Eliza Ward: Hey — good to be here.
Brian Reed: Yeah, likewise. You saw the FDA news this morning?
Eliza Ward: That's actually — yeah, that's exactly where I want to start. August 19th. Ultragenyx Pharmaceutical just got accelerated approval for a gene therapy called Genglycos. And the thing that matters here: this is the first FDA-approved treatment of any kind, ever, for GSD-Ia.
Brian Reed: First ever. For a disease that's been managed with, what — cornstarch?
Eliza Ward: Cornstarch. Like, literally raw cornstarch, multiple times a day. And the GlucoGene study — the Phase 3 trial, 46 participants, 48 weeks, randomized double-blind — what it showed was a 31% greater reduction in daily cornstarch intake versus placebo. That's the number the FDA approved on.
Brian Reed: Hang on — 31% reduction in cornstarch intake is the pivotal measure? Not, like, fewer hypoglycemic episodes or something more... clinical?
Eliza Ward: Right — and that's the whole tension here. The FDA called it a surrogate endpoint. Karim Mikhail, who's the acting director of CBER at the FDA, called it a 'great milestone,' but the approval is accelerated, which means confirmatory data still has to come.
Brian Reed: So the milestone is real, but what it proves is... still being worked out.
Eliza Ward: Still being worked out — which is actually the thing that makes me want to back up a step, because why is cornstarch the entire disease management in the first place? Like, why has nothing pharmaceutical ever touched this before?
Brian Reed: Right, so — okay, the simplest version: your body can store glucose, it just can't release it. Think of it like a car with a fuel tank that has no output valve. The engine runs out, you stall. So patients have to manually pour fuel in — every three, four hours — or they go hypoglycemic. That's it. That's the whole disease.
Eliza Ward: The G6PC gene. Glucose-6-phosphatase. The enzyme that opens the valve is just — absent.
Brian Reed: Gone, yeah. And cornstarch works because it digests slowly — it's basically a timed drip. But picture a parent setting a three a.m. alarm to wake their teenager, measure out cornstarch, mix it with water, watch them drink it, go back to sleep. Every night. For the rest of that child's life.
Eliza Ward: Wait — every night, indefinitely.
Brian Reed: Emil Kakkis — CEO of Ultragenyx — he's described this regimen as one of the most burdensome in all of rare disease. And David Weinstein, who's spent his career researching GSD-Ia, has said essentially the same thing: the management burden is relentless in a way that most chronic disease management just isn't.
Eliza Ward: And there are roughly fifteen hundred to twenty-five hundred patients in the U.S. carrying all of that. So — okay, I want to name the obvious question: if the disease is this brutal, why is it only now that something worked? Janice Chou at the NIH had a program on this. Dimension Therapeutics had a program. Decades of infrastructure.
Brian Reed: That's actually — I mean, the NIH work goes back years, and the fact that it took Dimension Therapeutics and then Ultragenyx inheriting that program to get here tells you something. The liver is hard to target, AAV8 delivery into hepatocytes is genuinely complicated, and the safety bar for a one-time gene therapy in kids is enormous. It wasn't that no one was trying.
Eliza Ward: No, it's that the biology was intractable long enough that cornstarch — uncooked cornstarch, every few hours — was the best available answer for decades. That's the part the headline undersells.
Brian Reed: But here's where the coverage is getting this slightly wrong — the part I keep seeing is people reading the approval as proof that Genglycos prevents hypoglycemic crises. And I don't think that's what the FDA actually said.
Eliza Ward: It's not. That's the thing. The FDA accepted cornstarch reduction as a surrogate endpoint — their language is 'reasonably likely to predict clinical benefit.' That phrase is doing enormous work and I think most of the coverage is just... skipping it.
Brian Reed: But doesn't 'reasonably likely' get you most of the way there? Like, if you need less cornstarch, you're probably crashing less.
Eliza Ward: Probably, yeah — but that's inference, not data. The surrogate has never been validated as a proven predictor of fewer seizures or serious liver complications. One fewer cornstarch dose per day is what the GlucoGene trial actually measured. Whether that translates to fewer hospitalizations — that's what the confirmatory studies are for.
Brian Reed: And those studies are... two more years minimum.
Eliza Ward: FDA's floor is two additional years. And then there's Megha Kaushal — Acting Deputy Director of FDA's CBER Office — she's quoted saying the therapy 'targets the root cause of disease.' Eric Crombez at Ultragenyx uses almost identical language. And wait — I get why they say it, mechanistically it's accurate, the G6PC gene is the root cause — but patients hearing 'root cause' are going to hear 'cured.' That's not what an accelerated approval on a surrogate endpoint means.
Brian Reed: So the regulatory language and the patient-level takeaway are just — they're not the same sentence.
Eliza Ward: Not even close. And look, the post-marketing confirmatory trial is the only bridge between those two things, and that bridge hasn't been crossed. Which — actually, there's a related number that makes this feel even less settled, and we'll get to it in a minute: Ultragenyx is monitoring for ten years, the FDA mandated two, and the gap between those two timelines says something about what nobody's saying out loud yet.
Brian Reed: Right — and if the confirmatory data at year two shows the surrogate didn't hold, what actually happens? Does the approval get pulled? Do patients who are already on this stop?
Eliza Ward: That's the question nobody has a clean answer to — and the ten-year monitoring commitment is what makes me think Ultragenyx doesn't either. The FDA floor is two years. Ultragenyx is voluntarily running this for a decade, with untreated controls in the design. You don't do that if you're confident about durability.
Brian Reed: Wait — untreated controls? In a long-term trial for a disease this severe?
Eliza Ward: Yeah. That's — I mean, that raises its own ethical questions given how documented the disease burden is. But the structural point is: the inclusion of controls tells you they're not treating this as a solved problem. AAV8 gene expression in the liver — we just don't know how long it holds.
Brian Reed: And if it fades — if G6PC expression drops off at year four, year six — the patient is back to cornstarch. After a one-time infusion at, what, an implied price somewhere near that $362 million peak-sales analyst figure. No confirmed list price, but that math is not pointing toward affordable.
Eliza Ward: Right, and no confirmed price is actually the confirmed fact here. The $362 million is a peak-sales projection — that's speculation, not a number Ultragenyx has published.
Brian Reed: The part I don't get is how this even reaches patients logistically. Genglycos only goes through Qualified Treatment Centers. Think about what that means for a family in rural Montana whose kid has GSD-Ia — they're already traveling for every specialist visit. Now they need a QTC? The access cliff isn't hypothetical.
Eliza Ward: And some patients can't even get to that cliff — because of the AAV8 antibody exclusion. Prior exposure to AAV8 screens you out entirely. We don't have a proportion. Could be a small slice of those fifteen hundred to twenty-five hundred U.S. patients, could be significant.
Brian Reed: Ultragenyx has the UltraCare program — Gene Therapy Guides for insurance navigation — but sources don't actually confirm how many patients that realistically reaches. And then there's the Priority Review Voucher. Ultragenyx got one at approval. Valuable tradeable asset for the company. Does nothing for the family trying to get reimbursed.
Eliza Ward: That's the concrete thing to watch: whether payers cover this before the confirmatory data lands, or whether they wait for year two. Because if coverage decisions trail the durability evidence — that's where access actually closes.
Brian Reed: Two years. That's the factual horizon. Confirmatory trial reports back, and either cornstarch reduction tracks with fewer hypoglycemic emergencies — or it doesn't. And I genuinely don't know which way that goes.
Eliza Ward: Neither does the FDA, which is — I mean, that's what accelerated approval actually means. The approval happened. Whether the surrogate holds is still a live question.
Brian Reed: And if it does hold — if year two shows the 31% cornstarch reduction actually predicts fewer crises — that matters way past GSD-Ia. That becomes a template for every rare disease where a direct endpoint would take a decade to measure.
Eliza Ward: Right. And if it doesn't — if the surrogate doesn't validate — then the harder question is whether accelerated approval gave families real hope before the therapy could actually deliver on it. That's the reckoning nobody wants to name yet.
Brian Reed: Two years to know which story this is.
