GLUCOSE LOWERING DRUGS OR STRATEGIES, MAJOR ADVERSE CARDIOVASCULAR EVENTS AND HEART FAILURE OUTCOMES, AND ASSOCIATION WITH WEIGHT LOSS - META-ANALYSIS OF LARGE CARDIOVASCULAR OUTCOME TRIALS

Glucose-lowering drugs or strategies (GLDS) have varied effects on major adverse cardiovascular events (MACE) and heart failure (HF) in cardiovascular outcomes trials. Mechanisms driving cardiovascular risk reduction remain elusive. We searched Ovid MEDLINE, the Cochrane Central Register of Controlled Trials databases, PubMed, and meeting abstracts up to April 2019 for large CV outcome randomized controlled trials of glucose-lowering drugs or strategies in patients with or at risk for type 2 diabetes mellitus (T2DM). The primary endpoints were atherothrombotic major adverse cardiovascular events (MACE) and HF. We derived pooled risk ratios (RRs) with random effects models, explored results stratified by baseline CVD subgroups, and performed meta-regression by weight differences achieved across treatment arms. Data from 28 GLDS CV outcome trials were included. Among a total 212,221 patients, median age was 63 years, 65% were male, with 64% CVD and 11% HF at baseline. Over a pooled weighted mean of 3.7 years of follow-up, 20,118 (10%) patients experienced a MACE and 7,442 (4%) a HF event. Across all trials overall, novel GLDS lowered the risk of MACE compared with standard care (RR 0.92, 95% CI, 0.89-0.95; P < 0.00001) without effect on HF (RR 0.99, 95% CI, 0.90-1.09; P=0.81). Across varied drug classes or strategies, the magnitude and directionality of the RR for MACE was consistent (P-interaction=0.06) but markedly varied for HF (P-interaction < 0.00001) with meta-regression showing a change in RR of HF by 6% (95% CI 3%-9%) per 1.0 kg weight gain or loss achieved between treatment groups (P=0.0006; Figure). Nine trials evaluated strategies that produce marked weight loss including intensive lifestyle changes, GLP-1 agonists or SGLT2 inhibitors. Among these, risk reduction for MACE was confined to patients with as opposed to without established CVD (RR 0.89 [95% CI, 0.84-0.95] vs RR 1.02 [95% CI, 0.91-1.15]; P-interaction=0.01) with consistent effects on HF (RR 0.80 [95% CI, 0.72-0.88] vs RR 0.76 [95% CI, 0.56-1.03]; P-interaction=0.74). In >200000 patients in CV trials, magnitude and directionality of MACE benefit is consistent but markedly varies for HF with meta-regression showing a change in HF relative risk by 6% (3%-9%) per 1 kg weight gain/loss between treatment arms. Among diabetes drugs or strategies that lower weight, there is robust risk reduction in atherothrombotic and heart failure events with a consistent HF reduction among patients with and without baseline CVD whereas MACE reduction was confined to patients with established CVD.