GLP-1 Receptor Agonists and Perioperative Care: Comment

To the Editor: We read with mixed feelings the recently published multisociety clinical practice guidance on perioperative management of glucagon-like peptide-1 (GLP-1) receptor agonists (RAs).1,2 While it was gratifying to see a reached consensus, it concerned us that relevant evidence was not taken into consideration. For nearly 80 yr, since the days of Curtis Mendelson,3 anesthesiologists have been aware of the dangers of inducing anesthesia in “full stomach” patients. Hence, the most important variable upon which guidance ought to be based is the consistently higher incidence of increased residual gastric content in GLP-1 RA users (compared to nonusers)4 despite adherence to standard perioperative fasting guidelines. Indeed, increased residual gastric content is, in our opinion, the crux of the problem. It is not only the single most important risk factor, but also the only necessary (but thankfully not sufficient) condition for the development of bronchoaspiration under sedation/anesthesia. Accordingly, the presence of residual food during upper endoscopy increased the aspiration risk by 10-fold in one study.5 In contrast, several consensus recommendations appear to be based on studies quantifying GLP-1 RA–induced “delay in gastric emptying” and/or “gastroparesis,”2 which serve as proxies for (but are not equivalent to) the real outcome of interest—increased residual gastric content. Of note, a recent meta-analysis of gastric motility assays found that GLP-1 RA use resulted in a modest (36 min) gastric emptying delay as measured by (the gold standard)6 scintigraphy,7 which cannot explain the persistence of undigested gastric food after 8 h of fasting. Gastroparesis may, therefore, play only a partial role in the significantly higher incidence of perioperative increased residual gastric content in fasted GLP-1 RA users compared to nonusers.8–13 Other factors must be involved, whose nature remains unclear and beyond the scope of this discussion. This is not a singular issue. In a study involving 2,991 patients who underwent both elective esophagogastroduodenoscopy and scintigraphy gastric motility assessment, 68% of patients with endoscopically visible residual gastric content had normal gastric emptying.14 Hence, recommendations on perioperative management of GLP-1 RAs (including risk stratification and mitigation strategies) should ideally be based on studies focused on the primary outcome of interest (increased residual gastric content) rather than the experimentally measured “gastroparesis.”1,2 This distinction is important. For example, recommendation 1a suggests that some variables (therapy [escalation vs. maintenance] phase, [higher] dose regimens, gastrointestinal symptoms) are associated with higher risk of delayed gastric emptying and bronchoaspiration.2 As a corollary, recommendations 1b and 2 subsequently seem to imply that patients without these aggravating factors are “without elevated risk of delayed gastric emptying and aspiration” and could therefore be presumed to have an “empty stomach” and managed without any additional precautions.2 Such statements, however, are not supported by studies focused on the outcome of interest—increased residual gastric content. Three studies found no correlation between GLP-1 RA therapy duration and the presence/absence of increased residual gastric content.15–17 The average GLP-1 RA dose was also equivalent among patients who presented with and without increased residual gastric content in several reports.12,18,19 Similarly, while the “presence of gastrointestinal symptoms” should be considered “as elevating the risk of delayed gastric emptying and aspiration,”2 such association has been long known and applies to all perioperative patients, GLP-1 RA users or not.8,10 Importantly, even asymptomatic patients taking GLP-1 RAs had a significantly higher risk of presenting with increased residual gastric content compared to nonusers.10,19 As of yet, no studies have reported a difference in the incidence of increased residual gastric content that would justify treating patients differently on the basis of dose regimen and/or treatment phase, as recommendation 1a indicates.1,2 Similarly, the absence of ongoing digestive symptoms, while somewhat reassuring, should not be taken as a definitive sign of an empty stomach in GLP-1 RA users. While, individually, the cited reports have important limitations, collectively, they seem to point in the same direction and, therefore, must not be disregarded. Given the limited available data and the catastrophic consequences of bronchoaspiration, it would be prudent to err on the side of caution and consider their findings as more likely than not until more definitive data become available. So, what can clinicians do? In our opinion, the priority is to find a measure (or a combination of measures) capable of reducing/normalizing residual gastric content among GLP-1 RA users in the perioperative period. While larger prospective studies are urgently needed, emerging evidence suggests that a liquid-only diet for 24 h preoperatively may be effective for that matter.8,10,18,20–23 Additionally, preoperative GLP-1 RA interruption has been studied/proposed,1,2,8,16 but the level of supporting evidence remains weak/controversial, particularly with regards to ideal (and clinically practical) preoperative interruption intervals that would safely balance the risk of perioperative hyperglycemia against bronchoaspiration on anesthesia induction. Until enough evidence to support a reliably preventive strategy becomes available, gastric ultrasound appears a viable alternative to assess residual gastric content at the bedside (although keeping in mind every diagnostic test has both false negatives and false positives). Whenever in doubt, caution should be exercised, and the ultimate decision (e.g., case postponement or proceeding with rapid sequence tracheal intubation) should be made on an individual basis. Notably, bronchoaspiration may occur in GLP-1 RA users despite (or even during) rapid sequence intubation.24 In conclusion, we posit that increased residual gastric content is the crux of the problem and should, therefore, be used as the primary endpoint to guide recommendations pertaining to perioperative GLP-1 RA use. Because GLP-1 RA users are more likely to present with increased residual gastric content despite adhering to perioperative fasting guidelines, there is presently no evidence to justify not treating all GLP-1 RA users with similar precautions. The best mitigating strategies remain debatable. Whatever these strategies might be, however, they ought to be applied equally to all GLP1 RA users, irrespective of dose, therapy duration, presence/absence of symptoms, or any other factors, at least until more evidence becomes available. Competing Interests The authors declare no competing interests.