CHICAGO — Patients who develop acute pancreatitis while being treated with GLP-1 receptor agonists (RAs) show lower rates of recurrence if they continue with the treatment than if they discontinue, according to the results of a new study.

These findings are contrary to the FDA drug labels warning that GLP-1 RAs should be discontinued promptly if patients develop acute pancreatitis despite known rapid rebound effects, including the regaining of weight and reversal of other health benefits.

“In a large cohort of patients across the Cleveland Clinic Enterprise, we found no increase in recurrent pancreatitis among patients who continued GLP-1 RA medication after an episode of acute pancreatitis; in fact, their risk may actually be lower than patients who discontinued the medication,” senior study author Tilak Shah, MD, medical director of the Pancreas Center at Cleveland Clinic Weston Hospital in Weston, Florida, told Medscape Medical News.

The results of the real-world study were presented at Digestive Disease Week (DDW) 2026, in Chicago, along with other studies predominantly suggesting low risk — and possibly even protective effects — of GLP-1 RAs in pancreatitis.

Retrospective Cohort Study

Obesity and type 2 diabetes (T2D) are known to carry an increased risk for pancreatitis; however, data on the role that GLP-1 treatment can play in that risk is conflicting, with some case reports showing an association, but larger analyses failing to show a link.

To further investigate the issue, Shah and colleagues conducted a retrospective cohort study of 538 adults treated for idiopathic acute pancreatitis while receiving GLP-1 RA therapy in Cleveland Clinic between 2005 and 2023.

The study excluded patients with pancreatic cancer or acute pancreatitis due to alcohol, endoscopic retrograde cholangiopancreatography, and biliary etiologies.

Of the patients, 312 (58%) continued their GLP-1 therapy despite the pancreatitis diagnosis, while 226 (42.0%) discontinued the medication following the initial pancreatitis episode.

There were no significant differences between the groups in baseline characteristics, including age, sex, race, diabetes, hypertension, hyperlipidemia, and other factors.

The mean age was 57 years, and about 55% were female in both groups; rates of T2D were 87.8% and 84.5% in the GLP-1 continuation and discontinuation groups, respectively, with approximately 80% in each group having hypertension.

Over a follow-up of up to 90 days, 98 patients (18.2%) experienced a recurrence of acute pancreatitis.

The results showed that acute pancreatic recurrences were significantly more common among patients who discontinued GLP-1 RA therapy (22.1%) than among those who continued the drugs (15.4%; P = .046).

The only other factor significantly associated with an increased odds of recurrence was hypertension (odds ratio [OR], 2.75; P = .16).

The differences were consistent after a multivariate adjustment for factors including age, sex, diabetes, hypertension, hyperlipidemia, tobacco use, and alcohol use, with higher odds of recurrence among those discontinuing the drugs (adjusted OR, 1.65; P = .030).

“We hypothesized that even among patients who developed acute pancreatitis, GLP-1 [RAs] could be safely continued. These results support our hypothesis,” Shah said.

Mechanisms?

Shah speculated that GLP-1 RAs may have an effect on the fat accumulation that underlies pancreatitis, hence the potential protective effect.

“We believe an important cause of unexplained [idiopathic] pancreatitis is fat accumulation in the pancreas,” he said. “When fat accumulates in the pancreas, the pancreas is exposed to toxic fat metabolites that cause damage.”

“GLP-1 [RAs] likely reduce pancreatic fat deposition, which might explain the lower pancreatitis risk with continuing the medication.”

Overall, “there are now consistent data that GLP-1 [RAs] do not increase risk of pancreatitis,” he said.

“If a patient has a strong indication to take a GLP-1 [RA], such as diabetes or chronic kidney disease, then a history of pancreatitis should not be a reason to withhold the medication,” Shah said.

Similar Findings

Results from other analyses presented at the meeting support Shah’s findings, including a poster presented by Mustafa Atiq Arain, MD, of AdventHealth Medical Group Center for Interventional Endoscopy in Orlando, Florida. Arain and colleaguescompared two groups of 18,962 patients each with a history of acute pancreatitis who were and were not treated with GLP-1 RAs.

The results showed that those treated with the drugs had a significantly lower incidence of all-cause acute pancreatitis recurrence (OR, 0.125; P < .001). Recurrence rates were also lower in the GLP-1 groups in subsets of patients with prior alcoholic, biliary, drug-induced, and idiopathic etiologies than in those in the non-GLP-1 RA cohort.

Another study, said to represent the largest dataset to date assessing pancreatic risk in GLP-1 RA and SGLT2 users in adults with T2D and obesity, compared a cohort of 77,665 patients treated with GLP-1 RAs at centers in the Mayo Clinic Platform with 35,435 patients treated with SGLT2 inhibitors. The researchers found the same low rates of all-cause acute pancreatitis of 0.3% in both groups, with no significant differences at 6 and 12 months.

While GLP-1 users had lower incidences of chronic pancreatitis (0.2% vs 0.3%; relative risk [RR], 0.7), idiopathic acute pancreatitis (0.1% vs < 0.01%; RR, 0.07), and acute biliary pancreatitis (0.04% vs 0.07%; RR, 0.5), they did have higher incidences of gallstones (0.5% vs < 0.01%; RR, 29), pancreatic cancer (0.2% vs 0.1%; RR, 1.7), pancreatic cysts (0.4% vs 0.2%; RR, 1.7), and pancreatic disease (0.4% vs 0.2%; RR, 1.7; P < .05 for all).

The authors noted that the “higher gallbladder-related disease in GLP-1 RA users may reflect a higher proportion of females [55.3%] in this cohort [vs 46.1% in the SGLT2 cohort].”

Senior author Vivek Kumbhari, MD, of the Department of Gastroenterology and Hepatology at Mayo Clinic in Jacksonville, Florida, further added that substantial weight loss, itself, is associated with an increased risk for gallstones.

He agreed, however, that the evidence, including findings from Shah’s study, supports a rethinking of the GLP-1 RA discontinuation recommendations relating to acute pancreatitis.

“I think we really need to challenge this hypothesis that GLP-1s in a meaningful way elevate the risk of pancreatic disease,” he told Medscape Medical News.

“To me, it’s akin to the correlations such as those were reported early on with proton pump inhibitors and dementia, for instance,” he said.

“Whenever you take drugs en masse, there will be some correlations, but whether the cause and effect is clinically meaningful is a different story, and I wonder if that’s where we’re at here,” Kumbhari said.

In general, he added, patients who have diabetes have an underlying pancreatic dysfunction, “so that is huge confounding bias as to why they may develop pancreatic disease.”

In yet another poster presentation at the meeting, Fernando Gil López, also of Mayo Clinic, and colleagues conducted a systematic review and meta-analysis involving 16 studies, and a pooled total of more than 6 million patients in routine clinical settings, showing no significant increase in the risk for acute pancreatitis among those who were and were not treated with GLP-1 RAs (adjusted hazard ratio, 1.11).

That study did show a modest increase in the risk for biliary disease associated with GLP-1 RA use, but Lopez agreed that “in the future, we will see big changes [in guidelines] as far as the acute pancreatitis risk with this medication.”

“Every time we see patients with a history of pancreatitis, it has been exclusion criteria, and I think that will probably change,” he told Medscape Medical News.

A Caveat? FDA Adverse Event Reporting System (FAERS) Data

Of note, one study at the meeting looking at the FAERS data between 2019 and 2024 on reports of drug-induced pancreatitis showed that only three drug classes, including incretin mimetics (GLP-1 RAs), in addition to monoclonal antibodies and kinase or pathway inhibitors, accounted for nearly half of the reports.

The authors, including first author Akanksha Togra, of Texas Tech University Health Sciences Center El Paso at El Paso, Texas, and colleagues, concluded that “these findings emphasize the importance of heightened clinical vigilance for pancreatitis when prescribing these agents.”

Shah had no disclosures to report. Kumbhari’s disclosures included relationships with Boston Scientific, Medtronic, Fujifilm, MicroTech, ERBE, BariaTek, Endiatx, HydroGene Therapeutics, and Aurora Endosurgical. Lopez had no disclosures to report.