Performance evaluation of CVD risk models in T1D in the CARDT1 Study

Introduction and aims: Cardiovascular disease (CVD) is the leading cause of morbidity and mortality among individuals with type 1 diabetes (T1D). Nevertheless, assessing CV risk in T1D patients is still a challenging issue. Different prediction models have been proposed by guidelines; our aim is to evaluate the performance of different risk models in predicting the presence of asymptomatic coronary or carotid artery disease. Methods: A cross-sectional analysis was conducted in an Italian cohort of T1D patients without history of CVD who underwent coronary CT scan and ultrasound assessment of carotid intima-media thickness (IMT). Ability of CV risk models (diabetes duration, 2019 ESC Guidelines, Steno Type 1 Risk Engine-ST1RE or Scottish/Swedish CVD risk prediction tool-Swe/Sco) to predict atherosclerotic disease (presence of coronary plaque and/or IMT >1.3 mm) was quantified by the area under the receiver operating characteristic curve (AUC). Results: Eighty patients were included in the analysis (50% females) with mean age of 45 (±13) years and median diabetes duration of 21.5 (13.5;33.5) years, in good glycaemic control (mean HbA1c 53.3±9.7 mmol/mol), with BMI 24.4±3.6; 25 (31.3%) active or former smokers, 30 (37.5%) with one or more microvascular complications. Coronary CT scan showed coronary disease in 38 (46.2%) patients; IMT >1.3 mm was detected in 39 (48.8%) patients; 53 (66.3%) patients had either one and/or the other condition. Comparing AUC values, the prognostic accuracy of ST1RE and Swe/Sco tool in discriminating the presence of CAD was similar, with maximum effect on the combined endpoint (0.961 and 0.946, respectively; figure), significantly outperforming the predictive capacity of diabetes duration or CV risk stratification according to the 2019 ESC guidelines. Conclusion: In this cohort, both ST1RE and Swe/Sco tool showed reliability in detecting high risk patients when increased IMT and presence of coronary plaque are considered as proxy of cardiovascular risk. These results prove particularly useful in driving primary prevention strategies and diagnostic workup in asymptomatic T1D patients.