MON-564 Sulfonylurea efficacy in type 2 diabetes according to functional variants of candidate genes identified from SU-dependent patients - a pilot study

Abstract Disclosure: H. Son: None. S. Lee: None. J. Lee: None. M. Seong: None. H. Jung: None. Introduction & Objective: As genetic factors have a critical role in insulin secretion, sulfonylureas (SUs) offer potential for pharmacogenetic applications in type 2 diabetes (T2D). Previously, we reported whole exome sequencing (WES) analysis on SU-dependent T2D patients (ENDO 2024, abstract #7199), identifying 10 candidate genes with enriched functional variants in those with exceptional SU responsiveness. This pilot study explores whether any functional variants in these 10 genes predict glycemic responses to SUs in broader T2D population. Method: Participants were retrospectively enrolled from the Korea Biobank Network (IRB # H-2310-141-1479). Eligible participants were adults treated with SUs for more than 1 year. Individuals with conditions affecting drug prescription or glycemic control, such as systemic corticosteroid use or eGFR <45 mL/min, were excluded. Buffy coat samples underwent WES using the Agilent SureSelect V8 kit and Illumina NovaSeq 6000. Reads were aligned to hg19 with bwa-mem and processed per GATK guidelines, including variant calling with HaplotypeCaller. Variants were annotated using ANNOVAR. Changes in HbA1c levels following SU initiation and the proportion of participants achieving HbA1c levels < 7.0% were analyzed based on the number of candidate variants. Results: Forty participants were enrolled (the mean age 60 ± 9 years, male 75%, and the mean baseline HbA1c 8.5 ± 0.9%), and 33 (82.5%) carried at least one candidate variant. The median number of variants per individual was 2 (range: 0-4). Participants with two or more variants (n=21, high-variant group) exhibited a nominally greater reduction in HbA1c in 6 months compared to those with zero or one variant (n=19, low-variant group) (-1.74 vs. -1.46%-points, p=0.074). There was no significant baseline characteristics between the 2 groups. In logistic regression analysis for achieving an HbA1c below 7.0% at 1 year with clinical variables including age, sex, eGFR, and baseline HbA1c, -2 Log-Likelihood was 24.11. High-variant group achieved this goal in 81% and low-variant group in 68.4%. The inclusion of high-variant group as a variable in the logistic regression model slightly improved -2 Log-Likelihood to 23.18, with an odds ratio of 3.10 (p=0.175). Conclusion: This pilot study among Korean individuals with T2D provides preliminary evidence on the genetic determinants of SU efficacy, which were identified from SU-dependent patients. Expanding the cohort and developing a variant scoring system could lead to significant results, contributing to personalized treatment strategies for T2D. Funding: NRF (2022R1A2C2 004570), South Korea. Presentation: Monday, July 14, 2025