; adding CYP2C9 increased coverage to at least 98%. Coverage indicated that a record contained at least one medication linked to an included gene, rather than complete genetic coverage of every prescribed drug. Under the selected fee schedule, a combined panel covering 11 specified loci had a listed charge of CNY 2,200, compared with CNY 2,800 for three separately billed gene tests. This comparison concerned listed charges alone and did not establish clinical benefit, overall healthcare savings, or cost-effectiveness in routine practice across other settings or populations.
The study offers a practical starting point for selecting genes for future testing studies. Its estimates describe unweighted insurance samples rather than national prevalence, and annual exposure does not necessarily mean simultaneous medication use. No individual genetic results or validated clinical outcomes were available. Prospective research using contemporary prescribing data will therefore be needed to determine whether the proposed gene set improves treatment decisions, patient safety, and value in clinical care.
