Intended for healthcare professionals


Glucagon-like peptide-1 receptor agonists and risk of breast cancer

BMJ 2016; 355 doi: (Published 20 October 2016) Cite this as: BMJ 2016;355:i5519

Chinese translation


  1. Shari D Bolen, associate professor of medicine, epidemiology, and biostatistics13,
  2. Nisa M Maruthur, assistant professor of medicine and epidemiology46
  1. 1Division of General Internal Medicine, Department of Medicine, MetroHealth System/Case Western Reserve University, Cleveland, OH, USA
  2. 2Center for Healthcare Research and Policy, Case Western Reserve University at MetroHealth Medical Center, 2500 MetroHealth Drive, Rammelkamp Building, Room 234A, Cleveland, OH 44109, USA
  3. 3Department of Epidemiology and Biostatistics, Case Western Reserve University, Cleveland, OH, USA
  4. 4Division of General Internal Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA
  5. 5Department of Epidemiology, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD, USA
  6. 6Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, MD, USA
  1. sdb73{at}

Concern remains, and should feature in discussions with patients

Although many providers and some patients know about the potential risks of pancreatic or thyroid cancer with use of glucagon-like peptide-1 (GLP-1) receptor agonists,1 2 3 risk of breast cancer has also recently arisen as a potential risk with these drugs. The US Federal Drug Administration (FDA) and European Medicines Agency separately conducted pooled analyses of four weight management trials that investigated the use of 3.0 mg liraglutide (a type of GLP-1 receptor agonist).4 5 In these pooled analyses, 12 (0.29%) breast cancer events were reported in the liraglutide arms versus two (0.08%) events in the placebo arms.

Given the rare occurrence of these events, it was unclear whether this difference was due to chance alone or a true increase in breast cancer risk. The LEADER trial (liraglutide effect and action in diabetes: evaluation of cardiovascular outcome results) compared liraglutide at a lower dose of 1.8 mg with placebo, and showed no difference in breast cancer risk.6 However, this study was not designed to analyse breast cancer risk and had too few events for firm conclusions. Although GLP-1 receptors are not found in breast tissue, GLP-1 receptor agonists have been associated with increased expression of fibroblast growth factor, an important growth factor within the breast.7 8 Given the potential concerns regarding breast cancer risk, the FDA suggested additional studies to evaluate this risk.

In a linked paper in The BMJ, Hicks and colleagues (doi:10.1136/bmj.i5340)9 describe an observational study in the UK comparing GLP-1 receptor agonists and dipeptidyl peptidase-4 (DPP-4) inhibitors in adults with type 2 diabetes to further assess breast cancer risk. Although the study included 44 984 patients, the primary analysis focused on only 498 (1.1%) patients taking a GLP-1 receptor agonist and 2422 (5.4%) patients taking a DPP-4 inhibitor. The authors report that use of GLP-1 receptor agonists was not associated with an increased risk of breast cancer overall, compared with DPP-4 inhibitors (4.4 v 3.4 per 1000 person years; hazard ratio 1.40 (95% confidence interval 0.91 to 2.16)). However, when stratified by duration of use, a significant increase in breast cancer risk was identified in patients using GLP-1 receptor agonists for two to three years (2.66 (1.32 to 5.38)). The authors state that the risk diminished at more than three years of use, but the low number of events and people taking GLP-1 receptor agonists at this stage make it difficult to draw any true conclusions for longer follow-up periods.

Hicks and colleagues discuss that the lag in breast cancer risk might be due to detection bias, stating that the weight loss associated with GLP-1 receptor agonists could lead to enhanced detection of breast lumps. However, they acknowledge that no evidence has been identified to support that hypothesis and that this could represent a true increase in breast cancer risk. Furthermore, observational studies of drugs in particular run the risk of confounding by indication (that is, by selecting low risk patients for the drug treatment). Although Hicks and colleagues adjust for risk factors such as higher body mass index and use of hormonal treatment, other statistical techniques such as propensity scores might better address potential confounding by indication. Whether confounding exists or not, the increased events in the overall and time lag analyses represent a true concern of increased breast cancer risk.

How should we counsel patients being treated with drugs in this class? For adults with type 2 diabetes, GLP-1 receptor agonists are associated with strong benefits on weight and blood sugar control.10 Early evidence also suggests that these agents—liraglutide in particular—could be associated with fewer cardiovascular events or deaths than other standard diabetes drugs.6 Although GLP-1 receptor agonists are not currently a first line treatment for adults with type 2 diabetes, they are used in combination with other diabetes treatments for blood sugar control. Before prescribing, providers should discuss the balance of risks and benefits with all eligible patients, including a possible small increase in risk of breast cancer. Further studies are warranted to determine with greater confidence whether this risk is real, at what doses the risk appears, and whether it is specific to individual drugs within the larger class of GLP-1 receptor agonists. As these drugs come off patent and are used more widely in the future, outcomes on larger numbers of patients will be available to compare risks of rare events, such as breast cancer, across drug types and doses and over longer exposure periods.


  • Research, doi: 10.1136/bmj.i5340
  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and declare the following interests: None.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription