Initiation of glucose-lowering drugs reduces the anticoagulant effect of warfarin – but not through altered drug metabolism in patients with type 2 diabetes

Abstract

Aims Drug metabolism might be altered in patients with type 2 diabetes, and we have previously shown that initiation of glucose-lowering drugs might reduce the efficacy of warfarin. We aimed to evaluate the impact of initiation of glucose-lowering drugs on warfarin efficacy and drug metabolism using a translational approach. Methods First, we conducted a register-based self-controlled cohort study on Danish and Scottish warfarin users. Individuals initiating a glucose-lowering drug during warfarin treatment were identified and warfarin efficacy (International Normalized Ratio (INR)) was compared before and after initiation of glucose-lowering drugs, overall and stratified by response (change in HbA1c). Second, we conducted a clinical pharmacokinetic trial comprising individuals with treatment-naive type 2 diabetes. In a self-controlled design, patients ingested probe drugs for drug-metabolizing enzymes (the Basel Cocktail) before initiating glucose-lowering treatment, and at follow-up after three and 12 weeks of treatment. Drug metabolism, glycemic control, and inflammation were assessed on each visit. Results: In the Danish and Scottish cohorts, initiating glucose-lowering drugs reduced warfarin efficacy (n=982 and n=44, respectively). INR decreased from 2.47 to 2.21 in the Danish cohort (mean difference -0.26; 95% CI -0.35;-0.17) and from 2.33 to 2.13 in the Scottish cohort (-0.21; 95% CI -0.52;0.11) after initiation of glucose-lowering treatment. This impact on INR was more pronounced among individuals with stronger effects of glucose-lowering treatment (HbA1c reduction > 10 mmol/mol) with mean differences of -0.44 (95% CI -0.71;-0.18) and -0.40 (95% CI -1.06;0.26) in Denmark and Scotland, respectively. In the clinical pharmacokinetic trial (n=10), initiating metformin did not affect drug metabolism after three and 12 weeks of metformin treatment (geometric mean ratio of CYP3A4 metabolic ratio: 1.12 (95% CI: 0.95;1.32) and 1.16 (95% CI: 0.93;1.44), respectively). Glycemic control improved during treatment, while inflammation remained low and unchanged during treatment. Conclusion Initiation of glucose-lowering drugs among chronic warfarin users is associated with a reduction in INR, particularly among individuals with a large decrease in HbA1c. However, this effect seems unrelated to CYP enzyme activity and warfarin drug metabolism. Registry number ClinicalTrials.gov identifier NCT04504045.