Abstract 013: Using the Diabetes Collaborative Registry (DCR) to Estimate the Potential Real-world Impact of the Iris Trial on Improving Outcomes in Patients With Cerebrovascular Disease

Abstract

Background: Thiazolidinediones use has markedly decreased since concerns arose regarding the CV safety of rosiglitazone. Recently, however, the IRIS trial demonstrated that pioglitazone reduced the risk of MI and stroke in patients with recent stroke or TIA and insulin resistance but without overt type 2 diabetes (T2D). We estimated the current use and potential impact of pioglitazone in patients enrolled in the Diabetes Collaborative Registry (DCR). Methods: DCR is a US-based outpatient registry of patients with diabetes or prediabetes seen in cardiology, endocrinology, and primary care practices and currently encompasses 374 practices and 5114 providers. Patients ≥40 years with HbA1c <7% on no diabetes medications (other than pioglitazone), prior stroke or TIA, and no history of heart failure were considered potentially eligible for IRIS. Although IRIS determined insulin resistance based on HOMA-IR levels, these were not available in DCR, and thus HbA1c levels were used as surrogates. Results: Among 242,590 patients in DCR with HbA1c data, 58,466 had prediabetes or diet-controlled T2D. Of these, 6,772 (12%) were deemed potentially IRIS-eligible. Mean age was 70 y, 52% were men, and mean HbA1c was 6.0%. Pioglitazone was used in 146 of these patients (2.2%); rates were slightly higher among patients seen by cardiologists vs. endocrinologists vs. primary care (2.6% vs. 1.4% vs. 0.0%, p<0.001). Patients treated (vs. not) with pioglitazone were more likely to be older (73 vs. 70 y), have higher body mass indices (32 vs. 30 kg/m 2 ), higher HbA1c (6.1 vs. 5.9%), and concomitant coronary artery disease (69% vs. 56%) (all p<0.02). If all potentially eligible patients in DCR were treated with pioglitazone for 4.8 years, 93 strokes and 86 myocardial infarctions may have been prevented, assuming similar risk reductions in this “real-world” patient population as observed in IRIS. Conclusion: In a large US-based outpatient registry, we found that 12% of outpatients with prediabetes or diet-controlled T2D met the main eligibility criteria for IRIS, partly due to the narrow inclusion criteria of the trial but also likely reflective of a general paucity of screening for prediabetes and insulin resistance in this cohort. Pioglitazone was rarely used in these patients but could have a substantial impact on CV outcomes in these eligible patients. The population impact could be even greater if the CV benefit of pioglitazone can be demonstrated in patients with cerebrovascular disease and overt T2D or even the broader population of patients with insulin resistance and cardiovascular disease. The effect of pioglitazone in these patients, who are also at high CV risk, should be the focus of future work.