Two large retrospective studies presented at the 2026 ASCO Annual Meeting report that women taking glucagon-like peptide-1 receptor agonists (GLP-1 RAs) were diagnosed with breast cancer less often than women who did not use the drugs, according to The ASCO Post. The findings cover both prevention in high-risk women with obesity and cancer detection during routine breast imaging. Researchers stressed that the observational designs cannot show cause and effect.
Why it matters
Obesity raises both the likelihood of developing breast cancer and the chance of a worse outcome after diagnosis, and few risk-reducing options exist. With GLP-1 drugs such as semaglutide and tirzepatide now widely prescribed for diabetes and weight loss, oncologists are asking whether that reach extends to cancer prevention. The meeting data point in a promising direction but fall short of answering the question, and the discussion that followed the presentations focused as much on bias as on benefit.
A 16% relative drop in a high-risk population
The first study, led by Nico-al Paolo Gotera, MPH, DO, an assistant professor at the University of Texas at San Antonio, drew on the TriNetX database. The team identified 80,480 adult women with a body mass index of at least 30 and elevated breast cancer risk tied to genetic, familial, or breast tissue factors. Women with prior breast cancer or endocrine therapy use were excluded. After matching users and nonusers one to one on propensity scores, the analysis compared 40,240 women on GLP-1 RAs with the same number who were not.
Over a median follow-up of roughly 2,700 days, breast cancer incidence was 2.95% among users and 3.07% among nonusers, a hazard ratio of 0.844 that was statistically significant, The ASCO Post reports. Colton Jones, MD, a hematology and medical oncology fellow at UT San Antonio, said the pattern held across several subgroups. The strongest signals appeared in women with a BMI above 40 and in those with a family history of breast cancer or benign mammary dysplasia. Significant reductions were seen with semaglutide, tirzepatide, and dulaglutide individually.
The association appeared among postmenopausal women but not premenopausal women. Jones said the team does not yet know why and will examine whether the postmenopausal group had more or more severe obesity. Gastrointestinal side effects were significantly more common in users, while no meaningful link emerged with venous thromboembolism.
Asked about clinical use, Jones was direct: he would not start a patient on a GLP-1 RA solely to lower breast cancer risk, because a prospective trial is needed to confirm the findings.
Fewer cancers found at imaging, and a trial on the way
A second study took the detection angle. Elizabeth McDonald, MD, PhD, a professor of radiology at the University of Pennsylvania Perelman School of Medicine, and colleagues reviewed electronic health records for 111,646 women aged 45 to 80 with a BMI above 25 who underwent breast imaging at a major academic center. Of these, 15,264 had been exposed to a GLP-1 RA and 96,382 had not.
The exposed group was younger and had more obesity, diabetes, and other conditions, so the researchers matched pairs on age, race, ethnicity, BMI, breast density, and diabetes status. In both the unmatched and matched analyses, prior GLP-1 exposure was tied to significantly less breast cancer detection: 1.6% versus 2.6% before matching and 1.6% versus 2.3% after, according to The ASCO Post. McDonald described an absolute risk reduction of 0.69% and acknowledged that observational data cannot establish causality.
She pointed to similar signals across multiple cancers at the meeting and called for investment in a randomized trial to test whether the drugs actually prevent cancer. The ECOG-ACRIN Cancer Research Group plans to launch such a study, INSPIRE, short for INtegrated Strategies for Breast Cancer Prevention and Interception in Elevated-Risk Women. It will test whether metabolic intervention lowers risk in women with a history of breast cancer or at elevated risk of the disease.
The case for caution
Rayjean J. Hung, PhD, FCAHS, associate director for population health at Sinai Health System and a professor at the University of Toronto, served as invited discussant. While praising the research, she listed factors that could create or inflate an apparent protective effect: selection bias in who receives a GLP-1 prescription, time-related biases, unmeasured risk factors, and the causal inference problems McDonald herself raised.
Hung said that before randomized controlled trials, the field needs supporting evidence from mechanistic studies and from real-world data that accounts for those biases. She also called for better comparators, such as active new users of drugs that neither raise nor lower cancer risk, and for study designs and analytical methods that strengthen causal inference, including target trial emulation.
The ASCO Post notes that the studies add to a growing body of work on GLP-1 drugs and cancer, and that experts continue to urge caution in reading the results.
What to watch: whether INSPIRE and similar randomized efforts move forward, and whether mechanistic research explains the postmenopausal-only signal seen in the TriNetX analysis.
