CAR-T Patent Landscape 2026: What Developers Must Know

Mapping the CAR-T patent landscape is now a prerequisite for anyone building a chimeric antigen receptor T-cell therapy, because the field’s commercial winners sit on a dense stack of foundational patents covering the target binder, the signalling domain and the manufacturing process. Seven products have reached the U.S. market since 2017, and the patents behind them — plus one landmark Federal Circuit decision — now define what a newcomer can build, buy or design around. This guide walks the landscape layer by layer.

Why the CAR-T Patent Landscape Is So Crowded

The CAR-T patent landscape is crowded because a CAR is not a single invention but a chain of them. A second-generation chimeric antigen receptor stacks an antigen-binding fragment, a hinge, a transmembrane region, a costimulatory domain and a signalling domain, and it is delivered by an engineered viral vector into T-cells through a multi-step manufacturing process. Each of those elements has attracted its own patents, often held by different academic institutions and companies.

The result is a thicket in which a product can be clear on the receptor yet blocked on the vector, or clear on the construct yet exposed on a method-of-treatment claim. Understanding the landscape means pulling it apart into layers rather than searching for a single ‘CAR-T patent.’

The Targets: CD19 Dominance and the BCMA Wave

Of the seven CAR-T products the FDA has approved, four target CD19 — Kymriah (tisagenlecleucel, 2017), Yescarta (axicabtagene ciloleucel, 2017), Tecartus (2020) and Breyanzi (2021) — while Abecma (2021) and Carvykti (2022) target BCMA for multiple myeloma, and Autolus’ Aucatzyl (obecabtagene autoleucel, 2024) returns to CD19 for B-cell acute lymphoblastic leukaemia.

CD19 is the most heavily patented target, and many approved products derive their binder from the same FMC63 antibody clone. That shared ancestry concentrates risk: a single binder family can read on multiple programmes, which is why the binder layer is usually the first stop in any CAR-T patent landscape review.

The Costimulatory Split: 4-1BB vs CD28

The sharpest fork in the landscape is the costimulatory domain. Kymriah and Breyanzi use 4-1BB; Yescarta and Tecartus use CD28. The two routes behave differently in patients — CD28 tends to drive faster, more intense expansion, 4-1BB a more durable one — but for an IP team the point is that each sits behind different foundational owners.

Choosing a signalling domain is therefore a freedom-to-operate decision as much as a clinical one. A developer that picks CD28 to improve potency may be walking into a different patent estate than one that picks 4-1BB, and the licensing economics can differ sharply between the two.

Foundational Owners and Key Patents

A few institutions anchor the CAR-T patent landscape. The University of Pennsylvania’s 4-1BB CAR work, associated with Carl June and exclusively licensed to Novartis, underpins Kymriah. St. Jude Children’s Research Hospital holds a 4-1BB construct patent (U.S. Patent No. 8,399,645) naming Dario Campana, which was licensed into the Juno/Bristol Myers Squibb portfolio. Memorial Sloan Kettering’s CD28 work, associated with Michel Sadelain, produced U.S. Patent No. 7,446,190.

On the commercial side the estates cluster around Novartis (Kymriah), Gilead’s Kite (Yescarta, Tecartus), Bristol Myers Squibb (Breyanzi, Abecma) and Johnson & Johnson with Legend Biotech (Carvykti). For a new entrant, the practical task is to map which owner controls which layer, because that determines who you negotiate with and where a design-around is realistic.

Juno v. Kite: The Decision That Rewrote the Landscape

No single event reshaped the CAR-T patent landscape more than Juno Therapeutics v. Kite Pharma. A jury had found Kite’s Yescarta infringed Sloan Kettering’s ’190 patent and the district court entered judgment of roughly $1.2 billion — the largest patent award the sector had seen. In August 2021 the Federal Circuit reversed, holding the asserted claims invalid for lack of written description because the patent claimed a broad genus of binders while disclosing only two working examples.

The Supreme Court declined to review the case in 2022, leaving the Federal Circuit’s reasoning in place. The lesson rippled across the field: the broadest, most intimidating CAR-T claims — the genus claims that appear to cover any binder against a target — are often the most vulnerable under 35 U.S.C. § 112. A claim that looks like an absolute block may not survive a challenge.

Vectors and Manufacturing: The Overlooked Layer

Beyond the receptor sits a quieter but equally real layer: the lentiviral and retroviral vectors used to engineer the T-cells, and the cell-processing methods that turn a patient’s blood into a dosed product. These carry their own patents, and because every autologous CAR-T relies on some delivery and manufacturing route, a vector or process claim can block a programme even when the construct is clear.

Patents are also not the whole exclusivity story. Because CAR-T products are approved under a Biologics License Application, the Biologics Price Competition and Innovation Act grants the reference product twelve years of regulatory exclusivity from first approval, layered on top of the patent estate.

Patent Timelines and the Coming Follow-On Competition

The CAR-T patent landscape is not static, and timing is part of the map. The earliest foundational filings date to the late 2000s and early 2010s, which means some of the broadest construct and method patents will begin to lapse across this decade, opening space that is closed today. A landscape review that ignores expiry dates overstates the blockers, because a family with only a year or two of life left is a very different obstacle than one with eight.

At the same time, the regulatory picture is shifting. The twelve-year BPCIA exclusivity on the first-approved products is a fixed clock, and as those clocks run down the prospect of biosimilar or follow-on cell therapies moves from theoretical to practical. Developers planning launches in the late 2020s and beyond should chart patent expiries and exclusivity end-dates on the same timeline, because the two together — not patents alone — decide when a window actually opens.

This is also why assignee monitoring matters. The large portfolios continue to generate new filings on next-generation features — armoured CARs, logic-gated receptors, allogeneic ‘off-the-shelf’ approaches — so the landscape a developer clears today will have grown by the time its product reaches the clinic. Treating the review as a living map rather than a one-off snapshot is the only way to stay ahead of it.

What the CAR-T Patent Landscape Means for New Entrants

For a developer, reading the CAR-T patent landscape is the start of a freedom-to-operate exercise, not the end of one. The landscape tells you where the thickets are; a clearance analysis tells you whether your specific construct reads on them and whether those patents would survive. We walk through exactly that sequence in our cell-therapy freedom-to-operate case study.

The developers who come out ahead treat the landscape as a design input — choosing a binder and a costimulatory route with the patent map open — rather than discovering the blockers after the manufacturing process is frozen. A structured freedom-to-operate analysis turns the landscape into an action plan: licence here, design around there, and build with confidence everywhere else.

How PerspireIP Can Help

At PerspireIP, our team helps innovators and businesses protect what they build. Whether you need a patent or trademark search, prior-art analysis, or an IP strategy tailored to your goals, we can help. Contact our team to discuss your next step.

Frequently Asked Questions