Considerations on peri-operative management of GLP-1 receptor agonists

We thank Hulst et al. [1] and Levy et al. [2] for their comments on our work [3]. Peri-operative use of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) has gained much attention recently.

We acknowledge the significance of power calculations in trial design to minimise type 2 error. It is essential to recognise, however, that power can be considered adequate with a smaller sample size if the effect size is large [4], as evidenced by our observed (clinically significant) differences between semaglutide users (40% incidence of increased residual gastric content) and non-users (3%, p < 0.001). For further clarification, we performed a post hoc power analysis based on our studied patients and parameter estimates. For an effect size of 0.7 based on the presence of increased residual gastric content in 43/107 semaglutide users and 3/113 non-users, and a significance level of 5%, using a χ² test, we achieved a critical χ² = 11.07 and a power (1-β error probability) > 0.9.

Our exclusion criteria accounted for conditions known to affect gastric emptying; hence our observed increased residual gastric content can be attributed primarily to semaglutide use. These exclusions strengthen, rather than limit, our findings' applicability. While we did not study patients with diabetes, we agree that future guidelines should focus on peri-operative management of GLP-1 RAs based on their primary use (weight loss vs. diabetes).

As for other classes of medications that can delay gastric emptying, it is impossible to call for revised societal guidelines for their peri-operative use when such guidelines do not exist. The impaired gastric emptying from these drugs has not been considered sufficiently relevant (unlike that induced by GLP-1 RAs) to warrant the attention of medical or anaesthesia societies to create specific guidelines.

Recently, several case reports have been published linking peri-operative GLP-1 RA use with bronchoaspiration and/or near misses [5]. Although anecdotal, it would be imprudent to disregard these reports and the growing body of evidence demonstrating a correlation between GLP-1 RA use and increased peri-operative residual gastric content [3, 6] as “lacking evidence” [2]. The recent development of multi-societal guidelines for the peri-operative management of GLP-1 RA [7] reflects this linkage.

Due to constraints inherent to our institutional protocol, we did not evaluate periods of discontinuation longer than 10 days and, consequently, were unable to make recommendations beyond this timeframe. Nevertheless, based on our findings (and other recent reports [6]), it does appear that 1-week pre-operative discontinuation suggested by the American Society of Anesthesiologists and other medical societies [7] may be insufficient to ensure an empty stomach. Indeed, while previous reports suggesting a 2–3 week semaglutide interruption pre-operatively were based primarily on pharmacologic principles, a recent study by our group suggests that similar intervals are required to reduce/normalise gastric content in semaglutide users [6]. While we acknowledge the limitations of recent reports, they align; therefore, given the catastrophic consequences of bronchoaspiration, it would be prudent to err on the side of caution until more definitive evidence emerges. Additionally, while peri-operative hyperglycaemia resulting from GLP-1 RA interruption remains a subject of debate [8], this can be mitigated by bridging regimens with less/no effect on gastric emptying. Notably, patients with diabetes on GLP-1 RAsare often already on insulin therapy in which case they can simply follow their baseline sliding scale, without the need for further delays (e.g. diabetologist consultation) as suggested by Levy et al. [2].

Hulst et al. suggested that gastric ultrasound has limited broad application. It is up to clinicians to decide whether this non-invasive tool, which has a rapid learning curve, should be part of their practice. In our cohort, gastric ultrasound prevented unnecessary surgical cancellations in semaglutide users. The peri-operative applicability of gastric ultrasound, particularly in the context of GLP-1 RA use, has gained attention recently and has been highlighted by anaesthesia and medical societies worldwide [7].

We recognise that ongoing digestive symptoms and semaglutide dosage/therapy duration may impact gastric emptying. Although we did not examine these variables, our study is the largest prospective cohort assessed via peri-operative bedside gastric ultrasound to date. Thus, despite its limitations, our findings are clinically significant.