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.

Ronald Turner

Appearances