Therapeutic Thursday: The Field Is Sorting Itself. Watch Who Is Leaving and Why.

Johnson & Johnson discontinued prizlo-cel in May. The drug worked. That was not enough.

That distinction sits at the centre of where cell and gene therapy is right now. Legend Biotech's Carvykti has made the company profitable for the first time. Lilly has spent $3.25bn on an in vivo platform with clinical data in four patients. A nonprofit consortium has built infrastructure specifically to rescue therapies pharma has abandoned. These things are happening in the same field, in the same quarter.

The clinical proof is real. The commercial model is sorting itself.

Today's edition maps both.

Lilly spent $3.25bn on Kelonia in April, four months after buying Orna in February. BMS acquired Orbital in October 2025. Chinese biotechs, Byterna, Starna, Vivacta, and DeliNova, raised venture capital on in vivo platforms with no clinical data. The field has picked its next architecture, and it is not the one that got it here.

The structural problem with autologous CAR-T has always been visible: collect cells from a patient who is actively sick, manufacture a bespoke therapy, ship it back within a window often measured in weeks. Manufacturing fails. Patients deteriorate before infusion. The clinical results in haematological malignancies are genuinely impressive. The commercial model underneath them is under constant pressure. Allogeneic CAR-T was supposed to solve this. It hasn't. No allogeneic product has reached approval, and Legend Biotech's CEO is direct about its own allogeneic programme alongside Carvykti: years away from commercial readiness.

In vivo removes the problem entirely. Gene-editing instructions go directly into the body, programming the patient's own T cells without cells ever leaving. Kelonia presented data at ASH in December on four patients in multiple myeloma, early and small, but striking enough that Lilly paid $3.25bn four months later. That price is not a valuation of the data. It is a valuation of what in vivo solves if the data holds.

THE CHATTER

The Chinese biotech raises are the more revealing signal. When capital moves ahead of any human evidence at all, the field has already made its decision. People close to the deal space note that autologous assets are being repriced in real time relative to in vivo platforms, quietly and without announcement.

WHY YOU SHOULD CARE: If you are in a CAR-T programme or evaluating one, the platform question is no longer theoretical. A programme that looked well-positioned twelve months ago may now be competing for partnership interest against platforms without IND filings. The window to position around in vivo, through partnership, acquisition, or building the clinical argument for your own platform's durability, is open and will not stay open.

LOOK OUT FOR

Lilly's first in vivo CAR-T IND from the Orna platform. ORN252, targeting CD19 autoimmune disease, was expected to reach the clinic in 2026. The question is not whether it works in four patients. It is whether Kelonia's ASH signals replicate on a different platform.

Legend Biotech has reported its first profitable quarter on Carvykti. The CEO is now publicly discussing allogeneic and in vivo programmes for oncology and autoimmune. It is the first time a CAR-T company has reached that conversation from a position of commercial strength rather than pipeline hope. Gilead is having the same conversation, framed less optimistically: can anito-cel in multiple myeloma rescue a cell therapy business that has underperformed since the Kite acquisition?

The commercial ceiling on autologous CAR-T is not the clinical results. It is delivery infrastructure. Infusion centres must be certified, staffed, and equipped to manage cytokine release syndrome. Multiple myeloma affects approximately 35,000 new patients annually in the US. The number receiving CAR-T is a fraction of that, not because the drug doesn't work, but because the system cannot keep pace.

The autoimmune opportunity is where the ceiling argument flips. Kyverna's pivotal data in stiff person syndrome points to a market where patients are younger, the disease is chronic rather than imminently terminal, and the value of a single-dose durable remission, replacing years of biologic infusions, is commercially legible to payers in a way that has not existed before for cell therapy. Companies that treat autoimmune as oncology CAR-T in a different indication will get it wrong. One friction compounding the throughput problem: eight years of real-world safety data show the theoretical cancer risk that justified a 15-year post-treatment monitoring requirement has not materialised. The case for reducing it to five years is now being made publicly. If the FDA accepts it, that is an administrative burden removed from treatment centres that has nothing to do with efficacy.

WHY YOU SHOULD CARE: Legend's profitability changes the commercial narrative but not the infrastructure constraint. The autoimmune indication is the more interesting commercial question right now. The payer logic, the patient population, and the reimbursement case are structurally different from oncology. Kyverna's BLA will be the first real test of whether that logic holds under regulatory and commercial scrutiny at the same time.

LOOK OUT FOR

Kyverna's BLA filing for mivocabtagene autoleucel. The 2026 timeline is stated. A clean submission and review would make mivo-cel the first CAR-T approved in autoimmune disease, resetting the commercial and competitive conversation in gMG and stiff person syndrome simultaneously.

The FDA's Office of Tissues and Advanced Therapies chief said it plainly in May: CMC readiness remains the primary roadblock for cell and gene therapy approvals. The same week, the agency published CMC flexibility guidance clarifying that phase-appropriate manufacturing standards are acceptable, that commercial-grade CMC is not expected at Phase I, and that non-traditional analytical methods can be valid if the science supports them. This is CBER's consistent message for three years. The fact that it still needs to be said publicly suggests it is still not being heard early enough.

UniQure's AMT-130 for Huntington's disease is what regulatory misalignment actually costs. The company built its strategy around accelerated approval using real-world evidence, a path apparently agreed with the FDA, then reversed. The reversal required a randomised controlled trial against a sham surgery control arm. UniQure has now filed with the MHRA using the same data package the FDA rejected. The MHRA's revised gene therapy definitions, currently out for consultation, may create flexibility the FDA's current posture does not allow. That the UK is now an explicit regulatory alternative for a US-developed therapy in a serious unmet need indication is not a footnote.

THE CHATTER

The UniQure situation is being watched closely by sponsors whose FDA interactions on accelerated pathways have stalled or produced unexpected reversals. The question being asked quietly: is MHRA approval without FDA alignment a commercially viable outcome, or a consolation route? The answer will depend on the MHRA consultation outcome and on whether any payer infrastructure exists for a product approved only in the UK.

WHY YOU SHOULD CARE: The CMC message from CBER has not changed in three years and is not changing now. The flexibility guidance is useful, but the underlying behaviour it is trying to correct, treating manufacturing as a late-stage problem, remains the norm. If you have a cell or gene therapy programme in development, the guidance gives you cover to take a phase-appropriate approach. It does not give you cover to delay. The UniQure situation is a live reminder that agreed regulatory pathways are not guaranteed ones.

LOOK OUT FOR

The MHRA's gene therapy definition consultation outcome. If the revised definitions create meaningful flexibility for novel modalities, watch how many programmes quietly pivot to the UK pathway in H2 2026. Separately, the FDA's response to the CAR-T follow-up reduction argument: if the agency signals openness to moving from 15 to 5 years, it removes real commercial friction for BLA filings currently in preparation.

J&J dropped prizlo-cel and JNJ-9530 in May. Prizlo-cel had demonstrated strong efficacy in large B-cell lymphoma. The discontinuation was not a clinical failure: the treatment landscape had moved, commercial returns no longer justified the investment, and the company had other priorities. Novo Nordisk offloaded three cell therapy assets since September, systematically, to Heartseed, Aspect, and Cellular Intelligence. Dr Reddy's exited CAR-T entirely, pivoting to GLP-1 generics. Each decision is rational from the perspective of the individual company. The aggregate effect is that patients with diseases those programmes were designed to treat have fewer options than they would if the field's commercial model were different.

Telethon, ASGCT, and OTXL have built a nonprofit coalition specifically to prevent scientifically valid therapies from disappearing because no commercial entity will fund them to approval. They call these programmes "commercially pre-viable": the clinical evidence exists, the biology is sound, the regulatory path is navigable, but the development economics do not currently work for a for-profit company. The coalition's existence tells you something about the volume of exits now occurring.

THE CHATTER

Bayer is the name being mentioned in the context of this sorting wave. The company has publicly declared its intention to raise its cell and gene therapy bet as others exit, arguing the exits create acquisition opportunities priced on commercial scepticism rather than scientific value. The question being asked is whether Bayer has the discipline to be selective. Its next acquisition will answer that.

WHY YOU SHOULD CARE: The exits are creating a buyer's market in programmes with scientific validity but commercial uncertainty. A mid-size biotech with balance sheet and risk appetite can now in-license a commercially pre-viable cell or gene therapy asset, one abandoned for portfolio reasons rather than clinical ones, at a discount that will not persist once the commercial model normalises. The window is the sorting wave itself. It will close.

LOOK OUT FOR

What Bayer actually acquires next. The company has declared its thesis. The next deal will reveal whether that produces disciplined acquisition criteria or opportunistic accumulation. There is a meaningful difference between the two.

The next ninety days will tell you more about where cell and gene therapy is actually going than the last two years of announcements. Kyverna's BLA, Lilly's first in vivo IND, the MHRA consultation outcome, and the FDA's posture on follow-up reduction are not separate stories.

They are the same question being asked four different ways: does the commercial model catch up to the science, or does the sorting wave continue?

See you tomorrow.