Steroid-Refractory Colitis, a Heart That Doesn't Tolerate Anti-TNF Well
Infliximab is the guideline's first-line answer to his steroid-refractory colitis, and his own heart is the guideline's own stated exception to it. The disagreement isn't which drug works faster — it's whether faster is worth what it might cost a heart that's already working with less reserve than it used to.
Arthur P., 66, has run in more than forty marathons since his forties, and had been quietly training toward a Boston-qualifying time this spring before the diarrhea started — three weeks ago now, and bad enough these past several days that he hasn't managed a training run at all, let alone the tempo miles his plan called for this week. He has metastatic melanoma on combination nivolumab and ipilimumab, twelve weeks in, with an early scan already showing real tumor shrinkage he was genuinely encouraged by before this started. He'd told his oncologist at that scan visit that he felt like himself again for the first time since his diagnosis — a comment his whole team remembers now, watching him instead describe how many times he's had to change his sheets overnight.
Colonoscopy confirmed diffuse colonic ulceration consistent with immune-related colitis, and five days of high-dose IV methylprednisolone at 2mg/kg have not moved him off eight to ten bloody stools a day — steroid-refractory by any reasonable definition, and the point at which NCCN and ASCO's own immunotoxicity guidelines call for escalation to a biologic, with infliximab named first. What complicates that otherwise straightforward next step is his heart: a reduced ejection fraction of 35% from a prior anthracycline exposure, NYHA class II on stable guideline-directed therapy. Infliximab's label restriction here is a dose ceiling rather than a ban: doses above 5mg/kg are contraindicated in moderate-to-severe heart failure, grounded directly in the ATTACH trial, where 10mg/kg produced higher mortality and heart-failure hospitalization than either 5mg/kg or placebo in patients with class III–IV disease and an ejection fraction at or below 35%. Two facts about him cut in opposite directions against that. His ejection fraction of 35% sits exactly at the trial's own enrolment ceiling, so on that axis he is inside the studied population; his NYHA class II puts him one step below it, and the label concedes outright that the drug has never been studied in mild heart failure. The question is therefore not whether the warning applies to him wholesale but whether a 5mg/kg dose — under the ceiling, and the dose ATTACH found no mortality signal for — is the version of it that matters. His cardiologist has followed him closely since his heart failure diagnosis three years ago, adjusting his regimen twice as his ejection fraction gradually improved from an initial 28%, and is the one voice in the room who has watched his physiology pushed toward decompensation firsthand.
On the ward, choosing the escalation drug
Infliximab is the guideline's first-line answer here, and for good reason — it's the fastest, best-documented option for steroid-refractory immune colitis, typically working within days. Eight to ten bloody stools a day after five days of high-dose steroids is a patient losing real volume and electrolytes right now. This isn't the moment for the slower option.
I understand the urgency, and I'm not disputing infliximab's speed. But his heart failure isn't incidental to this decision. ATTACH found the mortality signal at 10mg/kg — eight deaths at a year against four each on 5mg/kg and placebo — in class III–IV patients with an ejection fraction at or below 35%. I'll grant you the label's contraindication is a ceiling, above 5mg/kg, not a ban, and that a 5mg/kg dose sits under it. What I won't grant is that this makes him a studied patient. His ejection fraction is 35%, right at ATTACH's enrolment ceiling; his NYHA class is II, which the label says outright was never studied. That's not the same as "safe."
Speed matters, but a drug with a labeled cardiac warning in a patient whose heart is already compromised isn't a decision to make on speed alone.
There's a separate reason pointing the same direction. Abu-Sbeih et al., 2018, found vedolizumab effective in checkpoint-inhibitor colitis specifically, and its gut-restricted mechanism avoids the still-debated concern that systemic TNF blockade could blunt the same immune activation his checkpoint therapy is using against his melanoma — a real early responder we don't want to work against.
I won't pretend vedolizumab's slower onset is free, given how much he's already losing. Start it now, but with an explicit reassessment at day three — if he isn't clearly improving by then, we add infliximab anyway. His NYHA class is a real caution, not an absolute contraindication, and undertreated colitis is its own serious harm if we wait too long to act on that.
Agreed: start vedolizumab today, continue methylprednisolone at the current dose, correct electrolytes, and reassess explicitly at day 3 — if he isn't clearly trending toward fewer, less bloody stools by then, add infliximab at 5mg/kg and no higher, since undertreated colitis carries its own serious risk and that dose sits under the labeled ceiling.
Not agreed: the gastroenterologist felt a three-day window was too long to accept for a patient already this symptomatic, and would have started infliximab today with close cardiac monitoring rather than wait to see if vedolizumab worked first. The cardiologist held that a dose under the labeled ceiling still doesn't make him a patient ATTACH studied, and preferred giving the gut-selective option a real chance to work before accepting that risk. Neither position was overruled; the day-3 reassessment plan proceeded as written.