Glucagon-like peptide-1 receptor agonists (GLP-1RAs) do not appear to increase overall cancer risk based on currently available randomized trial data with a median follow-up of 70 weeks, according to a systematic review and meta-analysis published in Annals of Internal Medicine.
“However, because cancer is a long-latency outcome and the number of events is relatively small in our study, our findings should be viewed as reassuring but not definitive,” one of the study authors, Cho-Han Chiang, MD, of the Department of Medicine and Mount Auburn Hospital, Harvard Medical School, Cambridge, MA, told this news organization in an interview. “Clinicians should continue to individualize treatment decisions and remain attentive to new safety data as longer-term evidence emerges.”
GLP-1 receptor agonists are being used by millions of people worldwide for diabetes and obesity, yet their long-term cancer safety has remained uncertain. According to Chiang, early observational studies and case reports raised concerns about potential links with certain cancers such as thyroid and pancreatic cancers. “At the same time, other observational studies have suggested a lower risk of certain cancers, particularly obesity-related cancers, with GLP-1RA use,” he said. “This mixture of signals has contributed to uncertainty in the field.”
In a systemic review and meta-analysis, Chiang and coauthors used PubMed and other online sources to pull together cancer events across 48 randomized, placebo-controlled clinical trials involving 94,245 participants to provide an assessment of cancer risk with GLP-1RAs. The researchers used the Cochrane Risk of Bias 2 tool to evaluate the risk of bias, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to assess certainty of evidence, and random-effects meta-analysis to pool odds ratios (ORs).
Over a median follow-up of 70 weeks, the authors reported that trials with moderate-certainty evidence suggest GLP-1RAs “probably have little or no effect” on the risk of thyroid cancer (OR, 1.37), pancreatic cancer (OR, 0.84), breast cancer (OR, 0.95), or kidney cancer (OR, 1.12).
Meanwhile, trials with low or very low certainty evidence suggested that GLP-1RAs “may have little or no effect” on colorectal, esophageal, liver, gallbladder, ovarian, or endometrial cancer, as well as multiple myeloma or meningioma. The authors noted that this was likely due to sparse events, wide confidence intervals, and considerable variation across trials.
“One thing that stood out to us was how reassuringly consistent the results were across different trials and patient populations: we did not see a clear signal that GLP-1RA therapy increased overall cancer risk,” Chiang said. “Given the longstanding concerns about specific cancer types (such as thyroid and pancreatic cancer), the absence of an increased signal in those prespecified sites was reassuring.”
At the same time, he continued, some recent observational studies have suggested a reduced risk of certain obesity-related cancers among GLP-1RA users. “This is in contrast to our study, where we showed that GLP-1RA neither increased nor decreased the risk of obesity-related cancers,” Chiang said. “This discrepancy likely reflects differences in study design, confounding structures, patient populations, and the degree of weight loss achieved in real-world settings versus clinical trials.”
He emphasized that studies with 5–10 years or more of follow-up are required to fully determine whether GLP-1RAs influence the incidence or progression of cancer, particularly site-specific and obesity-related cancers. “Trials or large observational studies with prespecified, standardized cancer endpoints and adjudication will be key,” Chiang said. “In addition, evidence regarding the safety of GLP-1RAs in individuals with pre-existing cancer remains limited. It will be important to investigate whether GLP-1RAs affect the recurrence or progression among participants with a history of cancer.”
The researchers reported having no disclosures.
Source: Annals of Internal Medicine