What’s worth building in cell therapy?

What’s Worth Building in Cell Therapy?

What does “worth building” mean in cell therapy today?

A cell-therapy startup is worth building today when it can make the treatment much easier to deliver, open a much larger disease market, or restore a biological function that ordinary drugs cannot restore.

That bar has gone up. Cell therapy already works. CAR-T products are generating billions of dollars, TIL therapy has reached the market in solid tumors, regulatory T cells now have FDA validation, and stem-cell-derived replacement cells are producing meaningful results in diabetes and Parkinson’s disease.

The interesting gaps are more practical. Can we program the cells inside the patient instead of manufacturing them for weeks? Can immune reset move from very sick cancer patients into autoimmune diseases? Can replacement cells survive without lifelong immunosuppression? Can an off-the-shelf product perform well enough that doctors stop accepting the logistics of personalized CAR-T?

That is where we would look now.

Is cell therapy still growing, or are the original CAR-T drugs already peaking?

Cell therapy is still growing quickly, but the latest commercial numbers show that growth has shifted toward a smaller group of winning products while several older CAR-T franchises are shrinking.

The contrast is unusually clear in the latest quarterly results. Johnson & Johnson reported $657 million of worldwide Carvykti sales, up 49% from the same quarter a year earlier. Bristol Myers Squibb reported $484 million for Breyanzi, up 41%. Meanwhile, Gilead’s Yescarta fell 12% to $346 million and Tecartus fell 24% to $70 million. Gilead’s overall cell-therapy sales dropped 14% year over year.

A newer CAR-T such as Carvykti can now generate more than $2.5 billion at an annualized quarterly run rate while another major franchise is contracting. Physicians and patients are already moving between products as better clinical positioning, earlier-line use and competition change the treatment choice.

CAR-T product	Latest quarterly sales	Year-over-year change	What we see
Carvykti	$657M	+49%	Still expanding very fast
Breyanzi	$484M	+41%	Gaining strongly across indications
Yescarta	$346M	-12%	Mature franchise facing heavier competition
Tecartus	$70M	-24%	Clear pressure from competing treatments

Would we build another autologous CAR-T for blood cancer?

We would rarely build another conventional autologous CAR-T for blood cancer today unless it changed the patient experience or clinical outcome enough to make existing products look clearly outdated.

CD19 and BCMA are already crowded with approved CAR-Ts, bispecific antibodies and additional programs moving through development. A startup can produce impressive response data here and still end up with a difficult commercial position. Gilead’s current decline in Yescarta and Tecartus sales shows how quickly competition can affect even established products.

Access is also gradually improving for the incumbents. The FDA removed the REMS programs attached to the approved BCMA- and CD19-directed autologous CAR-T therapies in 2025, reducing some of the special certification requirements around treatment.

We would still pay attention to a company that could make an autologous product in a couple of days, avoid severe toxicity, work after previous CAR-T exposure, or produce much better durability. Changing the construct without producing a visible clinical advantage is much less interesting.

If you want more recent data on this point, please see our latest cell therapy market report.

Is in vivo CAR-T now the biggest opportunity in cell therapy?

In vivo CAR-T is currently our strongest cell-therapy startup opportunity because it could remove several of the category’s biggest problems with one change: engineering the patient’s immune cells directly inside the body.

The concept has recently crossed an important line. In a Phase 1 study published in Nature Medicine, EsoBiotec’s ESO-T01 was given intravenously to five heavily pretreated multiple-myeloma patients without leukapheresis, external cell manufacturing or lymphodepleting chemotherapy. Four of the five patients responded, including three stringent complete responses. CAR-T cells were successfully generated inside the patients.

Five patients cannot establish efficacy, and the safety profile was far from trivial. Four patients experienced cytokine release syndrome, including three grade 3 cases, and every patient had at least one grade 3 or higher adverse event. Still, we now have human evidence that a vector can enter the body, program T cells there and produce antitumor activity.

Large pharmaceutical companies have also concentrated a remarkable amount of money around the approach. AbbVie bought Capstan for up to $2.1 billion. Bristol Myers Squibb paid roughly $1.5 billion net of acquired cash for Orbital Therapeutics. AstraZeneca acquired EsoBiotec for up to $1 billion. Gilead’s Kite bought Interius for $350 million.

The attraction is easy to understand. A successful in vivo product could eventually remove leukapheresis, individualized manufacturing and part of the conditioning burden. The remaining risk is whether targeted delivery can become precise and safe enough to make that advantage clinically usable.

In vivo cell-therapy deal	Buyer	Headline value	What was acquired
Capstan Therapeutics	AbbVie	Up to $2.1B	Targeted LNP delivery of CAR-encoding RNA
Orbital Therapeutics	Bristol Myers Squibb	~$1.5B net of acquired cash	RNA medicines including in vivo CAR-T
EsoBiotec	AstraZeneca	Up to $1.0B	Targeted lentiviral in vivo cell engineering
Interius BioTherapeutics	Gilead / Kite	$350M	Targeted vectors for in vivo CAR generation

Can autoimmune CAR-T become a big market while it still needs chemotherapy?

Autoimmune CAR-T can already become a meaningful market in severe disease, but it probably needs to reduce or eliminate lymphodepleting chemotherapy before it reaches mainstream autoimmune care.

The evidence has moved quickly. Cabaletta Bio has now treated patients across myositis, lupus, systemic sclerosis and pemphigus with rese-cel. In its latest myositis update, five of six evaluable dermatomyositis patients reached the response level the company plans to use in registration, with those responses maintained off immunomodulatory drugs for as long as roughly a year and a half. Cabaletta is already moving systemic sclerosis toward a registrational study as well.

Kyverna has gone further down the regulatory path. Its miv-cel program is already in a rolling BLA submission for stiff person syndrome, while a Phase 3 trial is running in generalized myasthenia gravis. In the latest reported Phase 2 myasthenia gravis cohort, all seven treated patients achieved clinically meaningful improvement by 24 weeks, and the company reported durability extending toward one year. Kyverna is also testing the same CAR-T across rheumatoid arthritis, lupus and multiple sclerosis.

The same basic biological idea, deep depletion of pathogenic B cells followed by immune reconstitution, is now producing activity across several autoimmune diseases. That gives the category much more room than a single lupus program.

The limit is treatment burden. Current autologous programs can justify conditioning in severe, treatment-resistant patients, but millions of autoimmune patients have other options. Cabaletta has already treated early patients without preconditioning, Kyverna is exploring alternative or no-preconditioning regimens, and Cartesian’s Descartes-08 is in Phase 3 in myasthenia gravis using repeated outpatient infusions without preconditioning chemotherapy.

In vivo developers are pushing toward the same end point. AbbVie’s Capstan program is designed to generate CD19 CAR-T cells directly in patients without lymphodepletion, while Interius is pursuing the same broad objective with targeted viral vectors.

For severe refractory disease, chemotherapy may remain acceptable for quite a while. For rheumatoid arthritis, lupus and other very large chronic diseases, the biggest opportunity is likely to be an immune-reset therapy that no longer feels like an oncology procedure.

If you want more recent data on this point, please see our latest cell therapy market report.

Is off-the-shelf CAR-T finally starting to work?

Off-the-shelf CAR-T has recently become much harder to dismiss because several programs are producing responses that begin to approach autologous CAR-T while preserving immediate availability.

Caribou’s vispa-cel has the strongest mature dataset we see today. Among 27 optimized second-line large B-cell lymphoma patients, the latest long-term update showed an 82% overall response rate, a 67% complete response rate and median progression-free survival of 17.1 months after a single dose. The company has aligned with the FDA on a roughly 250-patient pivotal Phase 3 trial.

Caribou is also getting interesting results in multiple myeloma. Its CB-011 program produced an 83% complete response rate in the reported BCMA-naive dose-escalation group, with half of the patients who reached complete response or better still there at 15 months.

Allogene is attacking the market from another angle. Its ALPHA3 study is testing cema-cel as first-line consolidation in patients who remain molecularly positive after initial therapy. At an interim analysis, 58.3% of patients receiving cema-cel cleared measurable residual disease compared with 16.7% in the observation arm. Most treatment was handled outpatient, and the FDA has since granted the program both RMAT and Fast Track status.

The real advantage is immediate treatment. That could let CAR-T move into community centers, rapidly progressing patients and earlier disease settings where waiting for personalized manufacturing is a bigger problem.

We still want more durability before calling allogeneic CAR-T solved. CAR-NK and iPSC-derived immune cells belong inside the same opportunity, but their clinical evidence remains less mature than the strongest allogeneic T-cell programs.

Can cell therapy finally break through in solid tumors?

Cell therapy has now proved that it can work commercially in solid tumors, although the winning formats so far also show why another generic solid-tumor CAR-T would be a poor startup thesis.

Iovance gives us the strongest commercial proof. Its TIL therapy Amtagvi generated about $91 million in U.S. revenue in the latest quarter, up 68% from the same period a year earlier. Total company product revenue reached roughly $99 million, and gross margin improved to 56%.

The trade-off is operational complexity. TIL therapy requires tumor tissue collection, centralized manufacturing, lymphodepletion and specialist treatment centers. Iovance has spent heavily on its own manufacturing infrastructure to make the model work.

Engineered TCR therapy offers another useful lesson. Tecelra became the first engineered T-cell therapy approved for a solid tumor and recently received full FDA approval with an expanded age range. Yet the original developer, Adaptimmune, sold Tecelra and several related programs to US WorldMeds for $55 million upfront plus potential milestones after struggling financially. Patients must have the right cancer, express the MAGE-A4 antigen and carry one of a limited set of HLA types.

That history should make founders careful with narrow targeting strategies. A scientifically elegant therapy can end up serving a very small slice of an already uncommon cancer after biomarker filtering.

We would focus on technologies that solve antigen heterogeneity, improve trafficking, survive the tumor microenvironment, recognize several targets, or manufacture cells much faster.

Are regulatory T cells ready to become a real cell-therapy market?

Regulatory T cells are much more investable now because the FDA has finally validated the basic therapeutic idea in a randomized trial.

TREGZI became the first FDA-approved regulatory T-cell-based immunotherapy. In the 187-patient PRECISION-T study, 78% of patients receiving the Treg-containing transplant approach were alive without moderate or severe chronic graft-versus-host disease at one year, compared with 38.4% in the standard-transplant group. Serious chronic GVHD occurred in 12.6% versus 44%.

TREGZI is a specialized transplant product, so we should be careful about extrapolating too far. It combines donor hematopoietic stem cells, regulatory T cells and conventional T cells and is used after myeloablative conditioning in matched-donor transplantation. Still, the approval gives the industry a clean piece of evidence that deliberately manipulating Tregs can produce a large clinical effect.

The bigger startup thesis is engineered tolerance. Companies such as Quell are trying to engineer Tregs that suppress a particular unwanted immune response while leaving the rest of the immune system alone. AstraZeneca partnered with Quell around type 1 diabetes and inflammatory bowel disease in a deal carrying more than $2 billion of potential milestones.

If antigen-specific Tregs work, the applications could stretch from autoimmune disease to transplantation and cell replacement. We could imagine protecting an implanted cell therapy through targeted immune tolerance instead of putting the entire patient on chronic immunosuppressants.

We would call engineered Tregs a high-upside second-wave market today. Regulatory validation has arrived, but the precision-engineered products that make the startup thesis exciting still need much stronger human data.

If you want more recent data on this point, please see our latest cell therapy market report.

Is type 1 diabetes now the clearest cell-replacement opportunity?

Type 1 diabetes is currently the clearest proof that manufactured replacement cells can restore a function that patients have lost.

Vertex’s zimislecel has already produced results that would have sounded extraordinary a few years ago. In the 12 full-dose patients with at least one year of follow-up from its Phase 1/2 study, all 12 avoided severe hypoglycemic events while maintaining HbA1c below 7%, and 10 became insulin independent. Vertex has since continued enrolling and dosing its pivotal Phase 1/2/3 program.

The core biological question is therefore much smaller than it used to be. Stem-cell-derived islets can engraft, sense glucose and release enough insulin to eliminate injected insulin in many treated patients. The catch is that zimislecel currently requires chronic immunosuppression.

The latest pipeline moves show exactly where the industry is heading. Vertex has received FDA clearance to begin testing VX-017, another stem-cell-derived islet therapy intended for a broader type 1 diabetes population. Meanwhile, Sana is building SC451, an iPSC-derived islet product engineered with its hypoimmune technology.

Sana’s supporting human experiment is tiny but unusually interesting. A patient received immune-edited donor islets without immunosuppressive drugs, and the transplanted cells were still alive and producing insulin 14 months later. The dose was not intended to make the patient insulin independent, and a single person cannot prove a product works. It does show that immune-evasive islets can survive in a human body for more than a year without systemic immunosuppression.

For builders, this is one of the best setups in cell therapy: the desired function has already been demonstrated, the unmet need is enormous, and the main remaining obstacle is clear enough to attack directly.

Is immune evasion more valuable than making another replacement cell?

Immune evasion may now be more valuable than inventing another high-quality replacement cell because rejection is becoming the common bottleneck across regenerative medicine.

Type 1 diabetes makes the point neatly. Vertex can manufacture functional pancreatic islets, but its leading product relies on immunosuppressive drugs. Earlier attempts to physically shield islet cells inside devices have struggled to deliver enough functional benefit. Sana’s early human data, meanwhile, suggest that genetically hiding donor cells from immune attack can keep them functioning without systemic immune suppression.

The same problem appears whenever we put foreign living cells into a patient. Dopaminergic neurons, cardiomyocytes, pancreatic cells and other allogeneic tissues can all trigger immune rejection. If every product requires a different protective device or permanent immunosuppression, cell replacement remains confined to relatively severe disease.

A reusable immune-evasion system changes the economics. One well-designed cell line could potentially be differentiated into several therapeutic cell types while carrying the same protection from host immunity. That gives a startup a broader asset than a single disease-specific cell product.

There is a serious safety trade-off. Cells designed to avoid immune surveillance must still be removable if they grow abnormally, mutate or behave unpredictably. We would therefore care almost as much about kill switches and controllability as about the immune-evasion edits themselves.

For now, we would rather own a convincing immune-protection technology that can travel across several replacement-cell products than a slightly better recipe for producing one cell type the industry already knows how to make.

If you want more recent data on this point, please see our latest cell therapy market report.

Is Parkinson’s cell therapy real enough to build a company around?

Parkinson’s cell replacement is now real enough to take seriously because a stem-cell-derived neuron therapy has reached a randomized Phase 3 trial.

BlueRock Therapeutics is testing bemdaneprocel, a pluripotent-stem-cell-derived dopaminergic neuron therapy, in a pivotal study of roughly 102 people with Parkinson’s disease. The trial uses sham surgery and blinded assessment, which is important here because earlier open-label studies were too small to tell us how much improvement truly came from the transplanted neurons.

Earlier Phase 1 work showed that the transplanted cells could survive and produced encouraging motor trends over several years. Moving into a controlled Phase 3 program means the central question is now clinical efficacy rather than whether human dopaminergic neurons can be manufactured and implanted at all.

This market will develop more slowly than diabetes cell replacement. The procedure involves brain surgery, outcomes take longer to measure, and neurological symptoms are harder to interpret than insulin production. A startup may have to fund years of development before learning whether its cells meaningfully change disease.

Still, success here would create something important. Parkinson’s would become a human proof point for replacing neurons lost to neurodegeneration. That could pull more capital toward cell replacement in Huntington’s disease, retinal disorders and other conditions where a specific cell population disappears.

We would build in neural replacement when the company owns a clear advantage in cell quality, delivery, immune protection or a disease where the biological replacement target is unusually well understood. We would be much more cautious with broad “regenerative neurology” platforms that lack a precise cell-loss hypothesis.

Is cell-therapy manufacturing still worth building a startup around?

Cell-therapy manufacturing is still worth building around when the technology removes days, facilities or quality-control steps; generic automation and extra factory capacity look much less attractive.

A 2025 survey of 40 CAR-T manufacturing institutions found that 90% were already using some form of automation. Average vein-to-vein time was still 22 days. Respondents continued to cite equipment cost, regulatory requirements, quality control, facility constraints and specialist training as major problems.

That makes another automated box a difficult pitch. Automation has already entered the workflow, yet the end-to-end process remains slow and expensive.

Recent commercial preparations make the same point from the developer side. Cabaletta signed a 10-year commercial supply agreement with Cellares for rese-cel and has since added ElevateBio as another manufacturing partner. Kyverna recently secured commercial manufacturing capacity with ElevateBio as it prepares for a potential autoimmune CAR-T launch.

A useful manufacturing startup therefore needs to move a much bigger variable. Rapid potency assays that cut release time, truly closed point-of-care manufacturing, automated quality systems that make decentralized production possible, or a process that compresses manufacturing from weeks toward days could all matter.

The biggest threat comes from in vivo engineering. If cell therapy eventually arrives in a vial and programs the cells inside the patient, much of today’s personalized manufacturing stack disappears.

So what is actually worth building in cell therapy?

The best cell-therapy companies to build now are in vivo immune-cell engineering, immune-evasive cell replacement and autoimmune immune reset, with next-generation off-the-shelf CAR-T close behind.

We would put in vivo engineering first. Human feasibility has now been demonstrated, several major pharmaceutical companies have bought their way into the category, and a successful product could remove leukapheresis, individualized manufacturing and potentially lymphodepletion from the same treatment. No other current cell-therapy technology attacks that many bottlenecks at once.

Immune-evasive cell replacement comes second and may eventually produce the larger market. Type 1 diabetes has already shown us that manufactured cells can restore a missing physiological function. The remaining problem, keeping those cells alive safely without chronic immunosuppression, is both valuable and reusable across diseases.

Autoimmune immune reset ranks almost alongside it. Cabaletta and Kyverna are moving well beyond isolated academic case reports and into registrational development. If CAR-T can produce durable drug-free remission across several autoimmune diseases, the market becomes much larger than the first oncology indications. The next jump will come from making that treatment easy enough for rheumatologists and neurologists to use routinely.

Off-the-shelf CAR-T has also moved back up our list. Caribou and Allogene are producing increasingly credible data, and immediate availability could let CAR-T move into community centers and earlier disease settings. We still want more durability before ranking it with the top three.

Engineered Tregs deserve a serious watch after TREGZI’s approval, especially for precise immune tolerance. Solid-tumor cell therapy remains attractive when a company has a genuine answer to tumor biology or manufacturing. Neural replacement could become a major category if the current Parkinson’s Phase 3 succeeds, although development cycles are long.

The broad conclusion is sharper than it was a few years ago. We already know cells can become powerful medicines. The next big cell-therapy companies will be built around making those medicines behave less like bespoke medical procedures and more like treatments that ordinary healthcare systems can actually use at scale.

What we would build	Conviction today	Why there is still room	What still has to work
In vivo immune-cell engineering	Highest	Could remove manufacturing, waiting and conditioning from CAR-T	Precise delivery and much cleaner safety
Immune-evasive cell replacement	Highest	Functional replacement cells already work in humans	Durable immune protection with strong safety controls
Autoimmune immune reset	Very high	Multiple diseases are showing deep responses after one treatment	Easier conditioning and broader outpatient use
Next-generation allogeneic CAR-T	High	Immediate treatment is becoming clinically credible	Autologous-like durability
Engineered Tregs	High upside, earlier	FDA has validated therapeutic Treg biology	Engineered antigen-specific tolerance in humans
Solid-tumor cell therapy	Selective	Huge need and commercial proof now exists	Better targeting, trafficking and manufacturing
Neural cell replacement	Selective	Parkinson’s has reached pivotal development	Phase 3 efficacy and practical delivery
Incremental autologous CAR-T	Low	Existing market is large	Hard to create enough differentiation
Generic cell manufacturing	Low	Real demand exists	Difficult to defend as workflows change

If you want more recent data on this point, please see our latest cell therapy market report.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.