A Longitudinal HbA1c Model Elucidates Genes Linked to Disease Progression on Metformin

Area of Science:

• Pharmacogenomics
• Metabolic Disorders
• Computational Biology

Background:

• Metformin is a first-line treatment for type-2 diabetes, but approximately one-third of patients exhibit poor long-term response.
• The influence of genetic and non-genetic factors on metformin's long-term efficacy remains incompletely understood.
• Identifying predictors of treatment response is crucial for optimizing patient outcomes.

Purpose of the Study:

• To identify genetic and demographic predictors of long-term glycemic control (HbA1c) in type-2 diabetes patients treated with metformin.
• To develop a computational model for predicting individual patient response to metformin.
• To investigate the impact of specific genetic variants on disease progression.

Main Methods:

• Combined nonlinear mixed-effect modeling with computational genetic analysis.
• Analyzed genetic data (12,000 variants in 267 genes), demographic information, and long-term HbA1c levels from 1,056 patients.
• Utilized statistical methods to identify significant genetic variants and clinical factors associated with treatment response.

Main Results:

• Nine genetic variants accounted for roughly one-third of the variability in the disease progression parameter.
• Serum creatinine, age, and weight influenced symptomatic response but explained minimal variability.
• Specific single nucleotide polymorphisms (SNPs) in CSMD1 (rs2617102, rs2954625) and SLC22A2 (rs316009) genes were found to impact disease progression, with minor alleles correlating with different outcomes.
• Minor alleles in CSMD1 were associated with less favorable outcomes, while the minor allele in SLC22A2 was linked to more favorable outcomes.

Conclusions:

• Genetic factors play a substantial role in determining long-term HbA1c response to metformin in type-2 diabetes.
• The identified SNPs in CSMD1 and SLC22A2 are potential biomarkers for predicting treatment outcomes.
• The developed computational model, upon validation, could enable personalized metformin treatment strategies for type-2 diabetes patients.