Can Allogeneic Cell Therapy Solve the CAR-T Access Problem?

Key Takeaways

• Allogeneic CAR-T can separate product manufacturing from the timing of an individual patient’s treatment need, potentially eliminating the leukapheresis-to-manufacturing interval inherent to autologous CAR-T.

• Early allogeneic CAR-T trials have demonstrated clinically meaningful activity across hematologic malignancies, while persistence, host rejection, conditioning, and infection remain important considerations.

• Multi-dose manufacturing changes the CAR-T production model from patient-specific lots toward batches capable of supplying many recipients, increasing the importance of reproducibility, characterization, and batch-level control.

• Genome engineering used to reduce alloreactivity and immune rejection introduces additional CMC requirements involving editing efficiency, unintended genomic changes, genome integrity, and comparability.

• Off-the-shelf availability cannot by itself eliminate CAR-T access barriers related to geography, socioeconomic factors, referral pathways, treatment-center availability, conditioning, and clinical management.

The patient-specific manufacturing model that makes autologous chimeric antigen receptor T (CAR-T) cell therapy possible also creates some of its most persistent access constraints. Treatment cannot begin simply because a physician has decided that CAR-T is appropriate. Cells must first be collected, transported, manufactured into an individualized product, tested and released, returned to the treatment center, and ultimately infused into a patient who must remain eligible throughout that interval. For patients with aggressive disease, deterioration during the waiting period can affect whether treatment remains possible at all.¹

Allogeneic CAR-T proposes a fundamentally different model. Rather than beginning manufacturing after an individual patient is identified, cells from healthy donors can be engineered, expanded, cryopreserved, and banked in advance, creating an inventory of doses that could be available when patients need them. Clinical-stage manufacturing has already demonstrated the potential scale of that shift: one allogeneic process has been reported to generate approximately 100 cryopreserved doses from a single production run.^2,3

That change addresses one of the most distinctive structural limitations of autologous CAR-T, but it does not remove complexity. It relocates much of it. Healthy-donor cells introduce immunological challenges involving graft-versus-host disease (GVHD), host immune rejection, cellular persistence, and the intensity of lymphodepletion required to support expansion. Solving those problems increasingly depends on genome engineering, which adds manufacturing, analytical, comparability, and regulatory demands. Meanwhile, geographic, socioeconomic, referral, and treatment-center barriers would remain even if the CAR-T product itself were immediately available.⁴

The central issue is therefore whether allogeneic CAR-T can remove the delays inherent to patient-specific manufacturing without replacing them with biological, clinical, or manufacturing burdens that limit the access gains the platform is intended to create.

The access challenge surrounding autologous CAR-T begins well before infusion. Once a patient is selected, the therapy must progress through leukapheresis, manufacturing, quality checks, transportation, and preparation for treatment. Reported median intervals from leukapheresis to product delivery or infusion have varied substantially among products, including 17 days for axicabtagene ciloleucel, 36 days for lisocabtagene maraleucel, and 54 days for tisagenlecleucel in data summarized in the access literature. During that period, tumor burden can increase, organ function can deteriorate, and patients can lose treatment eligibility.¹

Manufacturing is not guaranteed to succeed, either. In a UK analysis of 981 patients with large B cell lymphoma (LBCL) approved for CAR-T therapy, 38 patients (3.87%) experienced manufacturing failure. Subsequent pathways included remanufacturing, use of an out-of-specification product, delayed treatment, and, for some patients, failure to receive CAR-T therapy.⁵

A failure rate of a few percent may appear modest when viewed only as a manufacturing metric, but an autologous failure occurs after a particular patient has already entered a time-sensitive treatment pathway. Manufacturing performance is therefore inseparable from clinical access because the product does not exist until that patient’s cells have successfully completed production.

Manufacturing, however, accounts for only part of the broader access problem. A 2026 analysis of CAR-T utilization for multiple myeloma in the United States found lower odds of CAR-T use among patients with a potential drive time of at least two hours to the nearest authorized treatment center compared with those living within one hour.⁴ Lower household income and nonmetropolitan residence were also associated with lower utilization. Referral pathways, provider familiarity, treatment availability, and other health-system factors create additional barriers.¹

Allogeneic CAR-T therefore needs to be judged against two benchmarks. The first is whether it can reduce the delay, variability, and failure risk created by individualized production. The second is whether those improvements can translate into broader treatment access in a system where manufacturing is only one part of the problem.

The attraction of allogeneic CAR-T begins with a reversal of the manufacturing sequence. Autologous manufacturing starts with the patient. Allogeneic manufacturing can start before that patient ever arrives.

Healthy-donor T cells can be engineered, expanded, characterized, cryopreserved, and organized into banks intended to provide ready-made doses at the time of clinical need. Instead of initiating a new manufacturing process after each treatment decision, the therapeutic product can already exist.

That shift can eliminate several patient-specific dependencies. Treatment no longer requires leukapheresis from the eventual recipient as the source of the therapeutic cells. Product availability need not depend on the manufacturing suitability of that patient’s collected material. The interval between collection and finished-product release can disappear from the individual patient’s treatment timeline, and manufacturing failure no longer has to occur only after the patient has committed time to the process.

The production model changes as well. The phase I ALPHA and ALPHA2 studies evaluated ALLO-501 and its successor, cemacabtagene ansegedleucel (cema-cel), respectively, both healthy-donor-derived, CD19-directed allogeneic CAR-T products engineered for use in unrelated recipients.³ The manufacturing process used for these products can generate approximately 100 doses from a single production run, which can then be cryopreserved and stored for use as needed. Instead of manufacturing a new CAR-T product for each patient after treatment is selected, one donor-derived production run can therefore create an inventory capable of supplying multiple patients.

The difference creates an economic hypothesis, but not yet an established economic conclusion. Allogeneic development has been associated with the prospect of greater cost-effectiveness, but the evidence reviewed here does not establish that current allogeneic CAR-T therapy has achieved lower commercial costs than autologous CAR-T.

For access, the stronger and better-supported claim is that allogeneic manufacturing can separate the timing of product manufacture from the timing of an individual patient’s need.

Donor-derived manufacturing decouples starting-material collection from the eventual recipient, but the resulting cells must function within an immune system that recognizes them as foreign.

That produces two opposing challenges. Donor T cells can recognize recipient tissues and cause GVHD, while the recipient’s immune system can recognize and eliminate the donor cells. The first problem raises safety concerns. The second can restrict expansion and persistence, potentially reducing the duration of therapeutic activity.^2,6

Much of allogeneic CAR-T engineering is consequently directed toward controlling those interactions. Disrupting the endogenous αβ T cell receptor can reduce the ability of donor cells to mount an alloreactive response against recipient tissues. Additional modifications can help the engineered cells resist host immune pressure or accommodate lymphodepletion strategies intended to suppress rejection.²

This creates a fundamental design tension. The cells need to be modified enough to behave safely in an unrelated recipient, but they must retain the functions required to expand, recognize malignant cells, exert antitumor activity, and persist long enough to produce a meaningful clinical effect.

Recent work with glycan shielding illustrates how directly this problem is shaping next-generation design. The approach was investigated as a strategy for reducing immune recognition while preserving endogenous T-cell receptor function. Early clinical observations remain too limited to establish a new standard, but they show that developers are still reconsidering which cellular functions must be removed and which may be advantageous to retain.⁷

Early studies have established that allogeneic CAR-T cells can be administered to unrelated recipients and produce antitumor responses. The more difficult questions concern the consistency, durability, and clinical burden of those responses.

In the phase I CALM study, 25 adults with relapsed or refractory B cell acute lymphoblastic leukemia received genome-edited, healthy-donor-derived UCART19. The overall response rate was 48%. Two patients developed grade 1 acute cutaneous GVHD, while grade 3 or higher infections occurred in seven patients and grade 3 or higher cytokine release syndrome (CRS) occurred in six.⁸

A different experience has emerged in LBCL. Among 33 CAR-T–naïve patients treated with cema-cel or ALLO-501 in the ALPHA and ALPHA2 studies, the overall response rate was 58%, with complete responses in 42%.³ CAR-T cells were detected for as long as four months, and no GVHD, immune effector cell-associated neurotoxicity syndrome (ICANS), or grade 3 or higher CRS was reported in the cohort.

Allogeneic activity has also been demonstrated beyond CD19-directed treatment. In the phase I UNIVERSAL study, 43 patients with relapsed or refractory multiple myeloma received the B cell maturation antigen-targeted product ALLO-715 following lymphodepletion that included the anti-CD52 antibody ALLO-647. Across the full treated population, 55.8% responded. Among 24 patients receiving the specified 320 million–cell dose with the fludarabine, cyclophosphamide, and ALLO-647 regimen, the response rate was 70.8%, and median duration of response was 8.3 months. Infections occurred in 53.5% of patients overall, including grade 3 or higher infections in 23.3%.⁹

These studies involve different malignancies, constructs, doses, patient populations, and conditioning regimens, so their response rates should not be compared as though they came from head-to-head trials. Collectively, however, they establish that engineered allogeneic CAR-T can produce clinically meaningful activity. They also show that meaningful responses can be achieved without substantial GVHD in some engineered allogeneic platforms.

These early studies have moved the question beyond whether an off-the-shelf CAR-T product can produce clinical activity. The more consequential issue is whether it can deliver sufficient durability and safety while preserving the availability advantage that motivated the platform.

The pharmacology of UCART19 demonstrates why persistence remains central. In the analyzed CALM population, CAR+ T cell persistence did not exceed 28 days in 10 of 25 patients and extended beyond 42 days in only four. Responders had greater UCART19 expansion and exposure than nonresponders, while the lymphodepletion environment also affected cell kinetics and host T cell reconstitution.¹⁰

The ALPHA/ALPHA2 experience shows that persistence can extend longer in another allogeneic platform, with detectable CAR-T cells reported for up to four months. Among patients achieving complete response, the reported median duration of response was 23.1 months.³

These findings make persistence important without defining a universal persistence threshold. They also complicate any assumption that successful allogeneic therapy must reproduce the exact pharmacokinetic profile of an autologous product. Greater expansion and exposure were associated with response in UCART19, but durable complete responses in ALPHA/ALPHA2 were observed even though indefinite cellular persistence was not demonstrated.

The unresolved issue is therefore not simply how to make donor-derived cells persist longer. It is how much expansion and persistence a particular product needs to achieve durable clinical benefit without increasing other biological or treatment burdens.

An inventory-ready cell product does not automatically create an immediately deliverable treatment.

The principal early allogeneic programs have relied on lymphodepletion to create an environment in which donor-derived CAR-T cells can expand. CALM used fludarabine and cyclophosphamide with or without alemtuzumab. ALPHA/ALPHA2 used fludarabine, cyclophosphamide, and ALLO-647, while UNIVERSAL also evaluated an ALLO-647-containing lymphodepletion strategy.

Clinically important infections occurred within these treatment settings. CALM reported grade 3 or higher infections in seven patients, while UNIVERSAL reported infections in 53.5% of patients, including grade 3 or higher infections in 23.3%.^8,9 These observations do not establish that lymphodepletion alone caused those infections, but they make clear that ready-made cells are still being administered within treatment regimens that involve substantial immune suppression and associated clinical management.

For access, that distinction matters. A dose may be sitting in inventory, but the patient still requires conditioning, infusion, monitoring, and management of treatment-related toxicities. If supporting allogeneic-cell expansion continues to require intensive host immunosuppression, some of the specialized clinical infrastructure associated with CAR-T delivery remains necessary.

One measure of progress for newer allogeneic designs may therefore be whether they can reduce dependence on intensive lymphodepletion while maintaining adequate expansion and persistence.

The multi-dose allogeneic model creates important similarities to more conventional batch manufacturing, but CAR-T retains distinctive manufacturing complexity.

Production still depends on specialized good manufacturing practice (GMP) systems, process control, quality control, analytical testing, cryopreservation, and carefully managed scale-up. Product complexity can increase further as developers add genetic modifications intended to control GVHD, host rejection, or other aspects of cell behavior.¹¹

The multi-patient batch also changes the consequences of manufacturing control. In autologous production, an unsuccessful lot principally jeopardizes the treatment of the patient from whom the starting cells were collected. An allogeneic production run may generate doses for many recipients. The approximately 100-dose manufacturing run described in ALPHA/ALPHA2 provides a concrete example of that change in scale.³

The U.S. Food and Drug Administration’s (FDA) 2024 draft guidance for expanded allogeneic cells reflects the distinction. Its risk framework considers the cells’ expansion potential, reagents used during manufacturing, and the number of individuals that the resulting product could treat.¹²

This places substantial importance on the reproducibility and characterization of multi-dose batches. Donor selection and starting-material qualification, expansion, potency, identity, cryopreservation, thaw performance, release testing, and comparability all contribute to the effort to ensure that scaling the manufacturing model does not increase uncertainty in the product delivered to patients. CAR-T development guidance also places chemistry, manufacturing, and controls (CMC), analytical comparability, and product-specific characterization at the center of development for autologous and allogeneic products.^11–13

For contract development and manufacturing organizations (CDMOs), these requirements place a premium on capabilities spanning cell sourcing and expansion, analytical development, potency testing, cryopreservation, aseptic operations, and release strategy, with genome engineering adding another layer of specialization where it forms part of the product design.

Many of the strategies intended to make donor-derived cells safer and more durable depend on genome editing.

Editing can help reduce alloreactivity, protect cells from host immune pressure, or enable specific lymphodepletion strategies. Each intended modification, however, must be understood not only biologically but as a product attribute created through manufacturing. Regulatory expectations for human gene therapies incorporating genome editing address product design, manufacturing and testing, nonclinical safety, and clinical development.¹⁴

Safety assessment also extends beyond confirming that the desired edit occurred. The FDA’s April 2026 draft recommendations for next-generation sequencing (NGS)-based assessment identify off-target editing, unintended genomic changes, and loss of genome integrity as potential risks requiring evaluation.¹⁵

The CMC challenge therefore becomes more demanding as product design becomes more sophisticated. Developers need to characterize intended editing efficiency, unintended changes, genomic integrity, and residual populations that do not carry the intended modifications. They also need to understand whether manufacturing changes affect those attributes and whether analytical comparability can be established when an editing platform or process evolves during development.

Some of the engineering intended to make allogeneic cells easier to use clinically can therefore make the resulting product more demanding to manufacture and characterize consistently.

If allogeneic CAR-T succeeds in making therapeutic inventory readily available, it could remove a bottleneck intrinsic to autologous therapy: the need to create a new drug product after each individual patient is identified.

That would not eliminate the other barriers already documented across CAR-T delivery. Geographic distance, socioeconomic factors, referral pathways, treatment availability, and provider familiarity can still influence whether eligible patients reach cellular therapy programs.

Allogeneic CAR-T should therefore be judged according to the access barriers it is positioned to change. It could shorten the interval between treatment decision and product availability, reduce dependence on patient-specific manufacturing success, make scheduling more predictable, and replace individual production campaigns with inventory generated from multi-dose manufacturing.

Whether those advantages eventually allow CAR-T delivery to move beyond today’s most specialized treatment networks remains unresolved. Conditioning, infusion, toxicity management, and clinical monitoring remain part of the treatment pathway. Broader delivery will depend not only on how cells are manufactured but on how safely and simply the overall therapy can be administered.

Allogeneic CAR-T has progressed beyond a theoretical manufacturing solution. Healthy-donor cells have been genome engineered, manufactured in advance, administered to unrelated recipients, and associated with clinically meaningful responses in early studies. Multi-dose production has also demonstrated a manufacturing model fundamentally different from the one-patient, one-product structure of autologous CAR-T.

If allogeneic platforms can combine rapid availability with adequate expansion and persistence, manageable conditioning, low GVHD and host-rejection risk, and reproducible multi-patient manufacturing, they could remove a structural barrier that autologous CAR-T cannot completely engineer away: the requirement to manufacture the therapeutic product from the patient after the patient already needs treatment.

That would represent a meaningful solution to the CAR-T access problem, but not a complete one. Geography, referral, infrastructure, and socioeconomic barriers would remain, and an inventory-ready product would still have to fit into a clinically manageable treatment pathway.

The promise of allogeneic CAR-T ultimately rests on more than putting cells on a shelf. Its success will depend on whether advance manufacturing can translate into reliably available treatment without shifting the limiting constraint from production time to persistence, conditioning, product complexity, or clinical delivery.

1. Nze, Chijioke, and Christopher R Flowers. “Barriers to accessing cellular therapy for patients receiving care in community practices.” Hematology: American Society of Hematology Education Program. 2023: 382–385 (2023).

2. Georgiadis, Christos, Roland Preece, and Waseem Qasim. “Clinical development of allogeneic chimeric antigen receptor αβ-T cells.” Molecular Therapy. 33: 2426–2440 (2025).

3. Locke, Frederick L, et al. “Allogeneic Chimeric Antigen Receptor T-Cell Products Cemacabtagene Ansegedleucel/ALLO-501 in Relapsed/Refractory Large B-Cell Lymphoma: Phase I Experience From the ALPHA2/ALPHA Clinical Studies.” Journal of Clinical Oncology. 43: 1695–1705 (2025).

4. Blue, Brandon J, et al. “Geographic Variation and Inequities in Burden and Access to CAR T-Cell Therapies for Multiple Myeloma in the US.” Clinical Lymphoma, Myeloma & Leukemia. 18 Jul. 2026.

5. Dulobdas, Vaishali, et al. “Risk factors for CAR T-cell manufacturing failure and patient outcomes in large B-cell lymphoma: a report from the UK National CAR T Panel.” Blood Cancer Journal. 15: 30 (2025).

6. Pancholi, Neha J. “Driving CAR T-cell Therapy in New Directions.” American Association for Cancer Research. 25 Aug. 2026.

7. Wu, Zeguang, et al. “Glycan shielding enables TCR-sufficient allogeneic CAR-T therapy.” Cell. 188: 6317–6334.e21 (2025).

8. Benjamin, Reuben, et al. “UCART19, a first-in-class allogeneic anti-CD19 chimeric antigen receptor T-cell therapy for adults with relapsed or refractory B-cell acute lymphoblastic leukaemia (CALM): a phase 1, dose-escalation trial.” The Lancet Haematology. 9: e833–e843 (2022).

9. Mailankody, Sham, et al. “Allogeneic BCMA-targeting CAR T cells in relapsed/refractory multiple myeloma: phase 1 UNIVERSAL trial interim results.” Nature Medicine. 29: 422–429 (2023).

10. Dupouy, Sandra, et al. “Clinical Pharmacology and Determinants of Response to UCART19, an Allogeneic Anti-CD19 CAR-T Cell Product, in Adult B-cell Acute Lymphoblastic Leukemia.” Cancer Research Communications. 2: 1520–1531 (2022).

11. Dias, Juliana, et al. “CAR-T cell manufacturing landscape—Lessons from the past decade and considerations for early clinical development.” Molecular Therapy – Methods & Clinical Development. 32: 101250 (2024).

12. Safety Testing of Human Allogeneic Cells Expanded for Use in Cell-Based Medical Products. Draft Guidance for Industry. U.S. Food and Drug Administration. Center for Biologics Evaluation and Research. 30 Apr. 2024.

13. Considerations for the Development of Chimeric Antigen Receptor (CAR) T Cell Products. Guidance for Industry. U.S. Food and Drug Administration. Center for Biologics Evaluation and Research. 31 Jan. 2024.

14. Human Gene Therapy Products Incorporating Human Genome Editing. Guidance for Industry. U.S. Food and Drug Administration. Center for Biologics Evaluation and Research. 29 Jan. 2024.

15. Safety Assessment of Genome Editing in Human Gene Therapy Products Using Next-Generation Sequencing. Draft Guidance for Industry. U.S. Food and Drug Administration. Center for Biologics Evaluation and Research, Office of Therapeutic Products. 15 Apr. 2026.