May 11, 2026

expert reaction to study on association between GLP-1 use and breast cancer survival and recurrence

A study published in Nature Medicine looks at an association between GLP-1 use and breast cancer survival and recurrence. 

Dr Mangesh Thorat, Honorary Reader in Wolfson Institute of Population Health, Queen Mary University of London and Consultant Breast Surgeon, Homerton University Hospital, Queen Mary University of London (QMUL), said:

“This study by Tatum and colleagues has several major weaknesses making it impossible to draw breast cancer related inferences with even a low level of certainty.

“Overall, there appear to be fewer deaths in patients using GLP-1 receptor agonists (GLP-1 RAs) but the chances of any such benefit being driven through reduction in breast cancer recurrences or deaths due to breast cancer are rather slim.

“The data presented in the paper appears to be highly incomplete and there is substantial attrition and selection bias.

“Outside the context of this study, GLP-1 RAs appear to have substantial metabolic pathway modifying ability and therefore their role in cancer treatment settings merits research through prospective clinical trials.”

Additional explanatory notes:

“In terms of selection bias, the proportion of patients with Ductal Carcinoma In Situ (DCIS) is reported in the paper to be 8-10%, while the normal expected proportion is 15-20%. The final cohort population also represents 2% of overall starting cohort population.

“In terms of incomplete data, the ER status is known for around 20% of population, whereas this figure should be near 100% in well-curated datasets. Similarly, other key clinical variables where information should be available for near 100% of patients, type of surgery type information is available in <10%, radiotherapy information in <5%, systemic therapy in around 25% of patients.

“In terms of attrition / skewing – For relapse-free survival (RFS), the number of patients contributing are smaller, but this loss of patients does not appear to be random. For example (see figures 2 & 3) – cohort 2 (T2DM insulin/metformin) the difference between numbers (events in parenthesis) for All-cause vs RFS for GLP-1 RA  and no GLP-1 RA is 208 (11 i.e. 5% opposed to expected 1-2%), 334 (-100 i.e. -30%, which doesn’t make sense since generally the number of RFS events should be numerically higher even if one loses 10-15% of the cohort) respectively. For cohort 3 (T2DM SLLT2-i) the numbers are 406 (-172 i.e. -42%) and 463 (-167 i.e. -36%).

“All-cause mortality is a hard endpoint but relapse free survival (RFS) is a relatively softer endpoint and the latter requires accurate information capture – and this is simply not the case here. To draw any breast cancer related conclusion, one ideally needs to show greater improvement in RFS – not at all possible here.”

Prof Paul Pharoah, Professor of Cancer Epidemiology, Cedars-Sinai Medical Center, said:

“This manuscript reports an observational study to investigates outcomes in women with early breast cancer that are taking glucagon-like peptide-1 receptor agonists (GLP-1 RAs) compared to women with breast cancer who were not taking these drugs.  These drugs include semaglutide (trade names Ozempic or Wegovy) and tirzepatide (Mounjaro/Zepbound).  It is not a randomized controlled trial and so assigning causality to any associations is fundamentally problematic.  Propensity score matching is used to try and ensure patients in the groups that are compared with each other are as similar as possible.

“The primary end point was all cause mortality.  While this is a ‘hard’ end point that is relatively easy to ascertain (and therefore not subject to ascertainment bias) it is problematic because treatments for obesity and type 2 diabetes might easily affect other causes of death (such as cardiovascular disease).

“Relapse free survival is used as a secondary end point, which is specific to breast cancer, but is prone to ascertainment bias.  Patients who are on novel drug therapies are likely to be followed up more often and so a cancer relapse might be detected sooner.

“Three groups of breast cancer patients were used.  One group was patients with obesity who were either taking GLP-1 RAs or were not taking any drug for obesity.  In this group GLP-1 RAs were associated with a lower risk of dying and of relapse. A second group were patients with type 2 diabetes who were either taking GLP-1 RAs compared to those taking insulin/metformin.  Again , GLP-1 RAs were associated with a lower risk of dying and of relapse.  The third group were patients with type 2 diabetes who were either taking GLP-1 RAs or SGLT2 inhibitors for diabetes.  In this group there was no difference between outcomes according to treatment.

“The authors state that the findings suggested a potential association between GLP-1 RA use and improved outcomes among patients with BC who have obesity and related metabolic conditions.  This, in itself, is very odd choice of language that avoids the key question of whether or not there is a causal association.  There is clearly an association, but for several reasons it seems unlikely to be causal.

“There are some immediate red flags to suggest that the results are not a reliable indication of causality.  The biggest red flag is that the effect sizes are enormous.  In the second group comparison there is a reduction in all-cause mortality of over 90%! Similarly, the reductions in relapse were more than 56% in group 1 and 67% in group 2.  To put this in perspective the most effective chemotherapy regimens reduce relapse by 38%.   These effect sizes simply cannot be due to the drug and so the only reasonable explanation is confounding/ bias. 

“Another red flag is that no association was found in group three.  If there were a causal association of GLP-1 RAs with breast cancer outcomes then it ought to have been detected in this group also.

“The patients are very similar in terms of their characteristics for the factors used for propensity score matching.  One might therefore expect the mortality in the patients treated with GLP-1 RAs to be similar in all three groups.  However, they are very different suggesting that other un-matched factor are important predictions of mortality.

“In summary, the headline results reported in this paper seems too good to be true.”

Dr Simon Vincent, chief scientific officer at Breast Cancer Now, said:

“We know that people living with obesity are less likely to survive breast cancer and there is some evidence that Type 2 diabetes can impact breast cancer risk and survival too. So we need to understand how best to treat these conditions when someone has them all at the same time. 

“While this US-based retrospective study doesn’t answer the question of whether weight-loss medications can directly reduce breast cancer recurrence and deaths, it highlights that this area is worth researching further.

“In this study, the researchers were looking at deaths from all causes, not just from breast cancer. And although we know what weight-loss and Type 2 diabetes drugs the women were prescribed, we don’t know whether they took the medication, how long they took it for, and if they made any other lifestyle changes. We also don’t know how women’s weight changed over time, so we can’t draw any firm conclusions about the cause and effect.

“With the growing use and popularity of weight-loss drugs, we urgently need more research to understand if, and how, these medications could benefit people with breast cancer – whether they have a long-lasting effect and if they can reduce the risk of breast cancer coming back and spreading as well as reducing breast cancer deaths.

“Anyone looking for support or information about breast cancer can speak to our expert nurses by calling our free, confidential helpline on 0808 800 6000.” 

Prof Naveed Sattar, Professor of Cardiometabolic Medicine/Honorary Consultant, University of Glasgow, said:

“I think these findings are far too vulnerable to confounding from unmeasured factors to support any strong conclusions. For example, individuals motivated to take GLP 1 receptor agonists may also be those more engaged in improving their health overall, meaning other lifestyle behaviours are likely more favourable and could explain the associations observed.

“The reported effect sizes on mortality (which appears substantial here) are also difficult to reconcile with what we already know from placebo controlled trials of these agents in diabetes and obesity (where mortality benefits are much more modest in size), which in turn raises uncertainty around the estimates for recurrence free survival.  Taken together, I don’t think these results alone make the case for undertaking a trial in women with breast cancer at this stage; other evidence would need to be gathered to support such a trial.”

Dr Charles Birts, Lecturer in Antibody Therapy, University of Southampton, said:

“This is an interesting and potentially important study suggesting that GLP-1 receptor agonist use is associated with improved survival and lower recurrence in some women with breast cancer who also have obesity or type 2 diabetes.

“However, as an observational study, it cannot show that the drugs themselves are responsible or fully separate the effects of GLP-1 drugs from other related changes. The apparent benefit could reflect factors such as weight loss, improved metabolic health, or differences in the patients prescribed these treatments, and the results were less consistent when compared specifically with SGLT2 inhibitors, where differences were smaller or not clearly evident.

“Overall, these findings are promising and support further clinical trials, but they should not yet change breast cancer treatment practice.”

‘Survival and Recurrence With GLP-1 Receptor Agonists in Breast Cancer’ by Kristina L. Tatum et al. was published in JAMA Network Open at 16:00 UK time Monday 11 May 2026. 

DOI: 10.1001/jamanetworkopen.2026.12133

Declared interests

Prof Naveed Sattar: “NS has consulted for and/or received speaker honoraria from Abbott Laboratories, AbbVie, Afimmune, Amgen, AstraZeneca, Boehringer Ingelheim, Carmot Therapeutics, Eli Lilly, Gan & Lee, GlaxoSmithKline, Hanmi Pharmaceuticals, Janssen, Kailera, Mass Medicines, Menarini-Ricerche, Merck Sharp & Dohme, Metsera, Novartis, Novo Nordisk, Pfizer, Regeneron, Roche, Sanofi, UCB Pharma and Verdiva Bio; and received grant support paid to his University from AstraZeneca, Boehringer Ingelheim, Novartis, and Roche.  No shares in any medical areas.”

Dr Charlie Birts: “No COI’s to declare.”

Prof Paul Pharoah: “I have no conflicts of interest to declare.”

Dr Mangesh Thorat: “none.”

Dr Simon Vincent: No COIs.