Let's Stop Treating Allogeneic Cell Therapy As One Thing

By Arnaud Deladeriere, Ph.D., Cell&Gene Consulting Inc.

Allogeneic cell therapy is often discussed as though it were a single modality moving along a single development path. That framing is increasingly unhelpful. A solid tumor program, an iPSC-derived regenerative therapy, and a mesenchymal stromal cell (MSC) product may all sit under the allogeneic label, but they carry different biological requirements, manufacturing assumptions, and regulatory burdens.

We've already proved that some products work, and others are producing an increasingly robust body of data, though perhaps largely out of the spotlight. To recenter allogeneic therapies and exploit them for all the therapeutic good they can do, we must now learn to judge each program against the right category, the right economic model, and the right clinical objective.

Part 1 of the Cell&Gene Foundry's June discussion explored biological considerations and the promising progress, including from recent trial readouts at Caribou Biosciences and others. In the second part, the group turned from biology to structure: how allogeneic is defined, how it is paid for, and what it will take for more programs to reach late-stage development.

Allogeneic Is Not One Category

One of the more useful reframings concerned durability. The reflex is to treat long-term persistence as a requirement. For some products it is, but the cancer data argues otherwise in that setting. In roughly 90% of patients, CAR T cells become undetectable in circulation past day 28, yet a subset of those patients hold multiyear complete responses, some beyond four years. When the disease is cleared, the cells do not need to remain. The transient profile carries a clinical benefit of its own, since immune reconstitution returns, B cells recover, and patients require less long-term immunoglobulin support.

That conclusion does not extend to every allogeneic product. An iPSC-derived pancreatic beta cell carries the opposite requirement, where the value of the therapy depends on the cells persisting and remaining immunologically invisible for years. More broadly, allogeneic functions as a catch-all term covering products with little in common. A heme malignancy product, a solid tumor product, an autoimmune reset, and a regenerative iPSC product carry different target product profiles and different definitions of success. Asking why allogeneic as a whole has been slow flattens distinctions that matter. In several regenerative indications, the relevant comparator is not autologous at all, because no existing therapy addresses the condition.

Editing And A Regulatory Classification Problem

If one CMC theme warrants more attention than it receives, it is the regulatory treatment of genome editing reagents. The development instinct, for sound CMC reasons, is to limit the number of edits, since each edit adds process validation, characterization, and release burden. The bigger issue, however, is regulatory classification.

The agency currently treats each editing reagent as a separate drug substance. A company using proprietary editing chemistry — in Caribou's case the advanced guides within its chRDNA platform — files a drug substance module for each one, validates the assays that release each one, and conducts process validation against each one. The cost per dose is modest once divided across the hundreds of doses produced from a single lot. The up-front development cost is not, and it scales with the number of edits.

No one in the Foundry argued that editing reagents should avoid scrutiny. The question was whether they should be regulated as a drug substance at all. These reagents are not administered to patients. The consequence of an underqualified editing reagent is a failed product lot, identified by release testing of the actual drug product, rather than a patient safety event. Reviewing an oligonucleotide drug substance module as though the oligonucleotide were the therapy applies a risk frame that the product's own characterization already covers. The recent FDA guidance on prior knowledge offers a possible lever, since platform and prior-knowledge arguments can reduce duplicated work, particularly for vector components already understood as platforms. Proprietary editing chemistries are harder to fold into that framing, which is where the cost concentrates. The point is worth stating plainly, because it currently shapes how much complexity early programs are willing to design into a product.

Price, Reimbursement, And The Unfinished MSC Story

On economics, the discussion was largely aligned. Allogeneic manufacturing produces hundreds of doses per batch, which provides a genuine cost-of-goods advantage and the option to price below autologous. Whether that option is exercised is a separate matter, because price tracks clinical value, indication, and the need to recover development cost far more closely than it tracks cost of goods. This mirrors a conclusion the Foundry has reached repeatedly: reduced manufacturing cost does not automatically translate into a lower price. The working expectation was that an allogeneic CAR-T will likely be priced near the autologous range, with the cost advantage appearing as margin rather than as a lower sticker. In vivo approaches are unlikely to break that pattern for the same reason.

The reimbursement logic beneath this is uncomfortable and worth stating directly. For a durable, potentially curative therapy in a rare indication, a high one-time price can still represent savings against a lifetime of existing treatment. The counterpressure is that a healthcare system does not capture the cost of patients who are not cured and who leave the system over time. The recent decision by Regeneron to provide its auditory gene therapy at no charge in the United States, enabled by a rare indication and a priority review voucher, sets a benchmark that the next program in that space will struggle to raise capital against, and a reminder that pricing in advanced therapy is rarely set by the developer alone.

Mesenchymal stromal cell products remain the unresolved case in this conversation. Mesoblast now holds a U.S. approval and held a Canadian approval more than a decade ago that never translated into uptake, for reasons that no one at the table could fully explain. The most likely interpretation is that the clinical impact, while real, has not been the kind of result that moves reimbursement and prescribing in the way an oncology cure does. The same question applies to the broader indications that several MSC programs are now pursuing, where the economics of an expensive cell therapy against a widespread condition remain unproved.

What It Will Take To Reach Late Stage

The most strategically useful part of the discussion concerned how an allogeneic program reaches and succeeds at late stage, and the answer had more to do with routing around autologous than with defeating it.

Caribou has reached alignment with the FDA on a Phase 3 study in lymphoma and is deliberately avoiding a head-to-head comparison against autologous CAR-T. A non-inferiority study against an autologous comparator is large, slow, and expensive, and while the program believes it could compete on that basis, the more capital-efficient route targets patients who are ineligible for or restricted from autologous therapy. The access gap is larger than the field generally acknowledges. By most estimates, 70% to 80% of eligible patients do not receive an autologous CAR-T, whether for reasons of cost, geography, or logistics. A therapy that reaches those patients does not have to outperform 15 years of optimized autologous data to deliver clear benefit.

 

In vivo programs will encounter the same constraint. To compete in the same markets, they will need to demonstrate superiority or non-inferiority against autologous, and they are some distance behind in building the broader-access argument that allows a program to step around that requirement. The limiting factor, as elsewhere, is capital. The head-to-head route is available. Raising enough to run it is the difficulty.

Beneath the specific topics, the recurring tension was between the pace investment expects and the pace biology permits. Teams design for the economics and for the theory of what a product should be, then wait for the biology to confirm or refute it. The biology is the readout, and it does not move on an investment timeline. That mismatch drives the cycles the field continues to experience, in which enthusiasm concentrates on a single modality, runs ahead of the data, and corrects sharply when persistence or durability fails to appear on schedule. A more rational allocation would distribute risk across modalities, which is not how the venture model tends to behave. This is part of why the group looked toward funding mechanisms outside the venture cycle, including Advanced Research Projects Agency for Health (ARPA-H), California Institute for Regenerative Medicine (CIRM), and the Cancer Prevention and Research Institute of Texas (CPRIT) model, and toward the faster-iteration approaches emerging from investigator-initiated trials in other regions, which trade a different risk and ethics profile for the ability to learn biological answers in humans earlier and at lower cost than the conventional path to IND allows.

Closing Perspective

Allogeneic cell therapy was never primarily a question of whether the modality works. The more accurate question is whether a given program has aligned its design, its donor and editing strategy, its CMC and regulatory burden, and its clinical path to a definition of success that it can both reach and finance. The programs reading out well at present are those that made those choices deliberately and early. For developers entering the space, that alignment is the part that remains within their control, and it is the part worth resolving before the clinical data arrives, not after.

Key Takeaways

About The Author:

Arnaud Deladeriere, Ph.D., is principal consultant at Cell&Gene Consulting Inc. Previously, he was head of MSAT and Manufacturing at Triumvira Immunologics, and before that, manufacturing manager at C3i. He received his Ph.D. in biochemistry from the University of Cambridge.