GLUCAGON-LIKE PEPTIDE 1 RECEPTOR AGONISTS (GLP1-RA) PRIOR TO ENDOSCOPY HAVE LOW GASTRIC RETAINED CONTENTS: A MULTI-CENTERED STUDY

Introduction:
Glucagon-like peptide receptor agonists (GLP1-RAs) are used to treat type 2 diabetes mellitus (T2DM) and obesity. Due to a gap of data, guidance from the GI societies and American Society of Anesthesiologist (ASA) remain in conflict on recommendations to holding GLP1-RAs prior to endoscopy (EGD). This multicenter study aims to address this equipoise and evaluate the effect of GLP1-RAs on food retention during upper EGD.

Methods:
This multicenter cross-sectional study included patients on confirmed GLP1-RAs receiving an EGD from 2021-2023. Demographics, GLP1-RA prescribing practices, and procedure outcomes were captured. GLP1-RA management of preoperative holding time was retroactively classified per recent ASA guidance. Logistic regression was performed to assess factors influencing retained gastric contents.

Results:
Of 815 patients identified (58% female, mean age 61 years), 81% had T2DM (Table 1a). A total of 70 (8.7%) patients had retained food in their stomach on EGD. Patients with retained food were more likely to have T2DM (93% vs 80%, p=0.009), higher mean HgA1C (7.6 vs 6.8, p<0.001), and pre-procedural blood glucose (148 vs 129, p=0.004). In these 70 patients, 10 (14%) had minimal (fully suctioned), 31 (44%) moderate (residue <50% gastric volume), and 29 (41%) large (residue >50% volume) amount of food residue. Only 1 (1.4%) patient required unplanned intubation and 14 (20%) had the procedure aborted. No aspiration events were recorded.

GLP1-RAs were held per ASA guidance in 406 patients (50%) (Table 1b). Those with GLP1-RA held were less likely to have retained food (4% vs 13%, p<0.001), but no significant difference to intubation (0% vs 2%, p=0.53) or aborting procedure rates (28% vs 18%, p=0.40). Of note 87% of patients continued on GLP1-RAs had no food retained. Patients with held GLP1-RAs had significant lower mean preoperative glucose levels (126 vs 136, p=0.01) but did not require correction.

In a multivariable model, likelihood of food retention increased 36% (CI 1.15-1.60) for every 1% increase in HgA1C and 3.45-fold (CI 1.50-7.91) when GLP1-RA was not held preoperatively. Differences by GLP1-RA type did not achieve statistical significance (Table 2).

Conclusion:
This study fills a highlighted gap to assess the impact of GLP1-RA on retained gastric contents for upper endoscopy. As retained contents during endoscopy is low at 8.7% with only 1.4% resulting in a clinically relevant change, a universal holding of GLP1-RAs prior to all endoscopies may be overly aggressive. We identify that the higher the HgA1C level, the increased odds of retained food regardless of holding medications prior to procedure.