Real-World Data Favor GLP-1 vs Aspirin in Colorectal Cancer Prevention

Data from a real-world, head-to-head comparison of a GLP-1 receptor agonist (GLP-1RA) use vs Aspirin (acetylsalicylic acid) use for the prevention of colorectal cancer (CRC) favor the use of GLP1-RA for risk reduction across subgroups and regardless of body mass index (BMI) or diabetes, according to results presented at the 2026 American Society of Clinical Oncology Gastrointestinal Cancers Symposium.

At a median 6-year follow-up for GLP-1RA users and 5 years for aspirin users, the use of a GLP-1RA reduced the risk of CRC by 35.7% compared with aspirin use (HR, 0.643; 95% CI, 0.531-0.778). In high-risk patients, the risk was reduced by 42.1% (HR, 0.579; 95% CI, 0.401-0.837). High risk was defined as having a first-degree relative with CRC, those with known genetic predisposition, and those with known high-risk comorbidities including ulcerative colitis and Crohn’s disease.

In a presentation of the data, Colton Frisco Jones, MD, a hematology oncology fellow at The University of Texas, San Antonio Mays Cancer Center, described the study as “the first real-world head-to-head comparison of the efficacy and safety of GLP1-RA vs Aspirin for the primary prevention of CRC.”

“The large population size of this study and the favorable safety profile of GLP-1s vs Aspirin could underscore a potential public health impact,” said Jones. “These findings warrant prospective validation.”

Younger patients (ages 18 to 44) saw a greater reduction in the risk of CRC at 41.7% (HR, 0.583; 95% CI, 0.349-0.974) compared with a 21.0% risk reduction (HR, 0.790; 95% CI, 0.623-1.003) in those ages 45 to 64 and a 27.1% risk reduction (HR, 0.729; 95% CI, 0.500-1.064) in those age 65 or older.

Benefit was seen regardless of diabetes or BMI statuses. Patients with diabetes had a 25.4% CRC risk reduction (HR, 0.746; 95% CI, 0.511-0.982), and those without diabetes had a 41.2% risk reduction (HR, 0.588; 95% CI, 0.471-0.734) when taking a GLP-1RA.

Between GLP-1RAs, the following incidence reductions relative to Aspirin use were observed:

• Semaglutide, 36.9% (HR, 0.631; 95% CI, 0.492-0.810)
• Tirzepatide, 28.9% (HR, 0.711; 95% CI, 0.440-1.147)
• Liraglutide, 56.4% (HR, 0.436; 95% CI, 0.220-0.861)
• Dulaglutide, 52.9% (HR, 0.471; 95% CI, 0.358-0.618)

Those with a BMI over 30 had a 38.9% CRC risk reduction (HR, 0.611; 95% CI, 0.492-0.759) with GLP-1RA use vs 56.2% (HR, 0.438; 95% CI, 0.298-0.643) in those with a BMI less than 29.

Safety Events in Patients Taking a GLP-1RA vs Aspirin

Patients taking Aspirin had a higher rate of gastrointestinal bleeding (2.1%) vs those taking a GLP-1RA (2%; HR, 0.852; 95% CI, 0.809-0.898; P = .018). The same was true for gastric ulcers (0.55% vs 0.5%; HR, 0.815; 95% CI, 0.735-0.904; P = .038) and acute kidney injury (2.8% vs 1.15%; HR, 0.369; 95% CI, 0.348-0.391; P = .0001).

Those taking a GLP-1RA had higher incidence of nausea and vomiting (10.5% vs 9.7%; HR, 0.980; 95% CI, 0.957-1.004; P = .0001), diarrhea (6.8% vs 5.4%; HR, 1.131; 95% CI, 1.096-1.166; P = .0001), and abdominal pain (19.0% vs 16.3%, HR, 1.055; 95% CI, 1.036-1.075); P = .0001).

Patient Selection and Observation

The primary end point for this study was CRC incidence with a secondary end point of incidence of adverse events related to GLP1-RA and Aspirin use.

Patients 18 to 90 years old from January 1, 2000, to January 1, 2024, were included, with the exclusion of patients taking other non-steroidal aromatase inhibitors (NSAIDs) besides Aspirin as well as those taking any hypoglycemic medications aside from GLP1-RA. Additionally, those with CRC prior to the study window were excluded. After exclusions, 140,828 patients in each arm were matched 1:1 to either cohort A, GLP-1RA users, or cohort B, aspirin users.

The average age at index event was 58 ±13.5 in cohort A and 58 ±13.7 in cohort B (standard mean difference, 0.0004). Most patients were White (67% in each cohort; P = .002) and female (67% in each cohort; P = .001).

Colton and co-authors used data from the TrinetX database, containing information on 150 million patients from 106 health organizations. Patients from cohort A were matched to patients in cohort B via propensity score matching. Clinically meaningful sensitivity and subgroup analyses were performed, and those with outcomes before the study window were excluded.

Regarding data analysis, multivariate logistic regression assessed associations, and Cox proportional hazards models determined time to event outcome.

GLP-1RAs in CRC Risk Reduction

Colton noted that Aspirin has been investigated as a means of CRC prevention but due to its “modest efficacy” and increased risk of bleeding, its use is limited; further, GLP-1RAs have been shown to have anti-inflammatory and antineoplastic due to PI3K/AKT/mTOR pathway inhibition. For these reasons, Colton and co-authors set to explore the use of GLP-1RAs as primary prevention for CRC.

Disclosures: Colton Jones reported no relationships to disclose.

Reference

Jones CF, Obomanu E, Neely A, et al. Glucagon-like peptide-1 receptor agonist vs aspirin for primary prevention of colorectal cancer: evidence from a real-world head-to-head comparison. J Clin Oncol. 2026;44(suppl 2):18. doi:10.1200/JCO.2026.44.2_suppl.18