Tirzepatide Outperforms Rival GLP-1 Drugs on Effectiveness Reports in Real-World Safety Data
A pharmacovigilance analysis of EudraVigilance finds tirzepatide has lower 'drug ineffective' reporting odds than older GLP-1 receptor agonists.
Summary
Researchers mined the European Union's drug safety database (EudraVigilance) to see how often tirzepatide, the dual GIP/GLP-1 weight-loss and diabetes drug, is reported as not working or being used outside its approved purpose. They compared it with other GLP-1 drugs such as semaglutide and liraglutide. Tirzepatide had lower odds of 'drug ineffective' reports than liraglutide, dulaglutide, exenatide and semaglutide in the full dataset. Among reports from healthcare professionals, it did not differ from semaglutide. Off-label use was the most common issue, with 1,521 reports. Most reports came from women, and about two-thirds of tirzepatide reports came from outside Europe. Because these are voluntary reports, the findings show reporting patterns, not true rates or causes.
Detailed Summary
Tirzepatide, a dual GIP and GLP-1 receptor agonist, was authorized by the European Medicines Agency in September 2022 for type 2 diabetes. The FDA added weight management in November 2023 and moderate-to-severe obstructive sleep apnea in obesity in December 2024. Randomized trials show it beats placebo and other GLP-1 drugs for weight and glycemic control. Even so, 9–15% of SURMOUNT-1 participants missed 5% weight loss at 72 weeks, and off-label use for cosmetic weight loss is spreading through social media. Real-world surveillance is needed to see how often these problems arise.
The authors analyzed publicly available, aggregated individual case safety reports in EudraVigilance through February 8, 2026. They compared tirzepatide with exenatide, dulaglutide, lixisenatide, liraglutide and semaglutide. They selected 57 MedDRA Preferred Terms in advance, covering suboptimal outcomes (such as lack of efficacy) and drug-use issues (such as off-label use, intolerance and non-compliance). Reporting odds ratios (RORs) were calculated for the full dataset and for healthcare-professional (HP) reports alone. A signal required more than 3 cases and a lower 95% CI bound above 1.
Of 131,586 reports across the six drugs, semaglutide accounted for 36.73% (48,335) and tirzepatide for 20.99% (27,617). For tirzepatide, 48.2% of reports involved adults aged 18–64 and 11.1% involved adults aged 65–85. About two-thirds (17,510) concerned women, versus 6,482 for men. More than 68% came from outside the European Economic Area, and more than 61% came from HPs. Gastrointestinal disorders dominated in volume (21,882 non-serious, 10,192 serious). Hepatobiliary (97%), metabolism and nutrition (84%) and psychiatric (84%) categories had the highest share of serious reports.
Among the selected terms, 'Off-label use' was reported 1,521 times, 'Drug ineffective' 425, 'Off-label use of device' 99, 'Drug intolerance' 60, 'Product use in unapproved indication' 15, 'Treatment non-compliance' 5 and drug resistance 2. Ineffectiveness reports mostly came from non-HPs. HP reports had lower odds than non-HP reports for 'Drug ineffective' (ROR 0.28, 95% CI 0.23–0.35), 'Off-label use' (0.44, 0.40–0.49) and 'Off-label use of device' (0.19, 0.12–0.31).
In the full dataset, tirzepatide had lower odds of 'Drug ineffective' than liraglutide (ROR 0.61, 95% CI 0.54–0.70), dulaglutide (0.72, 0.63–0.82), exenatide (0.78, 0.67–0.92) and semaglutide (0.81, 0.72–0.91). In HP reports, there was no difference versus semaglutide. 'Off-label use' had higher odds versus lixisenatide (3.40), dulaglutide (2.35) and liraglutide (1.67) but lower odds versus semaglutide (0.52, 0.49–0.55). 'Drug intolerance' was lower versus semaglutide (0.34, 0.26–0.45). The 'Off-label use of device' signal versus semaglutide (ROR 43.47, 95% CI 16.00–118.13) is highly uncertain.
These findings broadly agree with trial evidence of strong efficacy, and they flag off-label use as a real-world issue to monitor. But spontaneous reports cannot establish causation, incidence or true frequency. Reporting is shaped by regulatory alerts, media attention, notoriety bias, differing prescribing patterns and the fact that tirzepatide is newer than the comparators. Comparisons between drugs reflect reporting behavior as much as drug performance.
Key Findings
- Tirzepatide accounted for 27,617 of 131,586 EudraVigilance GLP-1 drug reports (20.99%), second only to semaglutide (36.73%, n=48,335)
- 'Off-label use' was the most reported term (n=1,521), followed by 'Drug ineffective' (n=425) and 'Off-label use of device' (n=99)
- Tirzepatide had lower odds of 'Drug ineffective' reports than liraglutide (ROR 0.61), dulaglutide (0.72), exenatide (0.78) and semaglutide (0.81) in the full dataset
- In HP-only reports, 'Drug ineffective' odds did not differ between tirzepatide and semaglutide, while odds stayed lower versus dulaglutide (ROR 0.32)
- Approximately two-thirds of tirzepatide reports (n=17,510) involved women versus 6,482 for men; 68% came from outside the EEA and 61% from HPs
- 'Drug intolerance' reports were less likely for tirzepatide than semaglutide (ROR 0.34, 95% CI 0.26–0.45)
- Hepatobiliary (97%), metabolism/nutrition (84%) and psychiatric (84%) categories had the highest share of serious tirzepatide reports
Methodology
This was a retrospective pharmacovigilance study using publicly available aggregated EudraVigilance data through February 8, 2026 (tirzepatide reporting began July 22, 2022). It analyzed 131,586 reports for six GLP-1 receptor agonists, including 27,617 for tirzepatide, and 57 MedDRA Preferred Terms chosen before data extraction. Disproportionality was measured with reporting odds ratios (95% CI) from 2×2 tables in MedCalc. A signal required more than 3 cases and a lower CI bound above 1, and analyses were run on all reports and on HP-only reports. There was no control for confounding, because only aggregated data were available.
Study Limitations
The authors note that causality, incidence and true event frequency cannot be established from spontaneous reporting data. Estimates may be distorted by reporting practices, regulatory alerts, scientific attention, notoriety bias and different prescribing patterns, and some CIs (such as the device-related term) are very wide. Conflict-of-interest details were not included in the supplied text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe:
