CVOTs applicability on SGLT-2i in a real world population of type 2 diabetes patients

OBJECTIVE OF THE STUDY To evaluate the generalizability of the results of recent CVOTs on SGLT2i (EMPA-REG OUTCOMES, CANVAS, DECLARE-TIMI 58 and VERTIS-CV) in patients with T2DM in the real world. DESIGN AND METHODS Database of AMD Annals 2018 was used in this study, including all patients cared for by 222 diabetes centers during 2016. Starting from the eligibility criteria adopted in the different trials, the analysis aimed to identify the subjects potentially eligible for each trial, compare their characteristics with those of the population recruited in the trials, and evaluate the current use of these drugs among the potentially eligible patients registered in AMD Annals database. RESULTS The evaluable cases (i.e. presence of information on all eligibility criteria) ranged from 149.064 for the CANVAS to the 342.205 subjects for the EMPAREG-OUTCOME. Overall, the eligible patients in AMD Annals ranged from 40,039 for the EMPAREG OUTCOME study (11.7%) to 144,166 (55.9%) for DECLARE-TIMI 58. The percentage of patients actually treated ranged from 4.4% (6.373) for DECLARE-TIMI 58 to 6.6% (2.917) for CANVAS. As compared to RCTs’ populations, AMD Annals population was older, included a larger proportion of women, showed slightly lower BMI and better metabolic control, and longer diabetes duration. Prevalence of major cardio-cerebrovascular events was lower, while the percentage of subjects with peripheral vasculopathy was higher; the proportion of subjects with reduced glomerular filtrate or albuminuria and diabetic retinopathy was higher. CONCLUSIONS The generalizability of CVOTs’ results to the real world patients is only partial, because the populations are quite different. A very low clinical use of SGLT-2i is documented. This is at least in part due to the strict AIFA criteria for reimbursement, which reduce the proportion of patients who could benefit from treatment with SGLT2i on major cardiovascular events and mortality; however, an identical advantage of the treatment obtained in RCTs cannot be hypothesized in the AMD Annals population. KEY WORDS type 2 diabetes; CVOT; real world; generalizability; SGLT2i