Among four classes of drugs used to treat type 2 diabetes (T2D), the GLP-1 receptor agonists (RAs) appeared to be most protective against major adverse cardiovascular events (MACE) — followed by SGLT2i, sulfonylureas, and DPP-4i — in a large comparative effectiveness study of adults with T2D.

“These new classes of drugs, GLPs and SGLT2 inhibitors, have the potential not only to improve glucose control but also to reduce heart attacks, strokes, and to delay mortality. So it’s very important for clinicians to correctly match the patient to whichever drug will give them the maximum benefit,” lead author Patrick J. O’Connor, MD, told Medscape Medical News.

Most of the randomized trials of these drugs enrolled people with preexisting cardiovascular disease or high cardiovascular risk, he added. “This study is different in that we took everybody with T2D who was a new user of one of these four classes of medicines. They didn’t have to have high cardiovascular risk.”

Only about 20%-25% of people with T2D have cardiovascular disease, so it’s important to know what to do with the other patients, he noted.

GLP-1s Lead the Pack for Minimizing MACE

The study included 296,676 adults from six large US healthcare delivery systems initiated on one of the four drug classes between 2014 and 2021. A subset of 241,981 with T2D were included in an analysis emulating a four-arm clinical trial, with adjustments for over 400 time-independent and time-varying covariates. The study population was 45.7% female. Just under half (49.8%) were White, 35.3% Hispanic, and 10.1% Black.

Initiated drugs were sulfonylureas in 87.6%, SGLT2i in 6.1%, GLP-1 RAs in 3.8% (about 50% liraglutide, 20% semaglutide, or 20% exenatide), and DPP-4i in 2.4%. At least two thirds of patients in all four groups used metformin at baseline. Insulin was used by 11.9% overall, ranging from 60.9% among those in the GLP-1 RA initiators and 43.8% of the SGLT2i users to just 7.2% of the sulfonylurea users.

The proportions discontinuing the initiated treatment ranged from 48.4% for sulfonylurea to 21.9% for SGLT2i.

In the per-protocol analysis, adjusted MACE risk estimates varied across the four drug class groups. Average risk difference over 2.5 years of follow-up showed a decrease in MACE risk for GLP-1 RAs compared with SGLT2i (-0.006; P < .001), sulfonylureas (-0.013; P < .001), and DPP-4i (-0.015; P < .001) over time.

Comparing SGLT2i with GLP-1 RAs, the 2.5-year cumulative risk difference was 1.5%, and for DPP-4i vs sulfonylureas, 1.9%, both differences significant.

However, there was a wide variation in risk by subgroups of individuals. The decreased MACE risk estimates in sustained users of GLP-1 RAs vs SGLT2i were greatest among individuals with atherosclerotic cardiovascular disease (ASCVD, 5.3%), those with heart failure (6.1%), those aged 65 years or older (3.6%), and those with moderate baseline kidney impairment (3.0%). In contrast, there was no significant decreased MACE risk in sustained GLP-1 RA users compared with SLGT2i in people younger than 50 years.

“These results, along with consideration of collateral clinical effects, cost, and availability, may inform treatment decisions for adults with T2D,” the study authors wrote.

How Much Benefit Is Driven by Weight Loss?

Asked to comment, Rozalina G. McCoy, MD, Associate Division Chief for Clinical Research, Division of Endocrinology, Diabetes, and Nutrition, Department of Medicine, University of Maryland School of Medicine, Baltimore, told Medscape Medical News that the authors “really did an incredibly rigorous and thoughtful job and were also very transparent in their analyses and assumptions, which I think is really important to interpret any study, not just observational.”

However, McCoy pointed out several limitations in the data, including short follow-up, heterogeneity of the patient population, and a lack of analysis of factors such as heart failure or chronic kidney disease (CKD) outcomes, hypoglycemia, or the treatment effect for obesity. “I think that’s the elephant in the room, how much of the benefit is driven by weight loss,” she said.

She also noted that the extent and sequence of insulin use was unusual. “[Having] patients who are on insulin and then started on non-insulins is clinically a very interesting scenario. That’s not normally what we do,” she explained. “They don’t say how many were on basal insulin versus multiple daily injections. Do they have type 1 and were misdiagnosed? I think it would have been helpful to know a little bit more detail about the insulin users,” she said.

In addition, the finding that people with ASCVD derive greater benefit from GLP-1 RAs than from other agents reflects that of previous studies, but the findings for people with CKD and heart failure do not, said McCoy.

“We know from even direct comparisons and network meta-analyses of randomized controlled trials that for people with heart failure and CKD, SLGT2i are much, much better. So it was interesting that they did not find that. It was the most surprising finding to me. I think we need to better understand why they’re seeing what they’re seeing in order for it to really inform practice.”

The study was funded by the National Institutes of Health (NIH) and the Patient-Centered Outcomes Research Institute (PCORI), and O’Connor reported receiving other NIH grants. McCoy reported receiving research funding from NIH, PCORI, and the American Diabetes Association (ADA), and she reported receiving funding from ADA as co-chair of the Professional Practice Committee (but was not speaking for ADA in this story).

Miriam E. Tucker is a freelance journalist based in the Washington, DC, area. She is a regular contributor to Medscape Medical News, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X @MiriamETucker and BlueSky @miriametucker.bsky.social.