Holding a GLP-1 dose before endoscopy cut retained stomach contents in Cleveland Clinic trial

In a 68-patient randomized trial, one in four people who kept taking their GLP-1 medication as usual before an upper endoscopy had clinically significant retained stomach contents after standard fasting, compared with 3.1% of those who skipped their last dose ^[1]. The results were published in JAMA Internal Medicine ^[1].

The OCULUS trial ran at two Cleveland Clinic centers and enrolled patients scheduled for esophagogastroduodenoscopy, endoscopic ultrasonography or endoscopic retrograde cholangiopancreatography between July 2024 and May 2025 ^[1]. The rate of clinically significant residual gastric volume was 25% among continuers versus 3.1% among those who held their medication, an absolute difference of 21.9 percentage points (90% CI 7%-36.7%, P=0.003) ^[1].

The trial defined clinically significant residual gastric volume as a composite: stomach contents that prevented the endoscopist from completing the exam, required stopping early or placing a breathing tube, or led to an aspiration event needing extended monitoring, unplanned treatment or hospital admission ^[1]. Every case in this trial fell into the first category, meaning the exam could not be done, and there were no aspiration events or unplanned intubations ^[1].

Subgroups differed. Among 35 patients who had only an upper endoscopy, 46.7% of continuers had significant residual volume versus 5% of holders (P=0.001) ^[1]. Lead author Tilak Shah, MD, MHS, of Cleveland Clinic Florida, said that rate "was far higher than we anticipated" ^[1]. In a separate group of 25 patients who had both an upper endoscopy and a colonoscopy, no one had a significant residual volume event, whether or not they held the drug; those patients followed a clear liquid diet plus bowel preparation ^[1].

How long a dose was held mattered for weekly drugs. Among patients on a weekly medication, 92.7% of those who held their dose more than three days had a clear stomach, compared with 62.5% of those who held three days or less (P=0.047) ^[1]. In the study, holders went without medication for 7 to 13 days for weekly drugs or at least one day for daily drugs; continuers took their usual dose on their usual schedule ^[1]. The single patient in the hold group who still had significant residual volume was taking daily oral semaglutide (Rybelsus) ^[1].

Symptoms did not predict who had food left in the stomach. The researchers found no link between gastrointestinal symptoms and retained contents; Shah said all of the patients with retained solid food had no upper gastrointestinal symptoms, and the one patient who reported nausea and vomiting did not have increased residual volume ^[1]. "Our study questions the validity of considering symptoms in decision-making," he said ^[1].

Average age was 63, about half of participants were female, and most were taking injectable semaglutide (Ozempic, Wegovy) or tirzepatide (Mounjaro, Zepbound) ^[1]. People with gastroparesis risk factors other than diabetes were excluded, which the authors note may limit how widely the findings apply, and the trial was not powered to compare individual GLP-1 medications or doses ^[1].

The American Society of Anesthesiologists said in 2024 that most patients should not stop GLP-1 drugs before elective surgery, reversing 2023 guidance that advised a one-week hold for injectables and a one-day hold for oral agents ^[1]. Current recommendations generally allow patients to continue medication up to the day of surgery while following a liquid diet for 24 hours beforehand, depending on circumstances; Shah said the trial adds evidence supporting that approach, since no patient who consumed clear liquids the day before had retained gastric contents ^[1].

Why it matters for patients

GLP-1 medications slow stomach emptying, and leftover stomach contents during sedation raise the risk of regurgitation and aspiration into the lungs ^[1]. In this trial, retained contents meant the endoscopy could not be completed as planned, so the practical result was a stopped or rescheduled procedure ^[1]. The findings suggest the day-before clear liquid diet did more to empty the stomach than holding the drug alone, and that holding a weekly dose longer than three days was tied to a clearer stomach than holding three days or less ^[1].

The trial also found that stomach symptoms were not a reliable signal of retained food, which is notable because the anesthesiology guidance points to symptoms such as nausea and constipation as indicators of delayed gastric emptying ^[1]. What this means for any individual patient's pre-procedure instructions has not been settled by this trial, and the study excluded people with several gastroparesis risk factors ^[1].

What happens next

Shah's team is running OCULUS-2, which is examining whether these medications affect the quality of bowel preparation before colonoscopy ^[1]. No completion date or results have been disclosed. For patients who arrive for a procedure without holding a dose, Shah said the options discussed include rescheduling, or, if the procedure is urgent, rescheduling for the following day with clear liquids ^[1]. He also said point-of-care ultrasound could be investigated as a way to avoid rescheduling ^[1].