Locally Advanced Basal Cell Carcinoma: When a Working Drug Becomes Unlivable
A single patient fourteen months into a drug that is visibly working and increasingly impossible to keep taking. The disagreement isn't about whether the tumor is responding — everyone agrees it is — it's about whether a responding tumor is reason enough to keep enduring what the drug costs to get there.
W.B., a 74-year-old man, has spent most of his retirement in the garage building furniture from reclaimed barn wood, work that requires steady hands and long hours standing — both of which have become harder over the past several months. A large basal cell carcinoma on his scalp, eroding toward the outer table of his skull and deemed unresectable without extensive reconstruction, was started on vismodegib fourteen months ago after a multidisciplinary review; the tumor has shown a genuine partial response, shrinking measurably on serial imaging and no longer threatening bony invasion the way it did at diagnosis.
What the imaging doesn't show is what fourteen months of vismodegib has cost him day to day: severe, near-continuous muscle cramping in his hands and calves, a persistent metallic taste that has driven a nine-pound unintentional weight loss because food no longer tastes like anything worth finishing, and a fatigue he describes as different from ordinary tiredness. ERIVANCE, the pivotal vismodegib trial, recorded adverse-event discontinuation in about twenty-two percent by its final thirty-month analysis; real-world cohorts run far higher, with Herms' French series finding seventy-four percent of those who stopped did so for side effects rather than progression. His own trajectory is not an outlier — it looks more like what happens outside a trial than inside one. His creatine kinase, checked given the severity of his cramping, came back normal, which is itself a real, useful reading: it argues against a rare but serious muscle-breakdown complication and toward the cramping being the drug's expected, if unusually severe, on-target hedgehog-pathway effect on skeletal muscle rather than something requiring a different workup entirely.
Cemiplimab, an anti-PD-1 antibody, is approved for advanced basal cell carcinoma in patients previously treated with a hedgehog pathway inhibitor — and Study 1620, the trial behind that approval, enrolled intolerance of prior hedgehog therapy as its own route in, alongside progression and non-response at nine months. He has been treated with one and cannot tolerate it, which places him inside both the indication and the studied population without his tumor having to fail first. His tumor hasn't progressed. It has responded. The question in front of the team is whether that response is reason enough to keep him on a drug he is, by any reasonable read of his own words, no longer able to tolerate.
His diabetes changes what monitoring means here rather than which drug he gets. On a hedgehog inhibitor his glucose is his endocrinologist's business and nobody else's; on checkpoint blockade, new hyperglycemia becomes one of the ways an immune-mediated endocrinopathy first announces itself — and a baseline condition that already makes new symptoms hard to attribute is precisely what blunts that signal.
A responding tumor and an unlivable drug
The tumor is responding. Fourteen months in, it's smaller than it was and it's no longer threatening the skull. I don't think we should walk away from a mechanism that's demonstrably working before trying to make it more livable first — pulsed or intermittent vismodegib dosing was tested directly in MIKIE, which maintained tumor response across both intermittent schedules it studied. That seems like the first thing to try, not the last.
The indication covers patients previously treated with a hedgehog inhibitor, which he has been — and Study 1620 enrolled intolerance as its own eligibility route alongside progression and non-response at nine months. That third route wasn't a technicality. It was written because the agency reviewing this evidence recognized that severe, unrelenting toxicity in a responding tumor is a real reason to stop, on its own, without needing the tumor to also fail.
I hear the argument for trying intermittent dosing first, and in a patient with milder symptoms I might agree. But he's lost nine pounds because food doesn't taste like anything to him anymore, and his hands cramp badly enough that the woodworking that structures his entire retirement is becoming difficult. That's not a dosing-schedule problem to titrate around. That's a drug he's told us, plainly, he can't keep taking.
I'd have leaned toward trying a lower dose or an intermittent schedule first if his toxicity were moderate. At the severity he's actually describing — near-continuous cramping, meaningful unintentional weight loss, fatigue distinct from his baseline — I don't think asking him to titrate through more months of a drug this size of the FDA indication exists specifically to let him stop taking is the right call.
Switch to cemiplimab. His tumor's response to hedgehog inhibition tells us the pathway matters, but it doesn't guarantee anything one way or the other about his response to checkpoint blockade — that's a real unknown we're accepting, not a certainty we're trading away.
Agreed: discontinue vismodegib and start cemiplimab, on the basis of documented intolerance rather than tumor progression, with close monitoring of whether the response achieved on hedgehog inhibition holds during the transition.
Not agreed: the Medical Oncologist would have preferred a one-month trial of reduced-dose vismodegib before switching classes entirely, still concerned that checkpoint blockade's efficacy against a hedgehog-driven tumor in this specific patient is unproven; the group proceeded with the switch given the severity of his current symptoms, with an explicit plan to revisit hedgehog inhibition at a lower dose only if cemiplimab fails to maintain control.