p.27 抽出テキスト

The DEVOTE (A Trial Comparing Cardiovascular Safety of Insulin Degludec Versus Insulin Glargine in Subjects With Type 2 Diabetes at High Risk of Cardiovascular Events) trial, a randomized, double-blind comparison of the ultra-long-acting, o.d. insulin degludec vs. insulin glargine U100, enrolled 7637 patients with T2DM with ASCVD or at high CV risk.181 Over a median follow-up of 1.8 years, there was no significant difference in the primary composite of CV death, non-fatal MI, or non-fatal stroke between groups. A significantly lower frequency of hypoglycaemia was observed in the degludec arm compared with the glargine arm.181

5.3.2.4 Glimepiride

Based on the findings of statistical non-inferiority of linagliptin vs. placebo in CARMELINA coupled with the non-inferiority of linagliptin vs. glimepiride demonstrated in CAROLINA, one might conclude that glimepiride is most likely not different from placebo with regards to CV safety.180 Thus, the long-lasting uncertainty about the CV safety of sulphonylureas may no longer be clinically relevant for glimeperide, at least in patients with a shorter duration of diabetes like those enrolled in the CAROLINA trial (median duration of T2DM ∼6 years).182

5.3.3 Cardiovascular considerations of older

glucose-lowering medications not tested in dedicated cardiovascular outcomes trials 5.3.3.1. Metformin Despite its long history as the recommended first-line treatment of hyperglycaemia for patients with T2DM, there have been no dedicated randomized trials to rigorously assess the CV safety or efficacy of metformin. Randomized trials that have reported CV outcomes with metformin are most-commonly limited by small sample sizes and few CV events for analysis, yielding low statistical power and substantially uncertain statistical precision of the estimates. The largest randomized trial with the most encouraging CV results for metformin was a nested randomized trial of 753 patients in the UKPDS who were overweight or obese at trial entry, comparing conventional glucose targets with a policy of intensive glucose lowering with metformin.183 In overweight and obese patients with newly diagnosed T2DM without previous CVD, metformin reduced MI by 39%, coronary death by 50%, and stroke by 41% over a median period of 10.7 years. However, with only 39 MIs and 16 coronary deaths in the metformin arm of the UKPDS, the precision of these efficacy estimates is largely uncertain. Initial randomization to metformin in the UKPDS was also associated with a lower incidence of MI and longer survival during an additional 8–10 years of passive follow-up.132 In meta-analyses of 13 randomized clinical trials evaluating the CV effects of metformin vs. placebo or active control, including the data from the UKPDS, none of the differences in assessed CV outcomes was statistically significant.184 The pooled HRs (95% CIs) were: all-cause mortality 0.96 (0.84–1.09); CV death 0.97 (0.80–1.16); MI 0.89 (0.75–1.06); stroke 1.04 (0.73–1.48); and peripheral vascular disease 0.81 (0.50– 1.31). While failing to demonstrate CV efficacy, the upper confidence limits of each of the outcomes analysed provide reassurance on the CV safety of metformin. Given the inconclusive results regarding the CV effects of metformin outlined above, metformin should not be a prerequisite for considering SGLT2 inhibitor or GLP-1 RA treatment for CV benefits. However, most patients in CVOTs with SGLT2 inhibitors or GLP-1 RAs were treated with metformin. Therefore, in patients already prescribed metformin, SGLT2 inhibitors and/or GLP-1 RAs should be added, independent of the need for additional glucose control. In patients with T2DM and ASCVD not treated with metformin, an SGLT2 inhibitor and/or GLP-1 RA should be given first line, and metformin should be considered for those who thereafter warrant additional glucose control. This Class IIa recommendation for metformin is based on the weight of opinion rather than the weight of evidence; results from meta-analyses of observational studies suggest associations with better CV outcome, but this is not supported by results from meta-analyses restricted to randomized trials in patients with T2DM and ASCVD, where no statistically significant effect of metformin has been observed for any major CV outcome.184,185

In patients without ASCVD or severe TOD at low or moderate CV risk, treatment with metformin should be considered based on the metformin data from the randomized sub-group with overweight or obesity from the UKPDS.183 For patients without ASCVD or severe TOD at high or very CV high risk, treatment with metformin may be considered based on expert consensus of the Task Force.

5.3.3.2 Sulphonylureas

Excepting glimepiride, which was assessed for CV safety and efficacy head-to-head against linagliptin in the CAROLINA trial, and gliclazidemodified release, which was compared with usual care (that could have included treatment with a sulphonylurea other than gliclazide) in the ADVANCE trial, dedicated CV safety assessments have not been conducted for the other sulphonylureas.132,173,174,184 In the UKPDS, which enrolled patients with newly diagnosed T2DM, the sulphonylureas chlorpropamide and glibenclamide (also known as glyburide) did not have statistically significant effects on CV outcomes, but importantly, no concerning signal of CV risk was observed.127 Likewise, in the ADVANCE trial evaluating more intensive glucose control vs. usual targets, patients randomized to the more intensive arm were randomized to treatment with gliclazide-modified release.128 While the gliclazidebased more intensive control strategy did not significantly improve CV outcomes, there were no major CV safety concerns observed. The relative CV safety of gliclazide and glimepiride is somewhat supported by results of contemporary real-world data analyses.186

5.3.4 Special considerations

5.3.4.1. Hypoglycaemia and cardiovascular risk Results from numerous studies have demonstrated associations between hypoglycaemia and CV events, with substantial uncertainty about whether these relationships are causal or simply associations. Results from randomized trials challenge a causal relationship between hypoglycaemia and adverse CV outcomes. For example, insulin degludec compared with glargine in the DEVOTE trial reduced the risk of hypoglycaemia, yet this did not translate into any difference in CV risk.181 Likewise, in the CAROLINA randomized trial, while glimepiride was associated with significantly more hypoglycaemia than the DPP-4 inhibitor linagliptin, MACE did not differ between the two randomized groups.180 The results of these two trials challenge, to some degree, the premise that avoiding hypoglycaemia may improve CV risk. In analyses of data from the TECOS randomized trial, which compared sitagliptin with placebo, hypoglycaemia events were independently associated with subsequent CV events, but importantly, the converse was also true.139 A non-fatal CV event was independently associated with subsequent hypoglycaemia. Similar results were confirmed in other trials.140,187,188

ESC Guidelines                                                                                                                                                                                          4069

Downloaded from academic.oup.com/eurheartj/article/44/39/4043/7238227 by guest on 07 October 2026