GLP-1s and Cancer: The 2026 Evidence Map Across Colorectal, Pancreatic, Breast, and Thyroid

In a KFF poll taken in late 2025, about one in eight U.S. adults (12%) said they were currently taking a GLP-1 drug such as Ozempic or Wegovy. That works out to roughly 30 million people. Until about 2024, most talk about these drugs centered on weight loss, blood sugar, and heart protection, such as the results of the SELECT trial. Cancer came up mainly as a warning: the boxed warning about medullary thyroid cancer on the Wegovy and Ozempic labels, and an older worry about pancreatic cancer that had mostly faded.

Since 2025, the cancer discussion has split into three separate questions. Do GLP-1 drugs lower the chance of getting cancer? Do they help people who already have cancer, for example by slowing its spread? And do they cause any cancers that need closer watching? Each question has its own evidence, and the one that matters most depends on your situation.

Most of the new evidence comes from observational studies. These look back at the health records of people who happened to take one drug or another. A randomized controlled trial (RCT) instead assigns treatments by chance, which is the only reliable way to show that a drug caused a result. With that limit in mind, the signal for lower cancer risk has grown stronger, and lab studies suggest GLP-1 drugs may act on some tumor cells directly as well as through weight loss.

One of the largest prevention studies so far was presented at the 2025 annual meeting of the American Society of Clinical Oncology (ASCO), held May 30 to June 3, 2025. Researchers led by Lucas Mavromatis of the NYU Grossman School of Medicine used health records from 43 U.S. health systems. They compared 170,030 adults with obesity and diabetes who started either a GLP-1 drug or a DPP-4 inhibitor. DPP-4 inhibitors are older diabetes pills that act on the same hormone system but cause little weight loss.

The study used target trial emulation, a method that sets up a comparison in health records to mimic a trial. It also used propensity score matching, which pairs patients with similar traits so the two groups look alike at the start. People on GLP-1 drugs had a 7% lower risk of developing one of 14 obesity-related cancers and an 8% lower risk of dying from any cause. In raw numbers, 2,501 of 85,015 GLP-1 users and 2,671 of 85,015 DPP-4 users were diagnosed with one of those cancers.

The 14 cancers were those of the esophagus, colon, rectum, stomach, liver, gallbladder, pancreas, kidney, breast after menopause, ovary, endometrium (the lining of the uterus), and thyroid, plus multiple myeloma and meningioma. The clearest results were for colon cancer, with 16% fewer cases, and rectal cancer, with 28% fewer. Women saw a benefit, while men showed no statistically significant difference. The GLP-1 group did not have a higher risk of any of the 14 cancers.

A 2025 study in eClinicalMedicine compared GLP-1 drugs with weight-loss surgery. Using records from Clalit Health Services in Israel, researchers matched 3,178 pairs of adults with obesity and diabetes who received either an early-generation GLP-1 drug or bariatric surgery between 2010 and 2018. Over a median of 7.5 years, obesity-related cancers occurred at similar rates: 5.89 cases per 1,000 person-years with the drugs and 5.62 with surgery.

The hazard ratio, a measure that compares how often an event happens in two groups, was 1.11, with a 95% confidence interval of 0.86 to 1.44. Because that range includes 1, the difference was not statistically significant. In a further analysis, the authors estimated that the drugs had a direct effect, separate from weight loss, equal to a 41% lower relative risk.

Taken together, these studies point to a modest drop in obesity-related cancers, in the same direction across different groups of patients. How much of it comes from weight loss and how much from the drugs themselves is still an open question.

The biggest treatment signal comes from a study that ASCO released on May 21, 2026, and that was presented at the 2026 ASCO Annual Meeting in Chicago, held May 29 to June 2. Mark David Orland of the Cleveland Clinic's Taussig Cancer Institute led the work. The team used the TriNetX health records database to study 12,112 people with stage I, II, or III cancer who started either a GLP-1 drug or a DPP-4 inhibitor after their diagnosis.

Each person had one of seven cancers: breast, prostate, non-small cell lung, colorectal, liver, kidney, or pancreatic. The researchers counted how many people progressed to stage IV, meaning the cancer had spread to distant parts of the body. For four of the seven cancers, people on GLP-1 drugs were 38% to 50% less likely to reach stage IV:

• Lung: 10% on GLP-1 drugs vs. 22% on DPP-4 inhibitors.
• Breast: 10% vs. 20%.
• Colorectal: 13% vs. 22%.
• Liver: 19% vs. 28%.

For prostate, pancreatic, and kidney cancer, the GLP-1 group also had fewer cases of spread, but the differences were not statistically significant. Side effects were similar in the two groups.

The team also ran a separate analysis using The Cancer Genome Atlas, a public research database of tumor samples. People whose tumors had high levels of the GLP-1 receptor (GLP-1R), the protein that GLP-1 drugs act on, had a 33% lower risk of death than people with low levels. In breast cancer, the difference was 45%. This analysis did not look at GLP-1 drug use. It suggests that the GLP-1 signaling system may affect how these cancers behave, which is one reason the authors think the drugs could have a direct effect.

A University of California analysis, published in Cancer Investigation in November 2025, studied more than 6,800 people with colon cancer across UC Health sites. Over five years, 15.5% of GLP-1 users died, compared with 37.1% of people not taking the drugs. After adjusting for age, body mass index, disease severity, and other factors, GLP-1 users still had significantly lower odds of death. The researchers called the results observational and said clinical trials are needed.

If GLP-1 drugs lower cancer risk, they could do it in two main ways, and the difference matters for who benefits.

The first is weight loss. The CDC lists 13 cancers linked to overweight and obesity. Extra body fat raises estrogen levels, raises insulin and a related growth signal called IGF-1, and keeps the body in a state of low-grade inflammation, all of which can help tumors grow. If GLP-1 drugs work only through weight loss, then other ways of losing weight, such as surgery or lifestyle programs, should give similar protection. The benefit should also grow with the amount of weight lost.

The second is a direct effect on tumor cells. Many cancer cells carry GLP-1 receptors. In lab studies, switching on those receptors can slow the growth and movement of some cancer cells. A 2025 study found that semaglutide slowed the growth, migration, and invasion of oral squamous cell carcinoma cells, a type of mouth cancer, and shrank these tumors in mice. Other lab work has described effects on breast cancer cells that do not depend on weight loss. Results in cells and mice do not show what happens in patients, but they make a direct effect plausible.

Current evidence fits a mix of both. The Israeli study's estimate of a direct effect beyond weight loss, and the link between tumor GLP-1R levels and survival in the 2026 ASCO analysis, both point toward a direct role. Weight loss is the better understood route for obesity-related cancers in general. The balance between the two probably differs by cancer type, and no human study has yet separated them cleanly.

Wegovy, Ozempic, Mounjaro, and Zepbound all carry an FDA boxed warning, the agency's strongest, about thyroid C-cell tumors. These include medullary thyroid carcinoma (MTC), a rare cancer of the thyroid's C cells, which make the hormone calcitonin. The warning comes from animal data. Rats and mice given semaglutide or tirzepatide at high doses for long periods developed C-cell tumors, likely because the drugs kept stimulating GLP-1 receptors on those cells.

Human data so far are reassuring but not final. Large population studies, reviewed by CancerNetwork in 2025, have not found a clear increase in medullary thyroid cancer or other thyroid cancers among people taking GLP-1 drugs. In the 2025 ASCO study, the GLP-1 group did not have a higher risk of thyroid cancer or of any of the other obesity-related cancers studied.

The labels say that people with a personal or family history of MTC, or with multiple endocrine neoplasia syndrome type 2 (MEN2), an inherited condition that raises the risk of MTC, should not take these drugs. That is the only thyroid-related contraindication, meaning a situation in which the drug should not be used. The warning stays on the label because rare cancers can take many years to show up in population data.

If you have a personal or family history of MTC or MEN2, these drugs are off the table. For everyone else, studies so far have not detected an increase in thyroid cancer, though they cannot rule out a very small risk.

Worries about pancreatic cancer date to 2013, when early reports on exenatide, a GLP-1 drug, and sitagliptin, a DPP-4 inhibitor, raised the possibility of a higher risk. One theory was that GLP-1 receptors on certain pancreas cells might promote early changes that can lead to cancer. The concern drew an FDA review and years of monitoring after the drugs reached the market.

Newer studies of semaglutide and tirzepatide have mostly not supported that concern. The 2025 ASCO study found no increase in pancreatic cancer among GLP-1 users. In the 2026 ASCO study of people who already had cancer, pancreatic cancer spread somewhat less often on GLP-1 drugs than on DPP-4 inhibitors, but the difference was not statistically significant. DPP-4 inhibitors also raise the body's own GLP-1 levels, so comparisons against them may understate any GLP-1 effect.

A study designed to answer this question was presented at the EASD diabetes meeting in Milan in late September 2026. Anton Pottegård of the University of Southern Denmark and colleagues used national health registries in Denmark, Sweden, and Norway. They followed 97,464 adults with type 2 diabetes who started semaglutide and a matched group who started insulin, a sulfonylurea, or an SGLT2 inhibitor. Counting began one year after treatment started, a common way to leave out cancers that were already present. From then on, there were 131 pancreatic cancers in the semaglutide group and 123 in the comparison group. The pooled hazard ratio was 0.91 (95% confidence interval 0.71 to 1.16), which means the study found no sign of higher risk.

The European Medicines Agency required this study as part of semaglutide's risk management plan, and Novo Nordisk paid for it. Its main weakness is time. After that first year, median follow-up ranged from 1.4 to 1.9 years, and pancreatic cancer can take many years to develop. The study also covered only semaglutide in people with diabetes, and it measured cancer, not pancreatitis.

The pancreatitis question is separate. Acute pancreatitis is a listed side effect of GLP-1 drugs, but it is uncommon. In Wegovy's adult weight-loss trials, a review committee confirmed acute pancreatitis in 4 patients on the drug (0.2 cases per 100 patient-years) and 1 on placebo (under 0.1). Chronic pancreatitis is a known risk factor for pancreatic cancer, but whether the acute episodes seen with GLP-1 drugs raise long-term cancer risk is not known. U.S. labels for Wegovy, Ozempic, Mounjaro, and Zepbound do not list a history of pancreatitis as a contraindication. Instead, they warn that acute pancreatitis, including fatal cases, has occurred with GLP-1 drugs, and they tell clinicians to stop the drug if pancreatitis is suspected.

The European Medicines Agency lists pancreatic cancer as a potential risk for the GLP-1 drug class, which calls for continued monitoring rather than a contraindication. The Nordic registry study is one product of that monitoring. In short, the 2013 pancreatic cancer signal has not held up for newer GLP-1 drugs, but acute pancreatitis remains a real, if uncommon, side effect.

Nearly all the cancer findings above come from observational data: health records, matched groups, and trial-emulation studies. None is the kind of randomized trial evidence the FDA usually requires before it adds a new use to a drug's label.

Observational studies have known weak spots:

• Confounding by indication: people prescribed GLP-1 drugs may differ from people prescribed DPP-4 inhibitors in ways that affect cancer risk on their own.
• Selection bias: people who tolerate the stomach side effects and stay on a GLP-1 drug may differ from people who stop.
• Immortal time bias: counting people as users only if they stayed on a drug for a while builds in a stretch of time in which nothing bad could be counted against the drug, which can make it look better than it is.
• Surveillance bias: GLP-1 users may see doctors more often and get more cancer screening. That can find cancers earlier and lower the share found at stage IV.

The studies tried to reduce these problems. The 2026 ASCO study compared two groups of people taking diabetes drugs rather than comparing users with nonusers. The 2025 ASCO study used target trial emulation and propensity score matching to make its groups as similar as possible. These methods help, but they cannot rule out every hidden difference.

Settling the question will take randomized trials. Large outcome trials such as SELECT, FLOW, SOUL, and TRIUMPH-Outcomes record cancers as part of their safety data, but they were designed to measure heart or kidney outcomes, so any cancer findings from them are exploratory. Small trials are starting to test GLP-1 drugs in people with cancer. One is a Rutgers Phase 2 trial, recruiting about 40 people with early-stage hormone receptor-positive breast cancer, that measures weight loss on tirzepatide during treatment. Trials like this are not built to show whether the drugs prevent cancer or slow its spread. The authors of the 2026 ASCO study have said they want to run randomized trials in people with cancer.

Probably not, if you do not otherwise need them for obesity or diabetes. A simple example shows why.

For someone at average risk, a 7% to 10% lower relative risk is a small change in absolute risk. If your chance of getting an obesity-related cancer over 10 years is 5%, a 10% relative reduction lowers it to 4.5%, a drop of half a percentage point. Whether that is worth the cost, the side effects, and the long-term commitment depends on your own risk.

The numbers change for people at higher risk, such as those with obesity, type 2 diabetes, a strong family history, or inherited conditions such as Lynch syndrome or BRCA1/2 mutations. If your 10-year risk is 20% and the reduction is 10% to 15%, the absolute drop is 2 to 3 percentage points, which is large enough to weigh. These figures are illustrations; the studies above did not measure the benefit in these specific groups.

If you already take a GLP-1 drug for obesity or diabetes, any cancer benefit comes on top of the reasons you are taking it. It does not change the decision to prescribe it.

If you have specific risk factors, such as a strong family history of colorectal, breast, or liver cancer, Lynch syndrome, or BRCA1/2 mutations, ask your oncologist or genetic counselor whether this evidence should play any part in your prevention plan. Current guidelines do not recommend GLP-1 drugs for cancer prevention.

There are three common situations.

If you have one of the cancers in the 2026 ASCO study and are not on a GLP-1 drug, the new data are worth discussing with your oncology team. Breast, colorectal, lung, and liver cancer showed the clearest results. A 38% to 50% lower rate of spread to stage IV is a large difference, even though it comes from an observational study. These drugs are not approved to treat cancer. The tumor GLP-1R analysis also did not test who benefits from the drugs, so measuring GLP-1R in your tumor is a research question rather than a standard step.

If you have a cancer outside this group, the cancer findings do not apply. The heart and metabolic benefits still apply if you have those conditions.

If you have cancer and already take a GLP-1 drug for obesity or diabetes, the studies above have not shown cancer-related harm from staying on it, and some suggest a benefit. Labels call for stopping the drug if pancreatitis is suspected, and people with a history of MTC or MEN2 should not be taking it at all.

One caution applies to everyone. GLP-1 drugs cause substantial weight loss, and losing weight during cancer treatment is not always good. People at risk of sarcopenia, the loss of muscle, or cachexia, severe weight and muscle loss caused by illness, need extra care. Decisions about starting or continuing a GLP-1 drug during treatment should involve both the oncology team and an obesity specialist, with attention to muscle and body composition as well as the number on the scale.

Observational studies suggest that GLP-1 drugs lower the risk of obesity-related cancers by a modest amount and may slow the spread of some cancers that are already present. The findings have appeared in several large datasets, and they fit both a weight-loss explanation and a direct effect on tumors. The 2026 ASCO result, 38% to 50% less spread to stage IV in four cancers, stands out, but it has not been tested in a randomized trial.

The thyroid and pancreatic cancer warnings have not been borne out in large human studies of newer GLP-1 drugs. The MTC/MEN2 contraindication remains for a small group of people, and acute pancreatitis remains an uncommon side effect that labels tell clinicians to watch for. For most people taking these drugs for obesity or diabetes, the cancer risks found so far look smaller than the possible benefits.

Medical practice has not caught up. Guidelines do not recommend GLP-1 drugs for cancer prevention, testing tumors for GLP-1R is not standard, and insurers do not cover the drugs for cancer prevention, since no drug is approved for that use.

If you already take a GLP-1 drug, the cancer evidence adds to its value without changing the decision. If you are thinking about taking one only to prevent cancer, that is probably not the right reason unless you have a high cancer risk and a doctor who knows this research. If you have an obesity-related cancer and are not on a GLP-1 drug, the question is now worth raising with your oncology team.

Update (October 4, 2026): We added the Nordic registry study of semaglutide and pancreatic cancer presented at EASD 2026.

Correction (October 4, 2026): The original version said GLP-1 drugs are contraindicated in people with a history of pancreatitis. U.S. labels do not list that as a contraindication; they warn about acute pancreatitis and tell clinicians to stop the drug if it is suspected. We also replaced an unsourced acute pancreatitis rate of 0.5% to 1% with the rate from the Wegovy label, and we removed a claim that GLP-1-related pancreatitis raises later cancer risk, which our sources do not support.

Update (October 5, 2026): We rewrote the article in plainer language and replaced references to "today" with dates. ASCO released the Abstract 3143 results on May 21, 2026, before its May 29 to June 2 meeting. We removed a mention of a "TIDES-CV" trial. We could not find it in any trial registry or published source, so we cite a registered trial instead. The figure of roughly 30 million users now rests on KFF's November 2025 poll, in which 12% of U.S. adults said they take a GLP-1 drug.

We also removed or softened claims our sources do not support. These include that the colon cancer result is the most often repeated finding, that the cancer benefit builds over time, and that the colorectal effect grows with weight loss. We dropped a 5% to 15% range for lower cancer risk and a claim that the 2026 ASCO study used propensity score matching. We no longer say that testing tumors for GLP-1R is reasonable to request, that cancer centers have begun discussing the drugs for prevention, or that the balance of benefit and risk "overwhelmingly" favors the drugs.

Correction (October 5, 2026): The original version listed bladder, prostate, and lung cancer among the 14 cancers in the 2025 ASCO study and left out gallbladder cancer. The correct list appears above. It gave June 10 as that study's date, but the 2025 ASCO meeting ran May 30 to June 3. It said the eClinicalMedicine study showed lower cancer rates than a historical control group. That study compared GLP-1 drugs with weight-loss surgery and found similar cancer rates. It also implied that the tumor GLP-1R survival finding came from people taking GLP-1 drugs. That analysis used The Cancer Genome Atlas and did not involve drug use.