OC 23.3 - Concurrent use of novel antidiabetic drugs and risks of adverse events among atrial fibrillation patients receiving chronic anticoagulation

Background: Novel antidiabetic drug use has not yet been much investigated in atrial fibrillation (AF) patients.

Aims: To examine associations between novel antidiabetic drug use and adverse events in anticoagulated AF patients.

Methods: Using nationwide data from Statistics Netherlands, patients with prevalent type 2 diabetes (T2DM), AF, and anticoagulation, who started dipeptidyl peptidase 4 inhibitor (DPP-4i), glucagon-like peptide-1 analogue (GLP-1a), or sodium-glucose co-transporter 2 inhibitor (SGLT2i) in 2017-2021 were identified, constituting three cohorts with novel antidiabetic drug use. Three control cohorts without receiving these antidiabetic drugs were further selected from individuals with T2DM, AF, anticoagulation, and matched demographic characteristics/valvular heart diseases. During follow-up until 31/12/2021, incidences of ischemic stroke/transient ischemic attack/thromboembolism (IS/TIA/TE), major bleeding, myocardial infarction (MI), heart failure (HF), and all-cause mortality were determined. The associations between novel antidiabetic drug use and the outcomes were evaluated by multivariable Cox regression adjusting for CHA2DS2-VASc and HAS-BLED score, antiplatelet and other antidiabetic drug use, and prior diagnoses of overweight/obesity, renal diseases, MI, IS, HF, hypertension, major bleeding, and cancer.

Results: Baseline characteristics differed by types of novel antidiabetic drug (Table). DPP-4i was associated with increased IS/TIA/TE hazard (aHR=1.54, 95%CI 1.22-1.94; 1-year risk: 2.6% versus 1.8%). GLP-1a was associated with decreased hazard of HF (aHR=0.82, 95%CI 0.70-0.96; 1-year risk: 8.0% versus 8.5%) and all-cause mortality (aHR=0.62, 95%CI 0.51-0.74; 1-year risk: 3.4% versus 6.3%). SGLT2i appeared to be associated with decreased hazard of IS/TIA/TE (aHR=0.74, 95%CI 0.54-1.02; 1-year risk: 1.4% versus 1.8%) and major bleeding (aHR=0.74, 95%CI 0.58-0.94; 1-year risk: 3.0% versus 3.2%), but increased HF hazard (aHR=1.31, 95%CI 1.17-1.46; 1-year risk: 19.6% versus 9.2%) (Figure).

Conclusion(s): GLP-1a or SGLT2i (versus non-users) did not seem to be associated with ischemic stroke/major bleeding in anticoagulated AF patients, but DPP4i users presented a higher ischemic stroke risk. Since residual confounding cannot be excluded, further confirmations are needed, ideally by trials.