Pharmacogenomic genome-wide association study of LDL response to statins using high-throughput electronic health records analysis

Authors

  • SB Goleva
  • JM Keaton
  • BJ Waxse
  • A Williams
  • AH Awan
  • JC Denny

Abstract

We developed a high-throughput, generalizable self-controlled case series method to determine drug effectiveness for pharmacogenomic analysis using electronic health record (EHR) data from the All of Us Research Program (All of Us). To demonstrate the utility of this method, we applied it to determine HMG-CoA reductase inhibitor (statin) effectiveness, which we quantified using the difference in LDL laboratory values before and after statin prescription. We identified any pharmacogenetic variants associated with statin effectiveness by conducting a trans-ancestry genome-wide association study (GWAS) adjusted for age at statin prescription, sex at birth, ten principal components, and baseline LDL value prior to statin initiation.

The mean reduction in LDL following statin prescription in All of Us was 28.3 mg/dL (95% CI = 27.1 – 29.4; n = 4,695). Our GWAS (n = 4,035) revealed a significant association, rs7412 (β= 7.19, SE = 1.32, p = 5.9e-09), a variant within APOE previously associated with hypercholesterolemia and atorvastatin response. Another significant association, rs73151468 (Beta = -8.35, SE = 1.52, p = 4.2e-08), resides within a lncRNA gene, ENSG00000234223. Data from the Gene-Tissue Expression (GTEx) project suggests ENSG00000234223 is only expressed in the liver, which may be relevant since lipogenesis occurs in the liver.

We used EHR data to determine statin effectiveness and identified genetic variants associated with statin response. Replication of the pharmacogenomic association in APOE validates the accuracy of this method, and a novel association demonstrate the possibility to discover clinically and mechanistically relevant pharmacogenomic targets.

Scientific Focus Area: Genetics and Genomics

This page was last updated on Tuesday, August 6, 2024