BackgroundRisk of atherothrombotic events is not uniform in patients with type 2 diabetes mellitus (T2DM). Tailored risk assessment may help guide selection of pharmacotherapies for cardiovascular primary and secondary prevention. ObjectivesThe purpose of this study was to develop a risk model for atherothrombosis in patients with T2DM. MethodsWe developed and validated a risk model for myocardial infarction (MI) or ischemic stroke (IS) in a pooled cohort of 42,181 patients with T2DM from 4 TIMI (Thrombolysis In Myocardial Infarction) clinical trial cohorts. Candidate variables were assessed with multivariable Cox regression, and independent variables (P < 0.05) were retained in the final model. Discrimination and calibration were assessed. Treatment interactions with dapagliflozin (sodium-glucose cotransporter-2 inhibitor) and evolocumab (proprotein convertase subtilisin/kexin type 9 inhibitor) were explored in the DECLARE-TIMI 58 (Dapagliflozin Effect on CardiovascuLAR Events-Thrombolysis In Myocardial Infarction 58) and FOURIER (Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk) trials, respectively. ResultsSixteen variables were independent predictors of MI or IS. The model identified a >8-fold gradient of MI or IS rates between the top vs bottom risk quintiles in the validation cohort (3-year Kaplan-Meier rate: 14.9% vs 1.4%; P < 0.0001). C-indexes were 0.704 and 0.706 in the derivation and validation cohorts, respectively. The model was well-calibrated in both primary and secondary prevention. Absolute reduction in the rates of MI or IS tended to be greater in patients with higher baseline predicted risk for both dapagliflozin (absolute risk reduction: 2.1% vs 0.2%) and evolocumab (absolute risk reduction: 3.2% vs 1.0%). ConclusionsWe developed and validated a risk score for atherothrombotic events, leveraging 16 routinely assessed clinical variables in patients with T2DM. The score has the potential to improve risk assessment and inform clinical decision-making.
Read full abstract