NK Cell Therapy Market (2026-2036)

The NK cell therapy market covers therapies based on natural killer cells, including unmodified cells, chimeric antigen receptor modified cells, gene-edited cells and cytokine-induced memory-like cells, sourced from cord blood, peripheral blood, established cell lines and induced pluripotent stem cell lines, and supplied for haematological, solid tumour, infectious and autoimmune indications. The growth of this market is driven by donor-derived supply being a property of the cell rather than an engineering achievement, by combination with antibody therapy, by induced pluripotent stem cell lines supporting repeated dosing, and by regulatory pathways that allow earlier clinical entry in parts of Asia.

No natural killer cell therapy appears on the FDA list of approved cellular and gene therapy products, where every approved chimeric antigen receptor product is an autologous T cell therapy and the one approved engineered T cell receptor therapy is also autologous. Revenue in this market therefore comes from clinical-stage supply, from manufacturing and process development, and from treatment delivered under national frameworks that permit it before a marketing authorisation, rather than from approved product sales. The forecast is weighted towards the later part of the period for that reason.

Natural killer cells do not cause graft versus host disease in the way donor T cells do, which makes donor-derived supply a property of the cell rather than something that has to be engineered into it, and that is why this platform is treated as the natural route to an off-the-shelf product. The same property explains why no product has been approved, because these cells are cleared quickly and do not persist, and persistence is what produced the durable single-dose remissions that made T cell therapy a commercial product. Safety and persistence are the same characteristic read from opposite ends.

A natural killer cell recognises and kills a target cell without needing to be primed against a specific antigen, using inhibitory receptors that read whether a cell is displaying normal human leukocyte antigen molecules and activating receptors that read signs of stress or infection. A tumour cell that reduces its human leukocyte antigen display to hide from T cells becomes more visible to natural killer cells for that reason, which is the biological argument for the platform and the reason it is pursued alongside T cell approaches rather than against them. The two mechanisms cover each other, since the escape route from one is what exposes a cell to the other. That complementarity is the reason the platform has retained support through a long period without an approval. It also means an eventual product would sit alongside T cell therapy rather than displacing it.

No natural killer cell therapy appears on the FDA list of approved cellular and gene therapy products, where every approved chimeric antigen receptor product is an autologous T cell therapy and the one approved engineered T cell receptor therapy is also autologous. Every product in this field is therefore at clinical stage, and the revenue recorded in this market comes from supply into trials, from manufacturing and process development work, and from treatment delivered under national frameworks that permit it before a marketing authorisation. A forecast for this market is a forecast of approvals rather than of uptake. The composition of current revenue also means that growth in the early years reflects trial activity rather than patient demand.

Donor-derived supply is where the platform holds a clear advantage. Natural killer cells do not cause graft versus host disease in the way donor T cells do, so a product can be manufactured from a donor and given to an unrelated recipient without removing a receptor by gene editing, which is the work that allogeneic T cell programmes have to complete first. Off-the-shelf supply is consequently a starting property of this platform rather than an objective. A developer here begins where an allogeneic T cell programme arrives after several years of editing work. The saving is in time and regulatory risk rather than in manufacturing cost.

Persistence is where the same property works against it. These cells are cleared within days to weeks rather than expanding and remaining for months, so a single infusion does not produce the sustained pressure on a tumour that durable remissions have required. Developers have responded by supporting the cells with interleukin-15 and related cytokines, by inducing a memory-like state before infusion, and by dosing repeatedly, each of which treats short persistence as a characteristic to design around rather than a defect to remove. None of these approaches reproduces the expansion an engineered T cell population undergoes after infusion.

Repeat dosing changes the commercial shape of a product in this field. A course of several infusions requires several times the manufactured material a single T cell infusion does, which raises the manufacturing burden per patient at the same time as it spreads revenue across a treatment course rather than a single administration. Supply has to be reliable across the course, which favours a source that can be produced to a schedule over one dependent on arranging a donor collection for each batch. Interruption part-way through a course is a clinical problem as well as a commercial one.

Combination with antibody therapy follows directly from how these cells work. Natural killer cells are the effector cells that mediate antibody-dependent cellular cytotoxicity, so an infused population increases the number of effectors available to an antibody already in use, and the combination has a mechanistic rationale that monotherapy does not. This places the natural comparator for the platform alongside antibody therapy rather than against single-dose T cell products, and it is the development path most programmes have adopted. It also means a developer is proposing an addition to a regimen rather than a replacement for one. Clinicians assess an addition differently from a substitution, and so do payers. The incremental framing lowers the evidence bar in practice even where it does not formally. Most programmes in the field are now designed around that framing, and a monotherapy study has become the exception rather than the starting point it once was.

Source determines what can be supplied and at what consistency. Cord blood provides cells without an elective donor procedure and is supported by existing banking infrastructure, peripheral blood provides larger collections from adult donors, established cell lines provide an unlimited supply of a uniform product that requires irradiation before infusion, and induced pluripotent stem cell lines provide an indefinitely expandable starting material that can be edited once and used for every batch. The last of these suits repeat dosing better than any donor-dependent route.

In July 2026 the Ministry of Health and Welfare in South Korea, following a resolution of the Deliberation Committee on Advanced Regenerative Medicine and Advanced Biopharmaceuticals, reclassified treatment using the immune cells of the patient, including autologous natural killer cells, from medium risk to low risk and removed the mandatory preliminary clinical study, allowing treatment plans to proceed to deliberation on existing clinical results and published literature. As of June 2026, 65 advanced regenerative medicine clinical studies had been approved in South Korea, of which 13 were classified low risk, 30 medium risk and 22 high risk, so medium and high risk projects accounted for 80% of the total. The reclassification applies to treatment using cells taken from the patient rather than to the donor-derived products that dominate development in the United States and Europe, so it widens access to autologous treatment in that market without altering the position of off-the-shelf products.

Engineering has concentrated on adding a chimeric antigen receptor to direct the cells at a defined target, which changes the basis on which they recognise a tumour from the innate reading of surface signals to a specified antigen. This brings the platform closer to the approach that has already been approved in T cells while retaining the absence of graft versus host disease, and it is the segment attracting most of the development capital, though it also inherits the persistence problem the unmodified cells have.