“NK-cell therapy” is not one interchangeable product. Natural killer cells may come from the patient, a healthy adult donor, banked umbilical-cord blood, the NK-92 cell line, or an induced pluripotent stem-cell (iPSC) master bank. Source affects speed, biological variability, engineering, storage, repeat dosing, and safety controls. It does not, by itself, prove that a treatment will work.

Most platforms discussed here remain investigational and are being evaluated mainly in oncology trials. No source is universally superior. A mature peripheral-blood cell may offer strong immediate function; a clonal platform may offer reproducibility; cord blood may offer bankability and progenitors; an autologous product may avoid donor rejection. Those advantages come with different trade-offs [1].

What does “source” mean in NK-cell therapy?

The source is the biological starting material, not the complete medicine. Manufacturing then adds selection, activation, expansion, genetic engineering in some programs, washing, formulation, storage, thawing, and release testing. Two products beginning with donor blood can differ substantially after different culture systems. Conversely, products from different sources can be designed toward a similar clinical goal.

NK cells recognize stressed or malignant cells through the balance of activating and inhibitory receptors, release perforin and granzymes, produce cytokines, and can mediate antibody-dependent cellular cytotoxicity (ADCC) through CD16. Source, donor history, maturation state, culture, and cryopreservation can change each of these functions [2].

PlatformMain strengthMain constraintDevelopment status
Autologous PB-NKPatient-matched; no allogeneic rejectionVariable yield and possible self-tolerance or disease-related dysfunctionExperimental clinical platform
Healthy-donor PB-NKMature cells with established collection workflowsDonor variability and finite starting materialExperimental clinical platform
Cord-blood NK / progenitor NKBanked units; expansion and engineering potentialImmature phenotype and low starting numbersExperimental clinical platform
NK-92Clonal, expandable, engineerableTumor-derived line; clinical doses require irradiationExperimental clinical platform
iPSC-derived NKRenewable clonal bank and multiplex engineeringComplex differentiation, maturation, and genomic safety controlExperimental/early clinical platform

Autologous peripheral-blood NK cells

Autologous NK cells are collected from the same patient who will receive them. Leukapheresis or peripheral-blood collection supplies mononuclear cells, after which NK cells are enriched and activated. The theoretical attraction is immunologic compatibility: donor rejection and graft-versus-host disease are not expected from the patient’s own cells.

The limitation is that the starting material may reflect the patient’s age, cancer, infections, previous chemotherapy, corticosteroids, malnutrition, or systemic inflammation. The collected NK population may be small, exhausted, or difficult to expand. Autologous NK cells are also educated to tolerate the patient’s own HLA, which can reduce activity against a tumor that retains inhibitory self-signals. Engineering or antibodies may partly redirect recognition, but they do not erase every biological constraint.

Autologous manufacture is a one-patient-one-batch process. Scheduling, collection, contamination control, chain of identity, and release all occur for one recipient. If the culture fails or disease progresses during production, there may be no banked alternative. For that reason, autologous NK has value in selected protocols but is not automatically the simplest approach.

Healthy-donor allogeneic peripheral-blood NK cells

Allogeneic PB-NK cells come from another person, usually a screened healthy donor. They can provide mature CD56dim CD16-positive cells with direct cytotoxic and ADCC capacity. Early haploidentical studies established that donor NK cells could be transferred and expand under selected conditioning and cytokine regimens [3].

Donor selection can consider infectious screening, HLA/KIR relationships, NK-cell number, expansion yield, phenotype, and functional assays. Yet there is no single universally validated “best donor” algorithm. KIR-ligand mismatch may be useful in some hematologic settings, but its relevance depends on disease, transplant context, conditioning, product composition, and study design.

A healthy donor can supply a cleaner and often fitter starting population than a heavily pretreated patient. A single collection may support several doses, and cryobanking can move manufacture toward off-the-shelf use. The trade-off is lot-to-lot and donor-to-donor variability. Host immunity can also reject the infused cells, limiting persistence. Allogeneic does not mean immune-invisible.

Cord-blood NK cells and cord-blood progenitors

Umbilical-cord blood can provide existing NK cells as well as hematopoietic progenitors that are differentiated into NK cells. Banked units can be screened, selected, and allocated without recalling an adult donor. Cord-blood NK cells tend to be less mature than adult PB-NK cells, with differences in KIR, CD16, adhesion molecules, and signaling; activation and expansion can strengthen function [4][5].

Starting cell numbers are limited, so manufacturing usually relies on substantial expansion or progenitor differentiation. Closed or semi-closed processes using cytokines, stromal support, or engineered feeder cells have generated clinical-scale products [6]. Each method changes phenotype and creates its own raw-material and safety requirements.

A landmark early trial used cord-blood-derived anti-CD19 CAR-NK cells engineered with IL-15 and an inducible caspase-9 safety switch. The small study supported feasibility in selected blood cancers, but it should not be generalized to unmodified cord-blood NK cells, other targets, solid tumors, or commercial products [7]. The result belongs to that defined construct and process.

Closed-system expansion, cryopreservation, and quality testing of NK cell products

NK-92: a cell line, not ordinary donor NK cells

NK-92 is an immortalized, lymphoma-derived cell line. It grows reproducibly, can be cloned, and is comparatively easy to engineer, making it useful for research and standardized manufacturing. Its receptor profile differs from primary NK cells; parental NK-92 notably lacks meaningful native CD16, so unmodified cells do not provide conventional ADCC in the same way as mature PB-NK cells [8].

Critical safety distinction

Clinical NK-92 products are irradiated before infusion because NK-92 is a continuously proliferating, tumor-derived line. Irradiation prevents continued division but also restricts persistence and can reduce later cytotoxic capacity. “Off-the-shelf” does not remove this requirement, and non-irradiated laboratory results cannot be assumed to predict an irradiated clinical product.

Published phase I experience has evaluated irradiated NK-92, not unrestricted proliferating cells [9]. Irradiation creates a deliberate safety-versus-persistence trade-off: the dose is more controllable, but repeated infusions may be needed, and activity declines. Experimental alternatives such as suicide switches or other proliferation controls require their own validation; they are not interchangeable with the established irradiation safeguard [10].

iPSC-derived NK cells

iPSC-derived NK cells begin with a renewable pluripotent cell clone that is differentiated through hematopoietic stages into NK cells. One master cell bank can theoretically support many standardized lots. Genetic edits can be introduced and characterized at the clonal iPSC stage before differentiation, which makes the platform attractive for CARs, high-affinity CD16, cytokine support, checkpoint modification, or safety switches.

Clinical-scale differentiation has been demonstrated, and preclinical studies have shown functional NK production and CAR engineering [11][12][13]. However, “renewable” does not mean simple. Differentiation can take weeks and may produce cells with immature or platform-specific KIR/CD16 profiles. Manufacturing must control the pluripotent bank, editing, differentiation efficiency, residual undifferentiated cells, genomic stability, identity, potency, and adventitious agents.

iPSC-NK products are experimental platforms, with early clinical programs evaluating specific engineered candidates. Results from one clone or construct cannot validate all iPSC-derived NK cells. A clonal starting point reduces some donor variability but concentrates dependence on the quality and genetic integrity of that clone.

Expansion changes the product

Expansion is not merely making more of the same cells. Cytokines such as IL-2 and IL-15, combinations including IL-12/15/18, or feeder systems expressing 4-1BBL and membrane-bound IL-15 or IL-21 can change proliferation, activation, metabolism, receptor expression, cytokine dependence, and exhaustion. Expansion duration and stimulation intensity influence both yield and function.

Feeder-free systems reduce exposure to another cell line but may achieve different yields or phenotypes. Feeder-based systems can be powerful, yet feeder cells must be irradiated, controlled, and removed or limited in the final product. Residual feeder testing, sterility strategy, and qualification of cytokines and media belong in the product specification.

Fold expansion alone is a poor quality claim. A large count is unhelpful if cells have low viability, weak serial killing, excessive cytokine release, contamination, or a phenotype inconsistent with the intended mechanism. Potency testing should reflect the proposed use and be linked to manufacturing consistency.

Cryopreservation: convenience with a functional cost

Cryopreservation enables inventory, transport, repeat dosing, and centralized release, but thawed NK cells may not behave like fresh culture. Post-thaw recovery, cytotoxicity, cytokine production, migration, proliferation, and persistence can all change. The effect depends on source, activation state, freezing medium, cooling rate, storage, thawing, washing, and post-thaw rest [14].

Some optimized products retain substantial function after freezing, showing that cryoinjury is not inevitable or identical across platforms [15]. Nevertheless, a pre-freeze count cannot substitute for post-thaw viable-cell recovery and potency. A product labeled “90% viable” can still have impaired function, and a fresh product is not automatically potent. Release should measure attributes at the clinically relevant time point.

What standardization should include

Specifications cannot be copied blindly from one source to another. CD16 expectations for PB-NK may be inappropriate for an intentionally engineered iPSC-NK product; irradiation is essential for NK-92 but not a routine requirement for primary donor NK. Standardization means defining critical quality attributes for the actual platform and showing that the manufacturing process controls them [16].

Safety is source-specific and protocol-specific

Potential risks include infusion reactions, fever, cytopenias from lymphodepleting chemotherapy, infection, cytokine-related toxicity, on-target/off-tumor effects from engineered receptors, and immune rejection. Primary allogeneic NK cells generally have a lower intrinsic graft-versus-host risk than conventional allogeneic T cells, but contaminating T cells, conditioning, cytokines, combination drugs, and the disease context still matter.

Engineering adds risks from the construct and manufacturing method. Cytokine-support genes may improve persistence but require control; CAR specificity can injure healthy antigen-expressing tissue; genome editing needs genomic assessment; iPSC products require exclusion of residual pluripotent cells; NK-92 requires validated irradiation and confirmation of loss of proliferative capacity. Long-term follow-up may be required for genetically modified products.

How should patients interpret source claims?

Ask for the exact platform, not simply “NK cells.” A credible protocol should disclose whether cells are autologous or allogeneic, primary or cell-line derived, expanded or minimally manipulated, genetically modified or not, fresh or cryopreserved, and investigational or authorized for the stated use. It should explain the trial identifier or regulatory basis, release tests, dosing rationale, preparative chemotherapy, supportive cytokines, and emergency plan.

Marketing phrases such as “young cells,” “universal,” “natural,” “GMP,” or “billions of cells” do not establish clinical benefit. GMP describes manufacturing control, not proven efficacy. A platform’s scalability also does not show that it reaches a solid tumor, persists long enough, or improves survival. Comparative claims require head-to-head evidence, which remains limited.

Frequently asked questions

Are autologous NK cells safer because they are my own?

They avoid allogeneic mismatch, but safety and quality still depend on collection, culture, contamination control, dose, co-treatments, and the patient’s condition. Autologous origin does not guarantee potency.

Are donor NK cells truly off the shelf?

They can be banked as multi-dose lots, but availability depends on validated freezing, post-thaw function, inventory, compatibility strategy, and regulatory release. Some programs still manufacture fresh donor-specific batches.

Is cord blood the same as iPSC-derived NK?

No. Cord blood is donated neonatal blood containing NK cells and progenitors. iPSC-NK begins from a reprogrammed pluripotent clone and uses a controlled differentiation process.

Why must NK-92 be irradiated?

NK-92 is a continuously growing tumor-derived line. Irradiation prevents proliferation after infusion; it also limits persistence and may reduce function over time.

Which NK source is best?

No source is best for every target and protocol. The appropriate comparison includes biological function, manufacturing reliability, clinical evidence for the exact product, safety controls, timing, and patient context.

Limitations and medical disclaimer

This article compares manufacturing platforms, not commercial products, and does not recommend one treatment. Most NK-cell products described are investigational, and evidence is concentrated in early-phase oncology studies. Product composition and outcomes can change substantially with source, construct, expansion, freezing, conditioning, disease, and combinations.

This information is educational and is not medical advice. Patients should discuss established alternatives and clinical-trial eligibility with an oncologist or relevant specialist. Never delay standard cancer treatment for an unverified cell product.

References

  1. Myers JA, Miller JS. Exploring the NK cell platform for cancer immunotherapy. Nature Reviews Clinical Oncology. 2021;18:85–100. doi:10.1038/s41571-020-0426-7
  2. Shimasaki N, Jain A, Campana D. NK cells for cancer immunotherapy. Nature Reviews Drug Discovery. 2020;19:200–218. doi:10.1038/s41573-019-0052-1
  3. Miller JS, Soignier Y, Panoskaltsis-Mortari A, et al. Successful adoptive transfer and in vivo expansion of human haploidentical NK cells in patients with cancer. Blood. 2005;105:3051–3057. doi:10.1182/blood-2004-07-2974
  4. Verneris MR, Miller JS. The phenotypic and functional characteristics of umbilical cord blood and peripheral blood natural killer cells. British Journal of Haematology. 2009;147:185–191. doi:10.1111/j.1365-2141.2009.07768.x
  5. Sarvaria A, Jawdat D, Madrigal JA, Saudemont A. Umbilical cord blood natural killer cells, their characteristics, and potential clinical applications. Frontiers in Immunology. 2017;8:329. doi:10.3389/fimmu.2017.00329
  6. Spanholtz J, Preijers F, Tordoir M, et al. Clinical-grade generation of active NK cells from cord blood hematopoietic progenitor cells using a closed-system culture process. PLoS ONE. 2011;6:e20740. doi:10.1371/journal.pone.0020740
  7. Liu E, Marin D, Banerjee P, et al. Use of CAR-transduced natural killer cells in CD19-positive lymphoid tumors. New England Journal of Medicine. 2020;382:545–553. doi:10.1056/NEJMoa1910607
  8. Klingemann H, Boissel L, Toneguzzo F. Natural killer cells for immunotherapy—advantages of the NK-92 cell line over blood NK cells. Frontiers in Immunology. 2016;7:91. doi:10.3389/fimmu.2016.00091
  9. Tonn T, Schwabe D, Klingemann HG, et al. Treatment of patients with advanced cancer with the natural killer cell line NK-92. Cytotherapy. 2013;15:1563–1570. doi:10.1016/j.jcyt.2013.06.017
  10. Navarrete-Galvan L, Guglielmo M, Cruz Amaya J, et al. Optimizing NK-92 serial killers: gamma irradiation, CD95/Fas-ligation, and NK or LAK attack limit cytotoxic efficacy. Journal of Translational Medicine. 2022;20:151. doi:10.1186/s12967-022-03350-6
  11. Knorr DA, Ni Z, Hermanson D, et al. Clinical-scale derivation of natural killer cells from human pluripotent stem cells for cancer therapy. Stem Cells Translational Medicine. 2013;2:274–283. doi:10.5966/sctm.2012-0084
  12. Zeng J, Tang SY, Toh LL, Wang S. Generation of “off-the-shelf” natural killer cells from peripheral blood cell-derived induced pluripotent stem cells. Stem Cell Reports. 2017;9:1796–1812. doi:10.1016/j.stemcr.2017.10.020
  13. Li Y, Hermanson DL, Moriarity BS, Kaufman DS. Human iPSC-derived natural killer cells engineered with chimeric antigen receptors enhance anti-tumor activity. Cell Stem Cell. 2018;23:181–192.e5. doi:10.1016/j.stem.2018.06.002
  14. Saultz JN, Otegbeye F. Optimizing cryopreservation and post-thaw recovery of natural killer cells is critical for off-the-shelf platforms. Frontiers in Immunology. 2023;14:1304689. doi:10.3389/fimmu.2023.1304689
  15. Oyer JL, Croom-Perez TJ, Dieffenthaller TA, et al. Cryopreserved PM21-particle-expanded natural killer cells maintain cytotoxicity and effector functions in vitro and in vivo. Frontiers in Immunology. 2022;13:861681. doi:10.3389/fimmu.2022.861681
  16. Goldenson BH, Hor P, Kaufman DS. iPSC-derived natural killer cell therapies—expansion and targeting. Frontiers in Immunology. 2022;13:841107. doi:10.3389/fimmu.2022.841107