The drug is often used not because it’s optimal, but because it’s cheap, says one group. That shouldn’t be the case.
Moderate-risk patients with diabetes treated with a sulfonylurea appear to be at a higher risk of major adverse cardiovascular events than those who receive a dipeptidyl peptidase-4 (DPP-4) inhibitor, with glipizide carrying the greatest hazard, according to a comparative-effectiveness study.
The higher absolute risk of myocardial infarction, ischemic stroke, heart failure hospitalization, or cardiovascular death with the sulfonylurea class was relatively modest at 1.1%, but investigators say the results still have clinical implications.
“The relative increase [in risk] was 13% for glipizide and smaller than that for glimepiride and glyburide,” lead investigator Alexander Turchin, MD (Brigham and Women’s Hospital, Boston, MA), told TCTMD. “We cannot speak to a high-risk population, but if the increase in risk is similar in the high-risk population, then glipizide is probably not really a good medication to use. I personally am probably going to abstain from using glipizide in the high-risk population. In the moderate-risk population, it’s probably still not going to be my first choice.”
In the treatment of type 2 diabetes, there are a number of different medications available after first-line metformin therapy. For those without established atherosclerotic cardiovascular disease (ASCVD) or chronic kidney disease, there are DPP-4 inhibitors, glucagon-like peptide-1 (GLP-1) receptor agonists, sodium-glucose cotransporter 2 (SGLT2) inhibitors, and thiazide diuretics, all of which can be used as second-line therapy depending on treatment goals. GLP-1 receptor agonists and SGLT2 inhibitors have been shown to reduce cardiovascular and kidney complications, and lead to weight loss, in addition to reducing HbA1c.
Sulfonylureas, on the other hand, are usually recommended in cases when cost is a barrier.
“Diabetes care has changed, and sulfonylureas are not as prominent as they used to be,” said Turchin. “They used to be the second most common medication after metformin, but they still play an important role because they remain cheap and effective. For people who may have difficulty paying for the more expensive modern medications, they are still a common option.”
Despite their use, the drug class is known to increase the risk of hypoglycemia and result in weight gain. Tolbutamide, which is no longer available in the US, was linked to a higher risk of cardiovascular events compared with insulin. While there are some observational studies with other sulfonylureas, the cardiovascular safety data are limited, said Turchin. Glyburide is known to be associated particularly with hypoglycemia and is less frequently used today.
I personally am probably going to abstain from using glipizide in the high-risk population. Alexander Turchin
In an editorial accompanying the study, Alexandria Ratzki-Leewing, PhD, Chixiang Chen, PhD, and Rozalina McCoy, MD (all University of Maryland School of Medicine, Baltimore), say the new findings are “disquieting” given that glipizide remains the most prescribed sulfonylurea in the US. While a 1.1% absolute increase in risk over 5 years might seem small, “the population impact is considerable,” they write.
“In reality, glipizide is often chosen not because it is optimal, but because it is affordable,” according to the editorialists. “With newer cardioprotective agents priced out of reach for many, a drug available for under $50 annually is liable to become the default. This creates a double bind: clinicians face pressure to contain costs, while patients absorb the risk.”
Modest but Significant Risk
Using data from the Observational Evaluation of Second-Line Therapy Medications in Diabetes (BESTMED) study, the analysis included people with diabetes who were treated at one of 10 US health systems or who were part of one of two national health insurance plans. All patients (median age 61 years; 47.1% female) received metformin for at least 3 months, and while none had a history of coronary artery disease, they were at moderate risk of ASCVD. The median HbA1c level at baseline was 7.8%.
Of the 48,165 eligible patients in BESTMED, 18,147 were started on glipizide, 14,282 on glimepiride, and 1,887 on glyburide. The remaining 13,849 patients were started on a DPP-4 inhibitor, which served as the reference group. DPP-4 inhibitors are clinical alternatives to sulfonylureas and they have not been shown to increase the risk of cardiovascular events compared with placebo, say investigators.
Over a median of 37 months, 6.6% of patients had a MACE, including 2.2% who had an MI, 2.6% who had an ischemic stroke, 3.0% who were hospitalized for heart failure, and 0.4% who died from any of those conditions. In all, 3.1% of patients died from noncardiovascular causes and 35.6% were lost to follow-up.
The estimated 5-year risk of MACE was 8.1% for those treated with a DPP-4 inhibitor. For those treated with a sulfonylurea, the 5-year MACE estimates were 8.4% with glyburide, 8.6% with glimepiride, and 9.1% with glipizide. Compared with DPP-4 inhibitors, the 5-year relative risk of MACE was 13% higher with glipizide, but the differences were not statistically significant with the other two sulfonylureas. The results were similar in multiple subgroups, including those stratified by age, sex, HbA1c, and kidney function.
In a sensitivity analysis that excluded 367 patients taking saxagliptin, a DPP-4 inhibitor linked previously with heart failure, the 5-year risks of MACE were in line with the main analysis: higher risks with glipizide, but not the other two sulfonylureas.
My practice since 2019 has been to use glimepiride when I use a sulfonylurea. John Buse
Diabetes researcher John Buse, MD, PhD (University of North Carolina at Chapel Hill), who wasn’t involved in the study, praised the group for the rigorous study. However, “the critical issue is that there is undoubtedly residual confounding, as inexpensive first-line drugs like glipizide almost always perform poorly compared to branded and thus much more expensive alternatives,” he told TCTMD.
For example, glipizide has a shorter half-life and lower risk of hypoglycemia than the other drugs and may be used more frequently in higher-risk patients. This heightened baseline risk might not be fully captured by statistical adjustments. While it’s possible that glipizide is associated with a higher risk of MACE compared with DPP-4 inhibitors, Buse said he isn’t certain.
“I do know from the CAROLINA randomized controlled trial published in 2019 that glimepiride is safe from a cardiovascular perspective as compared to linagliptin, a DPP-4 inhibitor,” he said. “So, my practice since 2019 has been to use glimepiride when I use a sulfonylurea.”
What Happens Next?
One strength of the analysis is that researchers emulated a target trial using data from the Patient-Centered Outcomes Research Institute’s PCORnet Common Data Model, which allowed them to include a diverse cohort across geographies, health systems, and payers. Doing so, they say, reduced the risk of immortal time and selection biases. Also, unlike in claims data, all patients in this analysis had baseline measurements of glycemic control, kidney function, and LDL cholesterol.
A helpful next step would be a dose-dependent analysis to understand if high doses are associated with greater cardiovascular risk, said Turchin. A randomized, controlled trial comparing the different sulfonylureas to each other, as well as other agents, would also be helpful.
“That would be a great thing to do,” said Turchin. “The problem is, who is going to pay millions of dollars to do it? I don’t know if that’ll necessarily be done. At the very least, it would be good if other people conducted similar analyses of observational data to see if their findings are similar to ours.”
In the absence of more data, physicians should simply be “cautious when prescribing glipizide, in general, but particularly when [doing so] in patients at high cardiovascular risk,” added Turchin.
In their editorial, Ratzki-Leewing and colleagues say that formulary reform is needed, and that payers and pharmacy benefit managers need to reassess glipizide’s place given the “now evident cardiovascular risk.” Professional societies should also stop treating sulfonylureas as a homogeneous class, they say. “Agent-specific risks merit agent-specific recommendations.”