GLP-1 Agonists and the Risk of Pulmonary Aspiration during Elective Upper Endoscopy: A Systematic Review and Meta-analysis

Glucagon-like Peptide-1 (GLP-1) receptor agonists, such as semaglutide, and synthetic dual agonists targeting both glucose-dependent insulinotropic polypeptide and GLP-1 receptors, such as tirzepatide, have been gaining popularity due to their weight loss properties and related benefits. Due to their potential as an alternative to bariatric surgery, patients frequently request prescriptions from their healthcare providers [1, 2]. However, the prevalence of nausea and other gastrointestinal symptoms is often the primary factor limiting continued use [3, 4]. Despite this, evidence in the literature suggests that these side effects are typically temporary and tend to resolve with continuation.

Glucagon-like peptide-1 (GLP-1) agonists work by enhancing insulin secretion and inhibiting glucagon release. Locally in the gastrointestinal tract, they delay gastric emptying by reducing antral and duodenal motility while increasing pyloric tightening. During anesthesia, delayed gastric emptying is a known risk factor for pulmonary aspiration [5]. Semaglutide, administered as a once-weekly injection, has gained significant attention from both medical practitioners and patients due to its favorable safety profile and low incidence of gastrointestinal side effects [1].

While the effects of GLP-1 agonists have been extensively studied in the context of weight loss and glucose regulation, their impact on delayed gastric emptying and the associated risks of Retained Gastric Contents (RGC) and pulmonary aspiration, remains less understood. For patients on GLP-1 agonists who require elective, urgent, or emergent procedures, optimal preoperative management is still under debate. Currently, there are no evidence-based guidelines, particularly regarding endoscopic procedures, on whether to hold these medications or how to adjust fasting protocols for this patient population.

Fujino et al. described a case of delayed gastric emptying in an obese 31-year-old female patient taking semaglutide for type 2 diabetes mellitus, even after following the American Society of Anesthesiologists recommendations for fasting guidelines. On endoscopy, food residue was found in the gastric body, and the procedure had to be aborted [6]. Strategies to decrease the risk of aspiration, like holding the medication up to four weeks before a scheduled procedure, were proposed by Gulak et al. after a case of a nondiabetic, nonobese patient with unexpected regurgitation of a large volume gastric content, despite a 20-hour solid and an 8-hour liquid fasting liquid [7]. Sherwin et al., using point-of-care (POC) ultrasound, showed that GLP-1 alters gastric emptying and retained gastric contents (RGC) two hours after clear liquid intake and after an overnight fast for solid food. This is concerning for patients who typically fast overnight before a procedure requiring anesthesia [8]. In a single-center electronic chart review, Silveira et al. found that patients taking semaglutide had increased residual gastric contents during elective esophagogastroduodenoscopy. However, interruption of preoperative semaglutide for 10–14 days was not predictive of increased RGC (24% vs. 5%) [9].

However, the American Gastroenterological Association, American Association for the Study of Liver Diseases, American College of Gastroenterology, American Society for Gastrointestinal Endoscopy, and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition released a joint statement on August 11, 2023, noting a lack of concrete evidence on best practices for patients on GLP-1 agonists and insufficient data on whether to stop medication before endoscopy [10]. They encouraged further studies focusing on these endpoints. The American Society of Anesthesiologists has provided consensus-based guidance suggesting holding GLP-1 agonists for one week in patients on weekly dosing and for one day in those on daily dosing prior to endoscopic or surgical procedures to reduce aspiration risk [11]. We, therefore, conducted a meta-analysis on pulmonary aspiration in patients on GLP-1 agonists undergoing endoscopy to guide perioperative management and fasting recommendations.

This review is reported following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, as indicated in the PRISMA checklist [12]. This systematic review, along with meta-analysis, is registered in the PROSPERO international database (ID CRD42024595241; www.crd.york.ac.uk /prospero↗).

Our systematic review and meta-analysis of seven retrospective studies and five case reports aimed to evaluate the adverse effect of GLP-1 agonists on pulmonary aspiration events during endoscopic procedures. The odds of aspiration in patients on GLP-1 receptor agonists were 1.45. However, the results were not statistically significant (P = 0.47). The case reports predominantly involved semaglutide, suggesting it may be a significant contributor to pulmonary aspiration among patients on GLP-1 agonists. In these reports, aspiration occurred despite withholding semaglutide for more than 6 days and across a range of doses, indicating that neither dose nor withholding duration alone reliably mitigated the risk. If gastric emptying from semaglutide, a long-acting agent, achieves tachyphylaxis over time, pulmonary aspiration may be due to factors other than GLP-1 agonist use.

In our manuscript, the majority of observational studies used long-acting GLP-1 agonists. There was no difference in pulmonary aspiration based on long-acting GLP-1 agonist use.

Compared to our meta-analysis, which included a broader population and various GLP-1 agonists, the case reports may have introduced a bias toward semaglutide-specific effects. However, since case reports typically highlight severe or unusual outcomes, their inclusion might have exaggerated the perceived risk and may not reflect the true incidence in larger populations. This is reflected in our pooled meta-analysis estimate, which shows no statistically significant increase in aspiration risk (OR 1.23; 95% CI 0.58–2.60) and substantial heterogeneity (I² = 75%).

Therefore, while the case reports provide important clinical context, they should be interpreted with caution given their inherent limitations, including lack of control groups, publication bias, and selective reporting.

A high level of heterogeneity was observed in our meta-analysis. This may be attributed to differences in the type of GLP-1 agonist used, duration of use, the length of time the medication was held prior to the procedure, and fasting periods. Additionally, the depth of sedation—including monitored anesthesia care, general anesthesia, and deep sedation—varied across studies, as did the methods used to identify aspiration events. Since pulmonary aspiration risk is strongly influenced by baseline patient characteristics, factors such as indications for GLP-1 agonist use (e.g., diabetes and obesity) and underlying comorbidities like GERD and autonomic neuropathy may also contribute to the observed heterogeneity.

Pulmonary aspiration is a known complication of gastrointestinal endoscopy, occurring at a rate of 4.6 per 10,000 endoscopies in a retrospective study by Bohman et al. Upper airway manipulation and stomach insufflation during the procedure can increase regurgitation and the risk of aspiration [30]. In our meta-analysis, the placebo group had nearly three times the aspiration rate compared to Bohman et al. and similar studies. This elevated risk in the placebo group may be attributed to several patient-related factors that differ significantly from the general endoscopy population in prior research. Conditions, such as diabetes, obesity, sedation-related suppression of airway reflexes, and comorbidities like autonomic neuropathy could contribute to this increased risk. The placebo group may represent a subset of patients with higher metabolic and gastrointestinal risk profiles, rather than reflecting the broader, generally healthier population undergoing routine endoscopy, as studied by Bohman et al.

Interest in aspiration risk associated with GLP-1 agonists arose after observations that patients on these medications frequently had residual gastric contents despite adhering to standard fasting guidelines [31]. Stark et al. reported a fourfold increase in residual gastric content, with 4/49 patients on GLP-1 agonists compared to 2/188 in the control group [31]. Similarly, Kobori et al. found a tenfold increase in residual gastric content in their propensity-matched study, 5.4% versus 0.49% in controls (P = 0.004) [32].

However, studies examining residual gastric contents and aspiration events related to GLP-1 agonists remain small in size,with some lacking statistical significance [19, 31, 32]. Case reports have documented residual gastric contents despite fasting longer than usual in patients on GLP-1 receptor agonists [33]. There is speculation that the recent initiation of GLP-1 agonists may increase the incidence of delayed gastric emptying, while tachyphylaxis from vagal nerve activation reduces these effects with long-term use or higher doses [11, 34].

Short-acting GLP-1 agonists (exenatide, lixisenatide) primarily delay gastric emptying, leading to suppression of postprandial hyperglycemia. The side effects of gastric emptying are more pronounced with short-acting formulations than with long-acting ones (liraglutide, injectable semaglutide, weekly exenatide, lemaglutide, dulaglutide, albiglutide). In a study by Drucker et al. comparing twice-daily short-acting exenatide with weekly exenatide [35], short-acting exenatide maintained its delayed gastric emptying effect after 14 weeks of continuous administration, whereas long-acting exenatides showed evidence of tachyphylaxis over days to weeks. There is a lack of studies examining the relationship between short-acting GLP-1 agonists and pulmonary aspiration in the literature. It is also speculated that in patients with underlying gastroparesis, GLP-1 agonists may further reduce gastric emptying.

The American Society of Anesthesiologists, in their guidance released on June 29, 2023, recommends holding GLP-1 agonists for one week before the procedure in patients on weekly dosing, and one day before in those on daily dosing. However, these recommendations are based on expert opinion rather than strong evidence [11].

The ASA also recommends monitoring upper gastrointestinal symptoms and fasting for 8 hours on solids and 2 hours on liquids in patients without upper GI symptoms. However, this may be inadequate, as not all patients with RGC exhibit upper GI symptoms [11]. Similar to the joint statement from gastroenterology societies, there is insufficient concrete evidence linking GLP-1 agonist use to an increased risk of pulmonary aspiration. Therefore, based on our study results, we cannot recommend withholding GLP-1 agonists prior to procedures.

Recommendations for a 24-hour clear liquid diet to reduce aspiration events have been proposed based on studies by Ghazanfar et al. [29], Silveira et al. [9], Nasser et al. [36], and Maselli et al. [25]. However, additional evidence directly comparing pulmonary aspiration rates between a 24-hour clear liquid diet and the standard 8-hour fasting period, especially in patients with diabetes and obesity, is needed before this strategy can be widely implemented for those on GLP-1 agonists.

Standard fasting times help reduce the risk of aspiration during procedures. Some studies and case reports suggest using POC ultrasound to detect RGC in patients on GLP-1 agonists, especially those who are symptomatic. Aborting procedures in patients with RGC can help avoid complications by delaying the procedure or modifying anesthetic plans [8, 37]. Until evidence-based guidelines define the optimal duration for holding GLP-1 agonists before procedures, POC ultrasound may be a useful tool to assess aspiration risk. In a study of twenty patients, Sherwin et al. demonstrated that POC ultrasound detected RGC even after an 8-hour fast [8].

ImplementingPOC ultrasound in clinical practice requires training gastroenterologists or anesthesiologists to identify RGC. This presents several challenges, including the need for structured training, a thorough understanding of gastric physiology, and managing subjective interpretation. Additionally, trainees must be familiar with anatomical variations, such as gastric bypass, obesity, and other causes of altered anatomy. Other factors to consider include logistics, time constraints, equipment availability, and the development of standardized protocols for clinical decision-making.

The type of sedation used can also affect the risk of aspiration. Rezaiguia-Delclaux et al. state that propofol is associated with a higher risk due to impaired airway protection reflexes [38]. Intubation can protect the airway and decrease the risk of aspiration. However, endoscopies are usually performed under Monitored anesthesia care (MAC)/ deep sedation, leaving the airway unprotected and causing loss of protective airway reflexes leading to aspiration. This is evidenced in the retrospective study by Bohman et al., where MAC sedations were associated with the highest rate of aspirations (6.9%) compared to general endotracheal intubation (6.4%) and registered nurse sedation (moderate intravenous (IV) sedation with fentanyl and midazolam) (0.9%).

The risk of pulmonary aspiration in patients on GLP-1 agonists is still inconclusive. Pre-procedural patient assessment, symptomatic evaluation, and, if needed, POC ultrasound may help identify patients at risk of aspiration events. There is limited evidence in the literature on the use of pro-motility agents to decrease the risk of pulmonary aspiration in patients on GLP1 agonists undergoing medical procedures. Promotility agents like metoclopramide and erythromycin may be useful for emergency procedures based on POC ultrasound findings or patient's comorbidities for the high risk of retained gastric contents. Singh et al. concluded in their study that the management of GLP-1 in the perioperative period should be individualized and based on communication with anesthesiologists before endoscopies [39]. Anesthesiologists should be quick to recognize gastric contents during emergency procedures and take measures to avoid aspiration, including rapid sequence induction, head elevation, and intubation if necessary. Rescheduling elective procedures may also be considered.

We conducted a rigorous meta-analysis following established methodologies and adhering to the PRISMA guidelines. Additionally, we registered our study in PROSPERO to ensure transparency and reproducibility. We strictly followed the pre-defined study protocol throughout our meta-analysis. The majority of the evidence in our manuscript included case reports, retrospective case-control, and cohort studies, which are inherently associated with a risk of bias, particularly when confounding factors are not adequately adjusted [40, 41]. Due to the retrospective nature of these studies, it is important to note that our findings do not establish a causal relationship between GLP-1 agonists and pulmonary aspiration.

It is worth mentioning that all the case reports and retrospective studies focused on long-acting GLP-1 agonists, and there was a notable lack of evidence regarding short-acting GLP-1 agonists. Further, many of the studies did not have information on the dose, duration of usage, duration of holding GLP-1 agonists, and fasting period prior to the procedure. Depth of sedation has a significant effect on the risk of aspiration during endoscopy procedures. Multiple studies in our meta-analysis did not mention the depth of sedation and anesthesia used. Importantly, aspiration events were identified using ICD codes in some studies, while other studies did not clearly define the method of identifying aspiration, introducing variability in outcome assessment. These factors could have contributed to the high heterogeneity noted in our meta-analysis.

Although the funnel plot demonstrates some asymmetry, potentially suggesting publication bias, our exhaustive literature search did not identify any unpublished studies that met the inclusion criteria. The asymmetry observed on funnel plots may be attributable to methodological heterogeneity, including variation in sample sizes, population differences, and outcome definitions across the included studies. It is also possible that the skewed distribution reflects selective reporting and inherent differences in effect sizes among smaller studies. If unpublished studies with null or negative findings had existed and been included, the overall effect estimate might have been attenuated. These considerations highlight the importance of interpreting our findings in the context of the available published evidence.

Furthermore, the majority of the studies analyzed were conducted in the United States; however, we believe that our findings are generalizable to broader populations.

Due to conflicting evidence and the small sample sizes of current studies, further retrospective research is needed. Future studies should aim to clarify the relationship between GLP-1 agonist use and aspiration risk, including factors such as duration of use, type of GLP-1 agonist (short- vs. long-acting), and underlying comorbidities like diabetes and obesity. Key variables to investigate include fasting times, residual gastric contents, and aspiration events. Additionally, risk should be assessed for individual drugs, the time required to develop tachyphylaxis to delayed gastric emptying, and the optimal cessation period needed to normalize gastric emptying. The effective fasting duration prior to procedures in patients on GLP-1 agonists also warrants further investigation.

In patients at high risk for retained gastric contents, extending the fasting period by a few additional hours beyond current recommendations may not consistently reduce gastric volume. This warrants further investigation. The potential benefits of withholding solid food and adopting a 24-hour clear liquid diet—similar to preparation for colonoscopy, should be evaluated. Additionally, the efficacy of pro-motility agents, such as metoclopramide and erythromycin, in accelerating gastric emptying in these patients should be explored in future studies [42].

This systematic review and meta-analysis found no statistically significant association between GLP-1 agonist use and increased risk of pulmonary aspiration during endoscopic procedures, though substantial heterogeneity was found among the included studies. Given the limitations in study design, inconsistent definitions of aspiration events, variation in GLP-1 agonist used and the absence of data on short-acting GLP-1 agonists, the evidence remains inconclusive.

Despite this, clinical vigilance is needed during endoscopy, particularly in patients with diabetes, obesity, or other risk factors for delayed gastric emptying. Until more robust evidence is available, we recommend individualized risk assessment for patients on GLP-1 agonists undergoing endoscopy. Point-of-care (POC) ultrasound may serve as a useful tool to identify retained gastric contents in high-risk or symptomatic patients, helping inform peri-procedural decision-making. Future studies are needed to determine optimal fasting times, the role of specific GLP-1 formulations, and the effectiveness of alternative strategies, such as clear liquid diets or promotility agents in minimizing aspiration risk.