Steroid-Refractory Checkpoint Hepatitis: Treating the Liver Without Losing the Cancer Response
A patient with a real, substantial tumor response to checkpoint inhibitor therapy develops steroid-refractory hepatitis from the same drug. Escalating immunosuppression is not controversial on its own. What happens to her cancer treatment in the same conversation is.
Yolanda R., a 58-year-old woman, ran a neighborhood bakery for over twenty years before selling it last year, not long after her melanoma diagnosis, to focus on treatment while her son took over day-to-day operations. She started pembrolizumab five months ago for metastatic melanoma, with a striking early response — her most recent imaging showed significant tumor shrinkage, a result her oncologist described as better than most patients see at this point in treatment. Three weeks ago her routine labs picked up rising transaminases, ALT climbing to six times the upper limit of normal, consistent with immune checkpoint inhibitor-induced hepatitis; pembrolizumab was held and she was started on high-dose prednisone.
Ten days into adequate-dose steroids, her ALT has not meaningfully improved, meeting the threshold the ASCO immune-related adverse event guideline uses to define steroid-refractory checkpoint hepatitis and to trigger second-line immunosuppression, typically mycophenolate — a recommendation that rests on case series rather than any randomized comparison. That escalation decision doesn't happen in isolation from the reason she's on this drug at all: her tumor response has been genuinely substantial, and whether — or when — pembrolizumab could ever be resumed is a real, separate question sitting underneath the more immediate one about how aggressively to treat her liver.
Escalating immunosuppression without losing sight of why she's on it
Ten days at adequate steroid dose with no meaningful ALT improvement meets the threshold for second-line immunosuppression by established criteria for this exact condition. I'd start mycophenolate now. Continued high-grade transaminitis carries real risk of progressing to more severe, potentially irreversible injury the longer it goes unaddressed.
I'm not opposed to escalation, but I want to confirm something basic first: pembrolizumab has been held, but has anyone actually confirmed how much drug is still circulating and immunologically active at this point? We're talking about adding a second immunosuppressant on top of steroids while the actual trigger for her hepatitis may still be exerting effect. Treating the consequence more aggressively without confirming the cause has actually stopped feels like it's skipping a step.
Holding the drug isn't the same as confirming its immunologic effect has cleared, and pembrolizumab's own pharmacodynamic effects are known to outlast its measurable serum levels by a meaningful margin — which is exactly why I don't think 'held' settles the sequencing question on its own.
Both of you are right about the hepatic side of this, and I want to name the part neither position is addressing directly: her tumor response has been genuinely exceptional, and permanently discontinuing pembrolizumab is not a cost-free safety decision just because it happens alongside treating an irAE. Escalate to mycophenolate now — that part isn't actually in dispute between you — but I'd want oncology and hepatology to have an explicit, separate conversation, once her liver stabilizes, about whether any future rechallenge is on the table, rather than letting this admission's urgency default her permanently off a drug that was working better than almost anything else available to her.
Mycophenolate started alongside continued prednisone; pembrolizumab remains held.
Explicitly not decided today, and documented as such rather than defaulted: whether pembrolizumab could ever be safely resumed once her liver recovers. Oncology and hepatology agreed to revisit that question jointly once her hepatitis is controlled, rather than letting the current admission's urgency settle it by default.