Caustic Esophageal Injury: Do Corticosteroids Actually Prevent Stricture
A confirmed second-degree caustic esophageal injury reopens a genuinely old, still-unsettled question: does adding corticosteroids to standard supportive care actually reduce stricture formation, or does the evidence behind that practice not hold up as well as the tradition does.
Marcus D., 38, works maintenance at a commercial laundry facility, and reached for what he thought was a rinsed-out water bottle on a supply shelf during a rushed overnight shift — it was industrial-strength drain-line descaler, decanted into the bottle by a coworker's mistake days earlier and never labeled. He swallowed roughly a mouthful before the taste registered, spat out the rest, and was brought to the emergency department within the hour, alert, in pain, and able to describe exactly what happened. Endoscopy performed within 24 hours showed circumferential, deep ulceration with patchy exudate in the mid-esophagus without evidence of perforation — Grade 2b caustic injury on standard endoscopic grading, serious enough to carry a real, meaningful risk of stricture formation over the following weeks to months, but without the full-thickness necrosis that would push toward Grade 3 and a different, surgical conversation entirely.
Corticosteroids have been added to caustic-injury management for decades on the theory that they blunt the inflammatory and fibrotic cascade that produces stricture — a plausible mechanism, and one that showed some benefit in early observational series. But the more rigorous evidence since, including randomized data specifically in Grade 2b injury, has not consistently shown a stricture-rate reduction large enough to be confident the practice actually works, and some analyses suggest steroids may mask early signs of perforation or superimposed infection without a proven fibrosis benefit to justify that risk. That gap between decades of practice and what the more rigorous evidence actually supports is the real disagreement here, not a dispute over the injury grading itself. The trial most often cited on both sides is Anderson and colleagues' (New England Journal of Medicine, 1990), which assigned 60 children with caustic ingestion to corticosteroids or no corticosteroids and found strictures in 10 of 31 treated against 11 of 29 controls — no difference, with stricture tracking only the severity of the initial burn. Katibe and colleagues' later meta-analysis of three randomized trials reached the same place. The residual argument for steroids in Grade 2b specifically rests on smaller, non-randomized series rather than on any subgroup of Anderson's own data, and none of it was gathered in adults presenting the way Marcus is today. Marcus was also asked, once his airway and hemodynamic status were confirmed stable, whether the mislabeled-bottle exposure at his workplace had happened before to anyone else on his shift — it hadn't, as far as he knew, though the team encouraged him to raise it with his employer regardless, a detail outside the clinical decision itself but one the treating team didn't want to leave unmentioned.
Whether decades of practice is actually backed by the evidence
I'd add a corticosteroid course here. The theory has been around for decades — blunting the inflammatory and fibrotic cascade that produces stricture — and this is exactly the injury grade, 2b, where that theory was originally aimed. He's 38. A stricture that narrows his esophagus for the rest of his life is a real, life-altering outcome, and I don't want to withhold something plausible on the chance it doesn't work.
I understand the instinct, but I want to be direct about what the more rigorous evidence actually shows. The randomized data specifically in Grade 2b injury — not the older observational series the practice originally grew out of — has not consistently shown a stricture-rate reduction large enough to be confident this works. Decades of practice isn't the same as confirmed benefit once better-designed trials exist and don't back it up.
I'd also name the real cost directly: steroids can mask the early signs of perforation or superimposed infection, which matters a great deal in an injury that's still evolving over these first several days.
I don't think this needs to be settled as all-or-nothing. If we do add steroids, I want it explicit that we're accepting a real monitoring cost, not a free addition to standard care — more frequent exams, a lower threshold for repeat imaging if anything changes, and a specific plan for what would make us stop them rather than continue by default. Given how genuinely unsettled the stricture-benefit evidence is, I'd lean toward holding steroids and focusing resources on that closer monitoring instead, but I'd support either choice if it comes with a real plan attached.
Agreed: standard supportive care with PPI and empiric antibiotics, NPO status maintained with a structured plan for advancing his diet as healing is confirmed, and serial exams every 8 hours for the first 72 hours given the injury's depth.
Not agreed, and left explicitly unresolved rather than papered over: whether the corticosteroid decision was the right one. The gastroenterologist's view that a plausible, decades-practiced intervention shouldn't be withheld from a 38-year-old facing real stricture risk was heard and not overruled — the group chose to withhold steroids based on the more recent randomized evidence, but agreed to revisit the decision explicitly if his course changes.