GLP-1 medicines and pancreatitis: what the MHRA’s strengthened warning means

What the source reports

The UK Medicines and Healthcare products Regulatory Agency strengthened the product information for GLP-1 receptor agonists and dual GIP/GLP-1 receptor agonists on 29 January 2026. The update highlights that acute pancreatitis can, in rare reports, be necrotising or fatal.

The warning covers UK-authorised medicines containing dulaglutide, exenatide, liraglutide, semaglutide or tirzepatide. Acute pancreatitis was already a recognised possible side effect; the MHRA says its overall frequency remains uncommon.

The practical purpose of the change is earlier recognition. Abdominal pain, nausea and vomiting can be mistaken for the more familiar gastrointestinal effects of these medicines, potentially delaying investigation of pancreatitis.

What the 1,296 reports include

Between 2007 and October 2025, the MHRA received 1,296 Yellow Card reports of pancreatitis associated with this group of medicines. The total combines several reported diagnoses, including acute, chronic, autoimmune, haemorrhagic, necrotising, subacute and obstructive pancreatitis.

Within that total, 24 reports described necrotising pancreatitis and 19 involved a fatal outcome. These are serious safety reports and explain why regulators want patients and clinicians to recognise possible pancreatitis promptly.

They are not 1,296 independently confirmed cases proved to have been caused by a GLP-1 medicine. A Yellow Card records a suspicion that a medicine may be connected to an event, including reports submitted by healthcare professionals, patients, members of the public and pharmaceutical companies.

Why ‘nearly double’ does not mean the risk doubled

The BMJ reported that the cumulative total of 19 fatal reports was nearly twice an earlier total of 10. That comparison describes reports accumulated by different cut-off dates; it does not demonstrate that the fatal-pancreatitis rate for an individual taking one of these medicines nearly doubled.

Use of GLP-1-based medicines has expanded rapidly. For context, the MHRA estimates that roughly 25.4 million packs were dispensed in the UK during the five years to September 2025, and says most online-only pharmacies were not captured in that estimate. The 1,296 reports, however, span almost 19 years.

Packs are not people, the reporting and dispensing periods differ, and spontaneous reporting misses some events while publicity can stimulate others. Without a consistent exposed population, comparison group and adjudication of causes, dividing one number by another would produce a misleading estimate of personal risk.

What clinical research has found

Randomised trials have generally recorded very few pancreatitis events, making small differences difficult to measure. A 2024 meta-analysis of 21 placebo-controlled semaglutide trials involving 34,721 participants found no statistically significant increase in acute pancreatitis; its estimated odds ratio was 0.70, with a 95% confidence interval from 0.50 to 1.20.

A broader 2025 meta-analysis of 62 GLP-1 trials found a modest increase in its main pooled analysis, but that association was no longer statistically significant when the studies were separated according to background medication use. The authors also noted that many trials with no events in either group could not contribute to the pooled risk ratio.

A 2026 analysis designed to emulate a clinical trial found similar one-year rates of all-cause pancreatitis among people starting a GLP-1 medicine and those starting a sulfonylurea. Each type of evidence has limitations: trials may be too short or selective for rare harms, while health-record studies can retain differences between treatment groups.

Pancreatitis has several possible causes

Acute pancreatitis is inflammation of the pancreas and requires medical assessment. NICE estimates that approximately half of cases are caused by gallstones, about a quarter by alcohol and the remainder by other factors.

Other possible causes include high blood lipids, medicines, injury, infection, metabolic conditions and hereditary factors. Obesity and type 2 diabetes—the conditions for which GLP-1-based medicines are commonly prescribed—can also occur alongside some of these risk factors.

This is one reason an event reported during treatment cannot automatically be attributed to the medicine. It is also why an individual episode should not be dismissed as an ordinary medication side effect without appropriate assessment.

Symptoms the MHRA says require urgent attention

The MHRA advises patients to seek urgent medical attention for severe, persistent abdominal pain that may spread to the back and may occur with nausea or vomiting. The pain and its persistence are important because nausea or temporary stomach discomfort alone are common with GLP-1-based treatment.

For clinicians, the regulator advises stopping the GLP-1 or dual GIP/GLP-1 medicine immediately when pancreatitis is suspected and not restarting it if pancreatitis is confirmed. It also advises caution when prescribing to someone with a previous history of pancreatitis.

A privately prescribed medicine may not appear in an NHS medication record. Giving an urgent-care team the product name, dose and date of the most recent dose can therefore help them assess the full picture.

What remains uncertain

The available figures do not provide a reliable absolute rate of medicine-caused pancreatitis or fatal pancreatitis. They also do not show whether risks differ between individual products, doses, treatment durations or patient groups.

Research has not resolved whether there is no increased risk, a very small class-wide increase, or greater susceptibility in a particular subgroup. The MHRA and Genomics England are using the Yellow Card Biobank to investigate whether genetic differences may help explain why some people experience acute pancreatitis during treatment.

Longer randomised follow-up, consistently adjudicated diagnoses and studies that account for gallstones, alcohol, triglycerides, diabetes severity and weight change would improve the estimate. Continued reporting remains valuable precisely because very rare harms are difficult to study before widespread use.

What this means for the individual

The warning deserves attention without panic. It does not mean that everyone taking a GLP-1-based medicine is likely to develop pancreatitis, and it is not evidence that a person without symptoms needs to abandon an otherwise appropriate prescribed treatment.

Treatment decisions remain individual. A history of pancreatitis, gallstones, heavy alcohol use, high triglycerides, other medicines and the expected benefits of treatment can all affect a clinician’s assessment of benefit and risk.

OurGLP1 can help someone retain their medicine name, dose history and a timeline of symptoms to discuss with a qualified professional. It cannot diagnose pancreatitis or decide whether a medicine caused an event.

Bottom line

The MHRA has strengthened an existing warning because rare reports of pancreatitis have included necrotising disease and fatal outcomes. The action is real and the symptoms warrant urgent attention.

The reported totals are a safety signal, not a calculation of individual risk and not proof that every reported event was caused by treatment. The most accurate reading is neither dismissal nor alarm: recognise the symptoms, investigate suspected cases promptly and interpret cumulative Yellow Card numbers with their limitations clearly stated.