麻豆传媒

SAN DIEGO -- Type 2 diabetes patients may have a lower risk of venous thromboembolism (VTE) when treated with a GLP-1 receptor agonist, regardless of their obesity status, a database study involving more than half a million patients found.

Propensity-score matched analysis showed that VTE rates at 1 year were 20% lower for patients starting on a GLP-1 drug rather than a DPP-4 inhibitor (6.5 vs 7.9 per 1,000 patient-years, HR 0.80, 95% CI 0.75-0.85, P<0.001), reported Rushad Patell, MD, of Harvard Medical School in Boston.

Subgroup analysis showed a lower risk associated with GLP-1 use both for pulmonary embolism (3.1 vs 3.9 per 1,000 patient-years, HR 0.78, 95% CI 0.71-0.85) and deep vein thrombosis (4.2 vs 5.0 per 1,000 patient-years, HR 0.82, 95% CI 0.75-0.88), according to findings presented here at the American Society of Hematology annual meeting.

An estimated 100,000 U.S. adults experience VTE each year, and incidence has been on the rise, said Patell, with one study showing that the serious and sometimes fatal events have increased by about 20% over the last decade. VTE can also lead to , which can affect patients' quality of life and productivity and prove costly for them and society.

As obesity is a known risk factor for VTE -- responsible for by some estimates -- the researchers hypothesized that the effect of GLP-1 receptor agonists on weight loss could have an impact on VTE as well. GLP-1 agonists -- which include blockbusters such as semaglutide (Rybelsus, Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound) -- were initially approved for type 2 diabetes before gaining indications for chronic weight management.

In the study, however, the reduction in VTE risk was observed across most BMI categories and the difference in VTE rates between the GLP-1 agonist and DPP-4 inhibitor groups became evident almost right after treatment initiation, said Patell. "It may be beyond just weight loss that there is an impact," he said, adding that preclinical studies have shown that GLP-1 receptors exist on platelets and endothelial cells.

If verified, the study findings could affect clinical practice.

"When I see somebody with a blood clot and I prescribe anticoagulants, if obesity's a risk factor we tell them about diet and exercise but sort of shrug our shoulders and do not really expect that to change," Patell said. "Maybe we finally have something [in GLP-1 drugs] that can impact this risk factor that is already , and if it actually does reduce the risk of VTE that could lead to some changes."

The retrospective, target therapeutic emulation study used data from the multicenter TriNetX Analytics Network database, which contains deidentified electronic health records on more than 250 million patients globally.

The two cohorts included type 2 diabetes patients newly starting either GLP-1 drugs or DPP-4 inhibitors, which were selected as the control antidiabetic medication as they have not been associated with weight loss.

Patients were excluded if they already had a VTE, were on oral anticoagulation, or had atrial fibrillation -- the most frequent indication for blood thinners. Patients who received both GLP-1 receptor agonists and DPP-4 inhibitors together were excluded, but crossover was allowed.

After propensity-score matching for indications, underlying comorbidities, laboratory values like HbA1c, demographics, and use of other antidiabetic agents, the researchers identified a matched set of 279,064 patients who received a GLP-1 drug and 279,064 who received a DPP-4 inhibitor.

Irrespective of BMI at baseline, GLP-1 drugs remained associated with a lower risk of VTE at 12 months:

• BMI 18.5-24.9: HR 0.61 (95% CI 0.42-0.89)
• BMI 25-29.9: HR 0.82 (95% CI 0.66-1.03)
• BMI 30-34.9: HR 0.77 (95% CI 0.63-0.94)
• BMI 35-39.9: HR 0.60 (95% CI 0.47-0.77)
• BMI ≥40: HR 0.74 (95% CI 0.60-0.91)

A limitation of the study was that the dataset did not capture changes in patients' weight, so the researchers could not evaluate whether weight loss correlated with VTE incidence. "It's something that we are hoping to do with different datasets in future studies," said Patell.

Comment