Edaravone for ALS: A Narrow Trial Population Against a Broader Real-World Request
A single patient, well outside the population edaravone's pivotal trial actually enrolled. The disagreement isn't about whether the drug works in early ALS — it's about what that finding means once a wheelchair-bound patient two years past that window asks for it anyway.
L.F., a 61-year-old retired librarian, has used a wheelchair for the better part of a year now, ever since her legs stopped reliably carrying her weight across a room. ALS was diagnosed three years ago, beginning as a dragging left foot she initially blamed on a bad hip, and it has moved through her steadily since — first the legs, more recently her breathing, with a forced vital capacity that has fallen from 78% predicted at diagnosis to 55% today. She lives with her sister, who has become, without either of them quite deciding it that way, her full-time caregiver. Her ALSFRS-R today totals 22 of 48, and reading the individual items rather than just the sum matters here: she scores a zero on walking and a zero on climbing stairs, meaning those functions are entirely gone, not merely reduced.
That distinction is exactly what separates her from the patients edaravone's pivotal trial actually studied. MCI186-19 enrolled only ambulatory, early-stage ALS patients — disease duration under two years, forced vital capacity at least 80% predicted, and critically, an ALSFRS-R score of at least 2 on every individual item, meaning no domain of function had yet been fully lost. In that narrow population, edaravone slowed ALSFRS-R decline by roughly a third over 24 weeks compared with placebo, the finding that led to its 2017 approval. L.F. would not have qualified for that trial on any of its three defining criteria: her disease is a year past its duration window, her FVC is well under the threshold, and two of her functional domains are already at zero. She has read about the drug in an ALS support-group forum and wants to know why that should matter.
A drug she asked for, outside the trial that approved it
I think we offer it. She is asking, she understands her prognosis better than most patients I see, and "the trial didn't study people like you" isn't the same as "the drug definitely doesn't work for you." Postmarketing series that include patients well outside MCI186-19's window have reported edaravone use with acceptable tolerability, and some patients report a perceived slowing they attribute to the drug — soft evidence, I'll grant that, but she has limited options and a disease that isn't going to wait for a better-designed trial to enroll patients like her.
I understand the pull of that, and I'm not questioning her judgment or her right to weigh in. But I want to be precise about what MCI186-19 actually showed, because it matters specifically for her: the roughly one-third slowing of ALSFRS-R decline came from a population selected to still have every functional domain at least partially intact — ambulatory, breathing well, under two years from diagnosis. Broader unselected-population trials run after the original approval did not replicate a significant benefit. That's the difference between "this drug modifies early disease trajectory" and "this drug does something for advanced disease," which nobody has actually shown.
Calling the postmarketing reports real-world evidence of benefit overstates what an uncontrolled series can tell you — perceived slowing in an inexorably progressive disease, without a comparator, is exactly the kind of signal that regresses to nothing under controlled testing, which is why MCI186-19 was run as a randomized trial in the first place. Twice-daily infusion cycles indefinitely is a real burden to ask a wheelchair-bound patient and her caregiver to take on for a benefit nobody has demonstrated outside her narrow trial population.
There's a version of this that respects both of you. The oral suspension formulation, approved in 2022, removes the infusion burden entirely — same drug, same unproven-outside-trial-population caveat, but without asking her sister to manage a ten-day IV cycle at home. I'd offer it with an honest framing: we don't have evidence this slows her disease the way it slowed MCI186-19's earlier-stage patients, and we're offering it because she's asking and the risk profile is low, not because the data support it in her stage.
I'd set a real functional checkpoint at three months — if her FVC and ALSFRS-R decline continue at the same slope as before starting, that's useful information for her, not proof of failure, but it's the honest number to have in front of her the next time this conversation happens.
Agreed: start oral edaravone, framed honestly to L.F. and her sister as unproven in her disease stage, with a functional recheck (FVC, ALSFRS-R) at three months.
Not agreed: whether offering an expensive therapy outside its evidence base is the right default once a patient asks, at a systemic level. The Clinical Pharmacologist's discomfort with that precedent wasn't resolved by today's decision, only outweighed in this instance by the low procedural risk of the oral formulation and L.F.'s own clearly stated preference.