Abstract A052: Trans-ancestry Validation of a European-Derived Polygenic Risk Score for Statin Response in Young Korean Ischemic Stroke Patients: Results from the GENE_YAS Study
Eung-Joon Lee, Yeonju Yu, Lee Jc, Jonguk Kim, Nakhoon Kim, Do Yeon Kim, Jihoon Kang, Byung‐Soo Kim, Han Moon-Ku, Jun Yup Kim, Kangho Choi, Joon-Tae Kim, Hyunsoo Kim, Man Seok Park, Hee-Yun Chae, Kyu Sun Yum, Dong Ick Shin, Jung Hwa Seo, Dae Hyun Kim, Jae-kwan Cha, Dong‐Eog Kim, Dong-Eog Kim, J. Y. Park, Kyusik Kang, M. G. LEE, Mi-Sun Oh, Byung‐Chul Lee, Hong-Kyun Park, Yong Jin Cho, Keun-Sik Hong, Chul-Hoo Kang, Joonggoo Kim, Jay Chol Choi, Hyungjong Park, Jeong-Ho Hong, Sung-Il Sohn, Tai Park, S H Park, Jin-Kyo Choi, Wook-Joo Kim, Jee-hyun Kwon, Jeongyoon Lee, Kyungbok Lee, Lee Jc, Kyungmi Oh, Joon‐Tae Kim, Jung Hoon Han, Keon-Joo Lee, HaeBong Jeong, Kwang-yeol Park, Lee Jc, Hoonji Oh, Joohon Sung, Soo-Ji Lee, Jeeeun Kim, Hee-Joon Bae
Background&Purpose: Polygenic risk scores (PRSs) for statin-induced low-density lipoprotein cholesterol (LDL-C) response have been developed in European cohorts, but their utility in non-European populations remains uncertain. We investigated the transferability of a European-derived LDL-C PRS to Korean patients with early-onset ischemic stroke. Methods: We analyzed 2,730 patients aged <55 years with acute ischemic stroke or transient ischemic attack (TIA) admitted within 7 days of onset to 17 academic or regional centers between 2017 and 2023 through the CRCS-K-NIH Registry. This cohort corresponds to the GENE_YAS (Genetic Analysis in Young Age Stroke Patients) study, a prospective sub-study of the CRCS-K-NIH designed to investigate genetic determinants of early-onset stroke. A PRS for on-statin LDL-C response was constructed using 35 single nucleotide polymorphisms identified in European cohorts. Follow-up LDL-C levels were measured 1–6 months post-stroke. Patients were stratified into PRS quartiles, and baseline characteristics, the absolute and percentage LDL-C reduction were compared. Multivariable logistic regression models estimated the association between PRS and treatment goals (LDL-C <70 mg/dL or ≥50% reduction), adjusting for age, sex, and atorvastatin-equivalent dose. Results: Mean age was 47.0 years, 72.4% were male, and baseline LDL-C was 119.3±40.5 mg/dL. Follow-up LDL-C decreased progressively across PRS quartiles (Q1: 81.4 mg/dL; Q2: 79.3 mg/dL; Q3: 75.4 mg/dL; Q4: 70.2 mg/dL; p<0.001). Percentage reduction increased from 27.5% in Q1 to 32.8% in Q4 (p=0.018). Absolute reduction ranged from 44.9 mg/dL in Q1 to 50.7 mg/dL in Q4 (p=0.021). Patients in the highest PRS quartile (Q4) had more than twofold greater odds of achieving LDL-C <70 mg/dL or ≥50% reduction compared with Q1 (adjusted OR 2.15, 95% CI 1.73–2.67, p<0.001). Each 1-SD increase in PRS corresponded to an additional 2.96 mg/dL LDL-C reduction (95% CI 1.37–4.55, p<0.001), independent of statin intensity. Discussion: A European-derived PRS for on-statin LDL-C response was significantly associated with both absolute and relative LDL-C reduction in young Korean stroke patients These findings support the cross-ancestry transferability of pharmacogenomic prediction tools and highlight their potential role in optimizing lipid-lowering strategies for secondary stroke prevention.