Taiho and Cullinan Stop Lung Cancer Trial Early — Data Too Good to Stay Blinded

One dose of LSD just cleared a Phase 3 anxiety trial — and the stock gave back a third of its pop before lunch. Both of those facts are the story.

Meanwhile, Taiho and Cullinan stopped a lung cancer trial early because the data was too good to stay blinded, and Karyopharm has a loan payment that pushes it into default the moment it makes it. Let's get into it.

Welcome to The Pharma Closeout for Thursday, August thirteenth, twenty twenty-six. I'm Alex Mercer.

And I'm Maya Patel.

Definium's DT120 — pharmaceutical-grade LSD, single dose — beat placebo by five point four points in Phase 3 generalized anxiety. Jefferies' Andrew Tsai says the street was modeling five. Stifel's Paul Matteis called it a clean win. That is a company with two late-stage successes in three months and a molecule nobody in this industry took seriously five years ago.

The separation isn't what I keep going back to. It's the shape of the curve.

Meaning what — the timing?

Meaning patients moved by day two and held through the core study period. In generalized anxiety, onset was never the risk. The risk is the placebo arm walking up to meet you around week eight — that's where a lot of anxiety programs have quietly died, and it's why the field stopped attracting capital. It didn't happen here. Twelve weeks out from one dose, the drug arm was still eleven point six points down on a fifty-six point scale. That's the finding.

Which is why the models look the way they do. Evercore ISI's Gavin Clark-Gartner carries two billion in peak annual sales for anxiety alone, with depression and PTSD stacked behind it. And that's before you get to label expansion — Matteis is explicitly framing what comes next as a catalyst path, not a launch.

Here's where I'd push, though. Everyone is modeling this as a pill. It is not a pill in any commercial sense that matters.

Go ahead.

Patients are monitored for up to eight hours. The average time to clear the end-of-session checklist was six point four hours. That's not a prescription. That's a scheduled clinical encounter — a room, a clinician, most of a working day. You don't dispense that. You build capacity for it.

That's fair, and the tape agreed with you. Stock opened up fifteen percent and had given most of that back by mid-morning. That fade isn't doubt about the efficacy. That's the market reading the delivery model in real time and finding it heavier than the headline.

Then the honest read is that the ceiling on this drug was never the label. It's chair-hours.

Say that again.

Chair-hours. You can win every endpoint in psychiatry and still be capacity-constrained on day one, because the binding constraint is how many monitored sessions the system can physically run in a year. Whoever solves site economics captures more of this market than whoever has the best separation number.

So the competitive read, if you're running commercial strategy in CNS: psychedelics just stopped being a science question and became a site-infrastructure question. The differentiated asset in this class two years from now isn't the one with the biggest delta — it's the one that gets a patient in and out in four hours. That's the spec sheet every program behind this one now gets designed against.

And whether any of it holds gets decided in September. We flagged the second anxiety trial yesterday as reading out next month — it hasn't landed, and that window is still open. Definium's CEO Rob Barrow is out calling this efficacy unprecedented. One trial doesn't earn that word. Two, with the same durability, and the standard of care conversation in anxiety genuinely changes.

Shifting to the deal and pipeline side — and to a trial that got stopped because it was working too well. Taiho Oncology and Cullinan Therapeutics said today zipalertinib hit in first-line EGFR exon twenty insertion lung cancer. Investigators halted at an early data check. William Blair's Matt Phipps notes the study was powered for a forty percent reduction in the relative risk of progression or death at final analysis — so an interim stop implies they saw more than that.

That's the number that reprices the field. But I'd hold the comparison loosely.

Because of the design?

Because of the design. Phipps puts Zegfrovy's monotherapy benefit at a thirty-five percent risk reduction against chemo. Zipalertinib is a chemo combination. Cross-trial, and a completely different tolerability burden on the patient. Cantor's Li Watsek is right that safety is where this actually gets decided — because in a mutation this rare, physicians choose on side effects once the efficacy is close enough.

J&J's Rybrevant and AstraZeneca and Dizal's Zegfrovy are both already on the market here, and the FDA is reviewing zipalertinib in second line with a decision expected by February twenty-seventh. So Taiho gets a shot at the front line before the second-line fight has even been settled.

And ArriVent felt it inside an hour — shares dipped as much as seven percent, with their own pivotal firmonertinib data due in the next few months. That readout is now walking into a bar that moved this morning, on a trial they designed two years ago.

Staying with the pipeline, but on the other end of the balance sheet. Karyopharm — we flagged the cash position Monday — reported Thursday, and this is worse than a runway story. Sixty-five point four million in cash, and a fifteen point eight million loan payment due September tenth. Underneath it, a ten million dollar minimum liquidity covenant. Making the payment is what triggers the default.

Wait. Paying the loan is the breach?

Paying the loan is the breach. Which is why the CEO is negotiating with lenders rather than writing the check, and why the company is openly evaluating financing and strategic alternatives. Xpovio, meanwhile, is doing what it's done for four years — thirty point eight million in the quarter, essentially flat against a year ago. There's no commercial rescue in that number.

Then the covenant isn't the thing to watch. The August filing is.

Make the case.

They're submitting a supplemental application this month for selinexor plus ruxolitinib in myelofibrosis, accelerated approval, built on the Phase 3 SENTRY data. It rests on spleen volume reduction as the surrogate, and the FDA has signaled that endpoint is reasonably likely to predict survival in this setting. That signal is the entire reason an accelerated path exists here — and it is a company asset that does not appear anywhere on the balance sheet.

So the asset is worth more than the equity.

The asset is almost certainly worth more than the equity, and every lender in that negotiation knows it. Which means this was never a conversation about fifteen point eight million dollars. It's a conversation about who controls a myelofibrosis filing four weeks from now — and distressed lenders don't usually take the sentimental view of that question.

On the regulatory side — we led yesterday with Lilly filing six lawsuits against sellers pushing counterfeit retatrutide. The filings aren't the story anymore. The enforcement math is. Lilly says it has referred more than two hundred people and entities to regulators, licensing boards and law enforcement, and flagged more than fourteen thousand websites, ads and listings across more than a hundred countries. Six lawsuits against that is a signal, not a remedy.

And it's a signal fired into a policy environment moving the other way. In July an FDA advisory panel endorsed broader use of six peptides without substantial evidence of safety and effectiveness. So Lilly is asking courts to close a gray market at the same moment federal advisors are widening the door that market walks through.

Which should worry medical affairs more than legal.

Why medical affairs?

Because retatrutide isn't approved anywhere on earth, and Lilly doesn't plan to file until next year. The defendants include operations selling material labeled research use only. So the safety database on this molecule is being written right now — outside the trial, by people Lilly cannot monitor, at doses nobody has characterized.

And none of that disappears when the real launch comes.

None of it. Every adverse event attributed to something called retatrutide over the next eighteen months attaches to this molecule's reputation whether or not Lilly made the vial. That's why the litigation is aggressive and early. The company isn't defending a market. It's defending a label it hasn't written yet — and you only get to write it once.

Looking ahead — September tenth is the hard date on Karyopharm. Payment, waiver, or default, resolved inside four weeks. Definium's second anxiety readout is the September event that either confirms this morning or reopens all of it. And February twenty-seventh is the FDA decision on zipalertinib in second line.

The nearer one is ArriVent. Firmonertinib pivotal data in the next few months, reading out into a landscape that changed before their trial finished — and Watsek's caution is the right one. It's premature to size the threat until we see the safety profiles side by side.

Three catalysts, three different flavors of risk — capital, replication, competition. If you only have room to track one this quarter, take the exon twenty field. That's where a standard of care gets rewritten in public, with three companies watching each other do it.

The thing I'll be sitting with tonight is chair-hours. If September holds, the hard job in psychiatry stops being drug development and becomes operations — and almost nobody in this industry is staffed for that yet. Have a good evening.

And that is your Pharma Closeout for Thursday, August thirteenth — Definium clearing Phase 3 in anxiety, Taiho and Cullinan halting a lung cancer trial early because it was working, and Karyopharm cornered by its own loan agreement. Follow the show on Spotify, Apple Podcasts, or wherever you listen — tomorrow's briefing lands on its own.