Esophageal Food Impaction: Glucagon While Awaiting Endoscopy
A patient with an acute food bolus stuck in his chest, waiting for the endoscopy team, raises a familiar emergency-department question with a genuinely modest evidence base: does IV glucagon actually help often enough to justify trying it, or does it mostly just delay what endoscopy will do anyway.
Ben H., 46, was three bites into a rushed sandwich between showings at the real estate open house he was running solo this weekend when a piece caught and wouldn't go down or come back up — he drove himself to the emergency department still holding a stack of listing flyers on the passenger seat. He reports mild discomfort and an inability to swallow his own saliva, but no drooling, no stridor, and no respiratory distress; he's talking in full sentences and clearly uncomfortable rather than in extremis. He has a known history of an esophageal ring found incidentally on an unrelated CT years ago, never formally evaluated, and this is his second food impaction in three years — the first resolved on its own before he reached a hospital.
The endoscopy team has been notified and is roughly forty-five minutes out. IV glucagon relaxes smooth muscle, including the lower esophageal sphincter, through glucagon-receptor-mediated adenylate cyclase activation, and has been used for decades as a bridge therapy while awaiting endoscopy on the theory that relaxing the sphincter might let a lodged bolus pass on its own. But the actual trial evidence behind that practice is genuinely modest — several studies have found success rates not clearly better than placebo or spontaneous passage, and a real, if uncommon, side-effect profile including nausea and vomiting, which carries its own aspiration risk in a patient who can't swallow his own secretions. That gap between how often it's still reached for and how well it's actually shown to work is the real question in front of the team. Peksa and colleagues' meta-analysis found relief in 30.2% of glucagon-treated patients against 33.0% of controls — a point estimate favoring no treatment — alongside adverse events in 15% of the glucagon group and none of the comparators. The multicenter randomized trial by de Benito Sanz and colleagues has since reported the same result directly: resolution in 23.6% on glucagon against 20.6% on placebo. The evidence is not equivocal so much as consistently negative, which makes the persistence of glucagon as a reflex first move the more interesting question. Ben's incidental esophageal ring, mentioned almost in passing in his history, turned out to be the actual site of this impaction once the endoscopy team arrived — a detail that shifted the conversation, once the bolus was cleared, toward whether it now warranted the formal evaluation it had never received.
A bridge therapy with genuinely modest evidence behind it
I'd give glucagon now. It's low-risk, it's been used as a bridge therapy for decades, and there's a real chance, even if modest, that it lets this pass before the endoscopy team even gets here. He's stable and not in distress — this is exactly the low-stakes window where trying it costs us very little.
I want to be honest about what the actual trial evidence shows, because 'decades of use' isn't the same as 'shown to work.' Several studies comparing glucagon to placebo or to spontaneous passage haven't found a clear success-rate advantage. I'm going to be doing this endoscopy in forty-five minutes regardless of whether we give it — the real question is whether it's worth the risk for a benefit that hasn't been convincingly demonstrated.
The risk that concerns me specifically is nausea and vomiting, which is a real, described side effect — and he currently can't swallow his own saliva, so vomiting carries genuine aspiration risk right now, not a hypothetical one.
I think the disagreement is really about how to weigh a modest, uncertain benefit against a real but manageable risk. If we give it, keep him upright, have suction and an antiemetic ready, and treat the nausea risk as something to actively manage rather than something we're hoping doesn't happen. Given his current stability and the short window to endoscopy, I'd support skipping it and going straight to supportive management — the evidence doesn't clearly favor giving it, and forty-five minutes isn't long enough that avoiding the procedure meaningfully changes his course either way.
Agreed: supportive management and upright positioning while awaiting endoscopy, glucagon withheld given the modest trial-level evidence and the short window to definitive treatment. Endoscopy proceeded on arrival of the team and retrieved the bolus without complication; biopsies were taken given his history of an incidental, never-formally-evaluated esophageal ring.