Human pluripotent stem cell-derived innate and adaptive immune cells for cancer immunotherapy

Summary

Allogeneic cell-based therapies hold great promise for cancer immunotherapy but face challenges like scalability, immune rejection, graft-versus-host disease, and toxicities. Human pluripotent stem cells (hPSCs), including embryonic and induced pluripotent stem cells, offer a scalable and adaptable platform to address these limitations. hPSCs provide an inexhaustible source of immune cells that can be genetically modified at the single-cell level to enhance anti-tumor activity and reduce immunogenicity. Recent advancements in generating iPSC-derived natural killer cells, T cells, and macrophages are opening the door to safer and more effective immunotherapies. This review examines the progress, challenges, and future directions in utilizing hPSC-derived immune cells to enhance cancer treatment and overcome barriers in allogeneic therapy.

Introduction

Adoptive transfer of engineered autologous immune cells has gained wide usage as a therapeutic approach for certain hematologic malignancies. The isolation of autologous T cells, followed by their genetic modification with tumor-targeted chimeric antigen receptors (CARs) and subsequent ex vivo expansion, has revolutionized adoptive cell therapy for various blood cancers ^1–4. Notably, autologous CAR-modified T cells designed to target CD19 have demonstrated remarkable clinical efficacy in refractory B cell leukemias and lymphomas ^2,3,5–7. To date, seven commercial autologous CAR-T (auto-CAR-T) cell products have been approved by the FDA for clinical use ⁸ including the recent obecabtagene autoleucel (Aucatzyl) ⁹. Despite the success of auto-CAR-T cell therapy in lymphoid malignancies, its efficacy against other cancers, such as acute myeloid leukemia (AML) ^10–12 and solid tumors ^13–15, remains limited. Auto-CAR-T cells carry the risk of cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), lengthy manufacturing time, batch-to-batch variation, and at times poor quality or insufficient quantity of primary T cells ^4,16–19. Due to these issues, it is estimated only about 25% of patients who could actually benefit from current auto-CAR-T cell therapies actually receive the treatment ²⁰. Conversely, allogeneic CAR-T (allo-CAR-T) cells can overcome most of these challenges; however, allo-CAR-T cell therapies face the risk of graft-versus-host disease (GvHD) and host-versus-graft (HvG) rejection, though cell engineering such as knock-out of the endogenous TCR via TRAC locus insertion of the CAR can be done to minimize the GvHD risk in allogeneic T-cell products ^7,21,22. Alternatively, allogeneic CAR-natural killer (NK) cells have emerged as a promising therapeutic approach ^23–26. Unlike T cells, allogeneic NK cells can be safely and effectively administered without strict HLA compatibility when treating heavily pretreated lymphoid malignancies ^24,26,27. Recent pre-clinical studies and autologous phase I clinical trial have tested the anti-tumor potential of CAR-engineered macrophages in solid tumors ^28–34. However, to date, clinical applications of allogeneic CAR-macrophages (CAR-Ms) have not been reported.

Compared with CAR-T, CAR-NK, and CAR-Ms derived from umbilical cord blood (UCB) or peripheral blood (PB) ^{7,16,23,24,34}, human pluripotent stem cells (hPSCs), encompassing both embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), offer an expandable and versatile cell source for immune cell generation. Early studies on hESC-derived NK cells demonstrated their potent tumor-killing capability in both in vitro and in vivo settings ^35–37. Although ESCs were used in initial investigations, the development of hiPSCs has markedly improved accessibility and expanded the genetic diversity represented among pluripotent stem cell lines. The advancement in the field has facilitated the generation of various specialized cell types from hPSCs, including vascular cells ^38,39, cardiomyocytes ^40,41, dopaminergic neurons ^42–44, mesenchymal stem cells ^45,46, insulin-producing cells ^47–49, neuronal cells ⁵⁰, and hematopoietic cells. The latter encompass macrophages ^32,33,51,52, neutrophils ^52,53, dendritic cells ^52,54, Langerhans cells ⁵², NK cells ^{35–37,55–57}, T cells ^58–60, gamma delta (γδ) T cells ^61,62, invariant natural killer T (iNKT) cells ^63,64, and mucosal-associated invariant T (MAIT) cells ^65,66, among others. Collectively, these developments highlight the growing translational potential of hPSC-derived products ⁶⁷.

A key advantage of using iPSCs to generate therapeutic cells is the ability to introduce genetic modifications at an undifferentiated stage, enabling the development of mature immune cells with stable expression or deletion of single or multiple genes. These modifications have significantly enhanced the anti-tumor activity of immune cells including CAR-T, CAR-NK, and CAR-Ms (as discussed below) ^{32,33,68–76}, and have also been leveraged to engineer “stealth” cells such as hypoimmune iPSC-derived products with improved immune compatibility for allogeneic recipients ^61,70,76,77. Furthermore, these genetic modifications at the undifferentiated PSC stage enable the production of clonal populations of differentiated cells, facilitating a standardized approach to cell therapy.

Reprogramming somatic cells into induced pluripotent stem cells

ESCs are derived from the inner cell mass of the blastocyst before gastrulation, a process marked by the formation of the three embryonic germ layers: endoderm, mesoderm, and ectoderm ^78,79. By contrast, iPSCs are generated by the reprograming of adult somatic cells using defined transcription factors (TFs) ^80,81. Like ESCs, iPSCs can propagate indefinitely as undifferentiated cells and retain the capacity to differentiate into all the three embryonic germ layers (Figure 1) ^82,83.

Figure 1.

Reprogramming of somatic cells and re-differentiation of somatic cell-derived hPSCs into therapeutic products for adoptive cell-based immunotherapy. Gr, granulocytes; Mac, macrophages.

Building on the prior context of iPSC applications in immune cell therapy, the ability to reprogram somatic cells into iPSCs has significantly expanded the potential for generating patient-specific and allogeneic therapeutic cell products, providing a variety of strategies for iPSC reprogramming. Initially, viral integration methods were used, such as Yamanaka’s cocktail of TFs; however, concerns regarding the oncogenic risks of transgene expression and potential insertional mutagenesis have made this approach less favorable ^84,85. To mitigate these risks, transgene-free reprogramming methods have been developed. One notable advancement is the use of non-integrating oriP/Epstein-Barr nuclear antigen-1 (EBNA1)-based episomal vectors or Sendai virus-based vectors, which do not integrate into the host genome, thereby leaving no genetic footprint ^86–88. In parallel, small molecules have been identified that either enhance reprogramming efficiency or replace TFs entirely. Although promising, the exclusive use of small molecules typically results in lower reprogramming efficiency compared with traditional methods ^89–91.

hPSC-derived hematopoiesis

hPSCs can be differentiated into hematopoietic cells using established protocols depending on the desired lineage ^92–97, with each protocol offering distinct advantages and limitations. In vitro differentiation of hPSCs into hematopoietic cells typically recapitulates extraembryonic hematopoiesis, including a primitive and a transient definitive wave characterized by erythro-myeloid progenitors (EMPs). The primitive wave of hematopoiesis generates primitive erythrocytes, tissue resident macrophages, and megakaryocytes, while the transient definitive hematopoiesis gives rise to erythrocytes, macrophages, neutrophils, megakaryocytes, mast cells, and NK cells (Figure 2) ^98,99. In hPSCs differentiation, extraembryonic hematopoietic progenitors emerge from a CD34-positive population enriched in hemogenic endothelium (HE) through endothelial-to-hematopoietic transition (EHT), a process closely mimicking physiological EHT observed in vivo ^100–103. The HE consists of bipotent endothelial cells capable of differentiating into both vascular endothelial cells, which form the vascular barrier and regulate nutrient transport and angiocrine signaling, and hematopoietic cells, which give rise to blood lineages ¹⁰⁴. The formation of primitive and transient definitive HE cells is regulated by Activin/Nodal and Wnt-β-catenin signaling pathways ^101,102,105.

Figure 2.

In vitro development of hPSC-derived hematopoietic cells from primitive and arterial type hemogenic endothelial cells. EBs, embryoid bodies; WNT, Wnt-β-catenin signaling pathways; CFU-Cs, colony-forming unit cells; Gr, granulocytes; Mac, macrophages; MK, megakaryocytes; NSG, NOD-Prkdc^scid IL2rg^null.

In a conventional hematopoietic differentiation system, formation of mesodermal lineage marks the first critical step in hematopoietic specification and HE development. This differentiation can be done either by formation of “spin-embryoid bodies” EBs or in a two-dimensional culture system. The EBs resemble hematopoietic organoids that contain hematopoietic progenitor cells, endothelial cells and stomal cells. The subsequent differentiation into erythrocytes, macrophages, neutrophils, megakaryocytes, and NK cells is highly dependent on specific combination of cytokines (Figure 2) ^{37,53,56,97,106}. NK cell generation from the spin-EBs involves replating the whole EB on plastic (tissue-culture plate) and allowing the EB-derived cells to form its own stroma with the NK cell differentiating on top of those adherent cells ³⁷. Other approaches to NK cell differentiation can also be utilized ^107,108. T lymphocyte progenitors are generated by diverting CD34⁺ populations toward definitive hematopoiesis through SB431542 inhibitor-mediated suppression of the primitive wave (transforming growth factor β [TGF-β]/Activin/Nodal pathway), leading to multipotent progenitors (MPPs) (Figure 2) ^53,109–112. These MPPs resemble in vivo EMPs; SB431542 has been specifically employed to promote the differentiation of lymphocyte progenitors (Figure 2) ^{52,59,112–116}. Significant advancements have enabled the successful generation of diverse hPSC-derived hematopoietic cells including red blood cells ^117–120, macrophages ^{51,52,121,122}, neutrophils ^52,123, DCs ^52,122, Langerhans cells ⁵², NK cells, T-, and B-lymphocytes (Figure 1 and 2) ^52,124–127.

A major milestone in this field has been the derivation of iPSC-based hematopoietic stem cells (iHSCs), as reported by Sugimura et al., ¹¹⁴ Piau et al., ¹²⁸ and Ng et al. ¹²⁶ (Figure 2). Compared with Sugimura et al., ¹¹⁴ who achieved iHSC generation through lentiviral-mediated TFs overexpression, both Piau et al. and Ng et al. demonstrated multilineage primary and secondary engraftment in mice following transplantation of iPSC-derived cells without the use of lentiviral transgenes. Piau et al. ¹²⁸ transplanted dissociated EBs; however, the relative contribution of hematopoietic versus non-hematopoietic cells to the bone marrow niche remains unclear. By contrast, Ng et al. ¹²⁶ transplanted enriched CD34 ⁺ cells and confirmed their contribution by detecting bone marrow erythroid, myeloid, B- and T-lymphoid cells, as well as CD45⁺CD34⁺CD38^low/− HSC-like cells. Additional evidence included splenic B and T cells, as well as thymic T cells. Engraftment was observed in 25–50% of recipients across four iPSC lines, with >70% human chimerism in the bone marrow of 19 out of 42 female mice at 16 weeks. Secondary transplantation resulted in low engraftment, likely due to limited primary cell input (0.3–2.0 × 10⁶ bone marrow cells) and the suboptimal non-irradiated NOD,B6.SCID^{Il2rγ−/− Kit(W41/W41)} (NBSGW) niche ¹²⁹. Notably, NK lineage contribution was not reported in either study.

hPSC-derived immune cells and immunotherapy

NK cells.

NK cells are essential components of the innate immune system, acting as first-line defenders against virus-infected and malignantly transformed cells without prior sensitization ^130,131. The ability of NK cells to discriminate between healthy and abnormal cells is governed by a dynamic interplay between activating and inhibitory receptors, which engage MHC class I molecules expressed on normal cells ¹³². The “missing self” recognition hypothesis, first proposed by Kärre and Ljunggren, explains how NK cells preferentially target cells with reduced or absent MHC class I expression while sparing those that retain it ^133,134. This innate mechanism enables NK cells to eliminate malignant cells that evade T cell-mediated immunity, a common tumor escape strategy involving MHC class I downregulation to avoid cytotoxic T cell recognition. Such evasion is especially relevant in the context of resistance to checkpoint blockade therapies ¹³⁵. Therefore, engagement of inhibitory NK receptors by self MHC class I ligands suppresses NK cell-mediated cytotoxicity, whereas low expression or loss of MHC class I ligands triggers NK-mediated killing ^136,137. Although allogeneic CAR-T cell therapy is attractive and can overcome certain limitations associated with autologous CAR-T cells, it poses risks such as GvHD and HvG rejection ^{4,6,7,16–19,22,138}. By contrast, allogeneic NK cells, even those that are HLA-mismatched, have demonstrated minimal toxicity and no GvHD in preclinical and clinical settings ^{24–26,139,140}. It has been reported that NK cells may be less prone to allograft rejection by downregulating the proliferation of CD8 T cells ¹⁴¹.

Like CAR-T cells, NK cell-based therapies have shown promise in hematologic malignancies but remain challenging for solid tumors ¹⁴². One major obstacle is the immunosuppressive tumor microenvironment (TME), which drives NK cell exhaustion through multiple inhibitory pathways.^143–145. For example, tumor cells often express HLA-E, which engages the CD94/NKG2A receptor on NK cells, CD155, which interacts with inhibitory receptors such as T cell immunoreceptor with Ig and ITIM domains (TIGIT)/CD96; and soluble factors like TGF-β1, which are secreted by both tumor cells and immune cells in the TME, to suppress NK cell activity ^145–153. Genetic disruption or pharmacological inhibition of NKG2A or TGF-β signaling has been shown to enhance NK cell cytotoxicity against HLA-E-expressing tumors ^{147,153–157}. Recent clinical studies have demonstrated enhanced anti-tumor activity of allogeneic NK cells against refractory non-small cell lung cancer (NSCLC) ^139,158. However, the sources of PB- and UCB-derived NK cells can be limited. iPSC-derived NK cells offer a scalable, off-the-shelf alternative that overcomes donor variability, expansion limitations, and challenges in single-cell genetic engineering associated with PB- or UCB-derived NK cells. Interestingly, preclinical studies have revealed that hPSC-derived NK cells demonstrate cytotoxicity comparable to that of PB- and UCB-derived NK cells ^{37,56,159–161}. NK cells generated from iPSCs expressing non-cleavable CD16a enhances antibody-dependent cell-mediated cytotoxicity (ADCC) whereas deletion of the cytokine-inducible SH2-containing protein (CISH) gene improves metabolism, persistence, and anti-tumor activity ^{68,69,71,162–164}. Despite these advances, iPSC-derived NK cells remain susceptible to HvG rejection. Disrupting the beta-2 microglobulin (B2M) gene eliminates surface expression of HLA class I molecules, including both class Ia and class Ib, avoiding recognition by allogeneic CD8⁺ T cells but increasing the risk of NK cell-mediated lysis ^146,165. HLA-E, -F, and -G (HLA class Ib molecules) serve as ligands for inhibitory receptors of NK cells, maintaining immune tolerance ^{146,166–170}. To mitigate NK rejection, forced expression of HLA-E in B2M-deficient iPSCs has been attempted ^171–173; however, due to endogenous NK cell heterogeneity, some subsets lack NKG2A and therefore may still attack iPSC-derived NK cells overexpressing HLA-E ^174–176. Given the pivotal role of HLA-C in suppressing NK cell activity through its interaction with inhibitory receptors such as killer-cell immunoglobulin-like receptors (KIRs) ^177,178, a preclinical study demonstrated that retaining HLA-C while deleting other class I and II molecules shows promise for evading both allogeneic NK and T cell responses. The study further suggests that, theoretically, a panel of 12 HLA-C-retained iPSC lines combined with HLA class II knockout (via disruption of class II major histocompatibility complex transactivator [CIITA]) could cover approximately 90% of the global population, offering broad applicability for regenerative medicine and immunotherapy ¹⁷⁹. In parallel, analysis of 10,000 consecutive UK cadaveric organ donors, compared with 6,577 patients on the UK kidney transplant waiting list, revealed that a panel of only ten donors homozygous for common HLA types could provide a complete HLA-A, HLA-B, and HLA-DR match for 37.7% of recipients and a beneficial match for 67.4% of the UK population ¹⁸⁰. Furthermore, a tissue bank comprising 150 selected homozygous HLA-typed volunteers could achieve matches for 93% of the UK population with minimal immunosuppression ¹⁸¹. In a similar effort, Yamanaka’s team has established a haplobank of HLA-matched iPSC lines covering ~40% of the Japanese population ¹⁸². Although promising, these strategies require rigorous validation in preclinical models before clinical translation. Nonetheless, iPSC-derived NK cells hold significant potential for universal, scalable, and effective cancer immunotherapy.

T-lymphocytes.

Genetic engineering of T cells with CARs has transformed adoptive cell transfer (ACT) for hematologic cancers, with CAR-T cells demonstrating remarkable efficacy in treating relapsed or refractory lymphoid malignancies ^{6,14,19,138,183–186}. However, autologous CAR-T cell generation from PB is labor-intensive and restrictive by limited cell availability, logistical issues, and most importantly lengthy vein-to-vein time ^{4,6,16–19,138,184–186}, necessitating the development of off-the-shelf allogeneic CAR-T therapies. Alternatively, innate-like T cells, including γδ T cells, iNKT cells, and MAIT cells, represent unconventional T cell subsets that bridge innate and adaptive immunity ¹⁸⁷. Unlike conventional αβ T cells, these populations arise early during thymic development and exhibit pre-programmed effector functions, enabling rapid responses to stress signals and microbial antigens without prior sensitization ¹⁸⁸. Importantly, they are less prone to GvHD, making them attractive candidates for allogeneic cell therapies ^{154,189–193}. γδ T cells develop from distinct thymic progenitors and express T cell receptors (TCRs) composed of γ and δ chains, which recognize non-peptidic antigens such as phosphoantigens and stress-induced ligands independently of classical MHC molecules ^{191,194–196}. iNKT cells express a semi-invariant αβ TCR that recognizes glycolipid antigens presented by the non-polymorphic CD1d molecule ¹⁹⁷. Despite their advantages, their rarity in PB (0.01–1% for iNKT and 0.5–5% for γδ T cells) limits efficient isolation and expansion ^198–201. MAIT cells express a semi-invariant TCRα chain, typically Vα7.2-Jα33 in humans (Vα19-Jα33 in mice) ^180,195. Their development requires the non-classical MHC class I-related molecule MR1, rather than classical MHC molecules used for conventional T cell selection. MAIT cells are particularly abundant in humans mucosal sites including gut, lungs, and liver, representing up to 40% of the resident T cells, compared with 10% in PB mononuclear cells ^202,203; however, CAR-MAIT cells have been less explored in cancer immunotherapy ^204,205. By contrast, iPSCs provide a renewable and adaptable platform for generating allogeneic CAR-T cells. These cells can be genetically modified to express CARs, alleviate GvHD and CRS, and produce clinical-grade T cells for ACT (Figure 3) ^{73,114,206–209}. The successful generation of TCR-less αβ T, γδ T, iNKT cells, and MAIT from iPSCs offers a viable alternative to autologous CAR-T therapies ^{61,65,75,210,211}.

Figure 3.

Genetic engineering strategies in hPSCs for producing HLA-unrestricted CAR-modified cells for cancer immunotherapy. Gr, granulocytes; Mac, macrophages.

Moreover, checkpoint molecules cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed death-ligand 1 (PD-L1), widely studied for their role in tumor immune evasion, have been used to protect iPSC-derived cells from allogeneic T cell rejection ²¹². In addition to T cells, PD-L1 expression may shield transplanted cells from immune rejection by PD-1-expressing immune cells including NK cells ^213,214, PD-1⁺ macrophages ^215,216, and PD-1⁺ DCs ^217,218. Furthermore, overexpression of HLA-E or HLA-G in iPSCs has been shown to inhibit NK cell-mediated rejection in HLA class I-deficient cells ¹⁷¹. These engineered iPSCs can be differentiated into hypoimmune T cells and other mature immune cell types for broad application in allogeneic ACT (Figure 3) ^75,219. This approach represents a major step towards scalable, off-the-shelf immunotherapies with reduced immunogenicity and enhanced clinical applicability.

Macrophages.

Macrophages play critical roles in tissue remodeling and repair ^220–222, apoptotic cell clearance ²²³, and maintaining immune homeostasis ²²⁴. These properties demonstrate remarkable plasticity with effector functions influenced by cytokines, pathogen-associated molecular patterns, metabolic cues, and tissue-specific interactions ^225,226. Within the TME, tumor-promoting macrophages enhance microvascular density, induce angiogenesis, facilitate invasion, and suppress immune responses through inhibiting NK and T cell activity ^227–232. By contrast, tumor-suppressive macrophages can infiltrate tumors and phagocytose cancer cells, remodel the TME, and potentiate adaptive immunity through antigen presentation, co-stimulation, and cytokine production ^{226,233–235}.

Recently preclinical and clinical studies have explored engineering macrophages with CARs. Studies using CAR-transduced human leukemia monocytic Tohoku hospital pediatrics-1 (THP-1) cells, bone marrow-derived macrophages (BMDMs), and UCB-derived macrophages have demonstrated efficacy with CARs targeting HER2, CD19 and carcinoembryonic antigen (CEA), respectively ^{31,34,240,236}. Unlike CAR-T cells, the autologous CAR-Ms product “CT-0508” has shown promising safety profiles in HER2-positive cancers, exhibiting no grade 3 or 4 CRS or neurotoxicity ³¹. Autologous CAR-Ms were tested in a single clinical trial that demonstrated safety and modest anti-tumor activity ³¹. Although these developments underscore the potential of CAR-M therapies, several challenges persist, including donor variability, low CAR transduction efficiency and short CAR-maintenance, limited expansion capacity post-engineering, and the lack of scalable allogeneic off-the-shelf products ^75,237,238.

To address these limitations, research groups have developed iPSC-derived CAR-macrophages (CAR-iMs). Notably, large-scale manufacturing of CAR-iMs has been successfully achieved for both infectious disease models ²³⁹ and cancer immunotherapy ^240,241. We and others have recently demonstrated that CAR-iMs engineered from an iPSC clone significantly extended survival rates in mouse models of cancer ^32,33. Despite these promising results, the clinical application of allogeneic CAR-M/CAR-iMs therapies remains unexplored. Key challenges in the field include the development of off-the-shelf products, achieving robust in vitro and in vivo expansion, and ensuring sustained phenotypic persistence of the final cell product. Continued innovation and optimization are critical to bridging the gap between experimental success and clinical applicability.

Allogeneic NK and T cell products in clinical trials

iPSC-derived immune cell products are emerging as transformative tools in cancer immunotherapy, addressing key limitations of conventional therapies. Their applications span preclinical and clinical settings, with a particular focus on NK and T cells.

NK Cells.

The safety and feasibility of allogeneic PB- and UCB-derived CAR-NK cell products have been reported in multiple clinical settings ^{24,26,242,243}. Allogeneic NK cells, even when HLA-mismatched, have shown minimal toxicity in both preclinical and clinical studies ^{23–25,244,245}. In addition, in vitro-expanded NK cells displaying potent cytotoxicity against B-cell chronic lymphocytic leukemia (B-CLL) ^246,247. Notably, the use of CB-derived CD19-targeting allogeneic CAR-NK cells achieved a 73% response rate in heavily pre-treated, relapsed/refractory cancers within 30 days post-infusion in phase I/IIa trials, without causing severe immune-related toxicities. These infused CAR-NK cells demonstrated the ability to expand and persist at low levels for up to 12 months in some patients, as confirmed by PCR analysis ²⁵¹. Recently, Marin et al. ²⁶ highlighted the pivotal importance of donor selection in optimizing clinical outcomes for allogeneic cell therapies. Specifically, receiving CAR-NK cells from a CB unit with nucleated red blood cells ≤8 × 10 ⁷ and a collectionto-cryopreservation time ≤24 h was the strongest predictor of superior outcomes. In their phase 1/2 trial of CB-derived NK cells expressing an anti-CD19 CAR and interleukin-15 (CAR19/IL-15) in 37 patients with CD19 ⁺ B cell malignancies, the overall response rates (ORRs) at days 30 and 100 were both 48.6%. One-year overall survival and progression-free survival were 68% and 32%, respectively, with no evidence of CRS or GvHD. Patients who achieved overall response had higher levels and longer persistence of CAR-NK cells (up to 12 months) ²⁶. These findings highlight the clinical utility of allogeneic PB- and UCB-derived NK cells, particularly in hematologic malignancies. However, opportunities for repeat dosing remain restricted, their off-the-shelf allogeneic nature is not fully realized, and scalability is constrained by donor dependence. These limitations are driving the development of next-generation platforms. The clinical utility of NK cells has further advanced with the development of iPSC-derived engineered NK cells (Table 1). Several studies have demonstrated the successful generation of clinical-grade hPSC-derived NK cells ^248,249. Most clinical trials involving iPSC-derived NK cells are in early stage (phase 1/2a), evaluating safety and feasibility with few trials focusing on efficacy. Fate Therapeutics has pioneered the development of multiple iPSC-derived NK cell products, including both non-engineered and engineered candidates such as FT500 (non-engineer NK cells with and without co-administration of anti-PD1 or anti-PDL1 checkpoint inhibitor antibodies), FT516 (expressing a novel high-affinity, non-cleavable CD16a Fc receptor for augmented ADCC), FT522 (expressing a novel alloimmune defense receptor, which is designed to target activated host immune cells expressing 4–1BB and uniquely enable effector cell proliferation and functional persistence in an allogeneic host immune system), FT538 (expressing a novel high-affinity, non-cleavable CD16a Fc receptor, IL-15/IL-15 receptor fusion protein that promotes NK cell persistence, and knockout of CD38 to mitigate NK cell fratricide by CD38-directed monoclonal antibodies), FT576 (a modified version of FT538 engineered to include a B cell maturation antigen [BCMA] CAR), and FT596 (expressing a novel high-affinity, non-cleavable CD16a Fc receptor, IL-15/IL-15 receptor fusion protein that promotes NK cell persistence, and a CD19 CAR) ⁶⁷. A first-in-human phase I trial investigating FT596, an iPSC-derived CAR-NK targeting CD19 and incorporating a high-affinity, non-cleavable CD16 receptor (hnCD16), was recently reported in heavily pretreated patients with relapsed or refractory CD19⁺ B-cell lymphoma. This phase I study demonstrated effective dual targeting of both CD19 and CD20 via use of the CD20 monoclonal antibody rituximab to enable killing of tumor cells that may lose CD19-expression. This study also found a favorable safety profile with no instances of GvHD, CRS, or ICANS ²⁵⁰. Over 50% of treated patients with refractory B cell lymphomas achieved a complete or partial remission, with some remissions lasting over a year. This includes patients who had previously relapsed after autologous CAR-T cell treatment. However, FT596 demonstrated shorter persistence compared with the 12-month durability reported in some patients by Marin et al. ²⁶ and Ghobadi et al. ²⁵⁰.

Table 1.

hPSC-derived immune cell products

Product Name	Developer	Cell Type	Characteristics	Applications	Benefits	Drawbacks	Development Stage	Reference
FT500	Fate Therapeutics	iNK cells	Universal, off-the-shelf, non-engineered	Cancer immunotherapy, advanced solid tumor	Scalable, consistent, Off-the-Shelf, well-characterized iPSC-clone	Limited in vivo persistence, No CAR, limited efficacy in solid tumor	Completed Phase I	NCT03841110
FT516	Fate Therapeutics	iNK cells	Enhanced CD16 expression	Hematologic malignancies (AML), COVID-19, advanced solid tumor	Improved ADCC, Off-the-Shelf, well-characterized iPSC-clone	Trial terminated	Phase I (Completed/Termina ted) by Fate Therapeutics	NCT04551885; NCT04363346
FT538	Fate Therapeutics	iNK cells	Triple knock-out, CD16+, IL-15Rα	AML, MM, advanced solid tumor	Enhanced persistence, Off-the-Shelf, well-characterized iPSC-clone	Limited clinical data	Phase I (Completed)	NCT04614636; NCT04714372; NCT05069935
FT576	Fate Therapeutics	iNK cells	BCMA-targeted CAR NK	Multiple Myeloma	Targeted cytotoxicity	Recruitment paused, early stage, limited in vivo persistence	Phase I (Active, not recruiting), evasion of BCMA− cancer cells	NCT05182073
FT596	Fate Therapeutics	iNK cells	CAR-CD19, IL-15, CD16	B-cell malignancies	Multi-functional) targeting both CD19+ and CD19− cancer cells, Off-the-Shelf, well-characterized iPSC-clone	Trial completed, comparatively low persistence than UCB-NK cells	Phase I (Completed)	NCT04245722;
FT819	Fate Therapeutics	iCAR- T cells	CD19-targeted CAR-T	B-cell malignancies, Systemic Lupus Erythematosus	Off-the-shelf CAR-T	Early-stage	Phase I (Active)	NCT04629729; NCT06308978
CNTY-101	Century Therapeutics	iNK cells	CD19-targeted CAR NK	Automimmune disease (Systemic sclerosis)	Allogeneic CAR NK	Early-stage	Phase I	NCT06255028
Shanghai iNKT Trial	Shanghai General Hospital	iNKT cells	Autologous iNKT cells	Advanced Pancreatic cancer	Innate and adaptive function	Early-stage	Phase I	NCT07055568

Beyond oncology, iPSC-derived CAR-NK cells have shown promise in autoimmune disease. A recent phase I study using QN-139b, a dual-targeting CD19/BCMA CAR-NK product, achieved significant B cell depletion and fibrosis reversal in a patient with severe systemic sclerosis. This product was extensively engineered to minimize allo-rejection (via B2M, CIITA, and CD16 disruption (to reduce risk of disease flares in autoimmune patients), and HLA-E/G expression) and enhance persistence and safety (via IL-2 and epidermal growth factor receptor [EGFR]) ²⁵¹. However, these findings are based on a single patient, and broader clinical validation is needed.

In vivo persistence remains a challenge for all allogeneic cell therapies. Compared with PB- or UCB-derived NK cell product, these findings highlight the potential of hiPSC-derived CAR-NK cells as off-the-shelf immunotherapies, mitigating the requirement for donor selection, offering unlimited scalability, comparative safety (e.g., minimal GvHD), improved persistence and in vivo proliferation (via IL-15 signaling), and multi-targeting capabilities through advanced genetic engineering features such as disruption of B2M, CIITA, CD38, and expression of HLA-E, HLA-G, hnCD16, truncated (t) EGFR as a safety switch and CARs (Table 2). Early clinical success in both refractory cancers and autoimmune diseases supports their therapeutic promise. However, their efficacy in solid tumors remains under investigation and may be hindered by TME-mediated immunosuppression. Moreover, the extensive genetic modifications required for these products raise important questions regarding long-term safety and durability of response, emphasizing the need for continued clinical evaluation.

Table 2.

NK cells comparision

Feature	PB/UCB-Derived NK Cells	iPSC-Derived NK Cells
Source & Scalability	Donor-dependent; limited scalability; variability in NK phenotype and function, limited availability for repeat dosing	Unlimited supply; clonal uniformity; GMP-standardized manufacturing, excess amount for repeat dosing
Safety Profile	Excellent safety in early trials; no CRS, GvHD, or ICANS reported	Excellent safety in early trials; no CRS, GvHD, or ICANS reported
Persistence	Enhanced in vivo persistence via IL-15/IL-15R fusion and other engineering strategies	Enhanced in vivo persistence via IL-15/IL-15R fusion and other engineering strategies; engineered hypoimmune NK cells have minimal allo-rejection
Clinical Efficacy	High response rates in refractory hematologic malignancies	High response rates in refractory hematologic malignancies; no big difference in vivo
Engineering Flexibility	Limited (CAR introduction feasible but complex),	Extensive: CARs, hnCD16 for ADCC, CD38 knockout, alloimmune defense receptor, HLA-E/G expression
Manufacturing Complexity	Requires donor screening and cryopreservation optimization	Fully standardized, requires cryopreservation as an off-the-shelf production

T Cells.

To improve accessibility and overcome donor limitations, gene-editing technologies such as CRISPR/Cas9 have been used to disrupt the TRAC locus in allogeneic PB-derived T cells, reducing GvHD risk while improving CAR expression and tumor-killing potency ²⁵². However, the scope is limited due to the limited product expansion and persistence in vivo along with the BCMA antigen loss in relapsed patients ²⁵³. Allogene Therapeutics has pioneered clinical development with phase I trials evaluating HLA-unmatched allogeneic CAR-T cells targeting CD19 in patients with relapsed/refractory large B-cell lymphoma (LBCL) ^254,255. These studies report an ORR of 58% and a complete response (CR) rate of 42% with a median CR duration of 23.1 months. The safety profile was manageable, with hematologic toxicities being the most common adverse events. Importantly, no GvHD was observed, reinforcing the potential of allogeneic CAR-T cells as viable alternatives to autologous therapies. However, in relapsed patients, tumors frequently exhibited loss of CD19 expression, as previously reported ²⁵⁶. Similarly, Wugen's phase 2 trial is evaluating WU-CART-007, a CD7-targeted allogeneic CAR-T therapy for relapsed/refractory T-cell acute lymphoblastic leukemia (T-ALL) and lymphoblastic lymphoma (LBL). To minimize self- and host T-cell killing and reduce the risk of GvHD, surface CD7 and TCR-alpha genes were disrupted. The study reported an 82% composite complete remission rate ²⁵⁷, however, the product was associated with grade 3–4 CRS and demonstrated limited in vivo persistence, lasting only 90 days. A phase I trial investigated an allogeneic CAR-T product utilizing Epstein-Barr virus (EBV)-specific T cells genetically modified with CD19-targeting CAR, confirming feasibility and safety with no dose-limiting toxicities; however, the product has limited post-infusion expansion and persistence ²⁵⁸. A phase I dose escalation study of a CD70-targeted CRISPR/Cas9-engineered TCR-deficient allogeneic CAR-T product (CTX130) demonstrated efficacy in relapsed/refractory T-cell malignancies with CRS occurring in 67% of patients, mostly Grade 1–2. Pharmacokinetic analysis showed rapid expansion by Day 7, followed by peripheral clearance by day 28. Notably, CD70-targeted allogeneic CAR-T cells exhibited shorter persistence compared to autologous CD19 CAR-T cells in circulation ²⁵⁹.

hiPSCs also offer a promising platform for generating allogeneic CAR-T cells. Fate Therapeutics introduced FT819, the first iPSC-derived CAR-T product with a highly scalable manufacturing process (>100,000 fold expansion), featuring TCR-deficient CD19-targeting T cells tested in patients with relapsed/refractory B-cell lymphomas (BCLs), B-CLL, and B-cell acute lymphoblastic leukemia (B-ALL) ²¹⁰ (Table 1). No dose-limiting toxicities, GvHD, or ICANS were reported, and CRS was limited to Grade ≤2 in three patients. However, this report did not provide details regarding the clinical response and in vivo persistence of the product.

Preclinical studies have demonstrated the feasibility of engineering hypoimmunogenic cells to enhance immune compatibility. For instance, targeted disruption of HLA-A and HLA-B while retaining HLA-C protected CD43⁺ hematopoietic cells from both allogeneic NK and T cell-mediated rejection, highlighting a strategy for generating hypoimmunogenic iPSC-derived blood cells ¹⁷⁹. Building on this concept, another study showed that T cells derived from genetically engineered iPSCs, lacking B2M, CIITA, CD155 (a ligand of NK cell-activating receptor DNAX accessory molecule-1 [DNAM-1]), and expressing single-chain-trimer HLA-E molecule largely evaded recognition by NKG2A⁺, DNAM-1⁺ NK cells, and CD4⁺/CD8⁺ T cells, while maintaining antitumor potency ⁷⁵. These advances have spurred clinical translation: BrightPath Biotherapeutics is conducting phase I trials of iPSC-derived CAR-iNKT cells targeting CD19 or HER2 ²¹¹, and BeiGene has reported preclinical success with iPSC-derived CAR-γδ T cells incorporating combinatorial gene knockouts to improve durability and antitumor efficacy ⁶¹. These findings underscore the therapeutic potential of gene-edited allogeneic CAR-T cells as scalable, off-the-shelf immunotherapies for hematologic malignancies. In PB- and UCB-derived CAR-T cells, gene-editing strategies such as TRAC locus disruption and TCR knockout have been employed to reduce GvHD risk and enhance CAR expression and cytotoxicity. Clinical trials using these products such as Allogene’s CD19-targeted CAR-T and Wugen’s CD7-targeted WU-CART-007 have reported encouraging response rates and manageable safety profiles, with no GvHD observed. However, PB- and UCB-derived products have limited capacity for extensive genetic engineering at the single-cell level, and their scalability remains constrained by restricted donor cell availability.

By contrast, iPSC-derived CAR-T cells offer additional advantages, including highly scalable manufacturing (e.g., >100,000-fold expansion in FT819), TCR-deficiency, and hypoimmunogenic engineering to evade host immune recognition (e.g., disruption of B2M, CIITA, and expression of HLA-E/G). These cells have demonstrated favorable safety profiles with minimal CRS, no GvHD, and no ICANS in early trials. Additionally, novel iPSC-derived T cell-subsets such as CAR-γδ T, CAR-iNKT and more recently MAIT cells are being explored for enhanced durability and anti-tumor efficacy. Despite these advances, challenges remain across platforms, including complex genetic engineering, maintenance of CAR expression in iPSC-derived T cells, limited persistence, and uncertain long-term efficacy particularly in solid tumors where TME-mediated immunosuppression may hinder therapeutic performance. Realizing the full therapeutic potential of PB/UCB- and iPSC-derived CAR-T cells will require continued innovation in cell engineering, manufacturing, and clinical trial design to address current limitations and expand their applicability across cancer types.

Concluding Remarks

While allogeneic CAR-T cells overcome certain logistical challenges associated with autologous approaches, they still face issues such as limited persistence and toxicity risks. Alternatively, allogeneic CAR-NK cells have not demonstrated GvHD and toxicity from use of these cells appears to be more limited ^{24,26,242,243}, although their long-term efficacy (over a year) remains to be established. Recent clinical trials demonstrate comparable clinical response rates of allogeneic CAR-NK cells to allogeneic CAR-T cells but the former showed lower rates of severe adverse effects such as GvHD, CRS, and ICANS in allogeneic settings ^{7,24,26,243,254,259,260}. More recently, autologous CAR-Ms have demonstrated encouraging activity against solid tumor without causing CRS ^31,34. However, expansion and maintenance of CAR-Ms remain challenging, and the safety and efficacy of allogeneic CAR-based therapies await validation in clinical trials. By contrast to PB- or UCB-derived counterparts, iPSC-derived immune cells offer a scalable and genetically adaptable solution, enabling the generation of standardized allogeneic immune cell products with potentially higher clinical compatibility.

Current Challenges and Future directions

The derivation of immune cells from iPSCs represents a promising frontier in next-generation cell-based immunotherapies. Early studies indicate that iPSC-derived immune cells share functional and phenotypic characteristics with their PB- or UCB-derived counterparts. However, comprehensive phenotypic and functional comparisons across immune subtypes and disease models remain limited.

hiPSC-derived immune cells have already addressed several challenges including immune rejection, genetic modification at the single-cell level, indefinite expansion of CAR-modified cells, off-the-shelf production, and logistical concerns. Preclinical and clinical studies have demonstrated that genetically modified CAR-T cells derived from hiPSCs can avoid GvHD, CRS, and neurotoxicity, mitigating host compatibility issues. Recent efforts have focused on developing hiPSC-derived macrophages, NK cells, γδ T cells, iNKT cells, and MAIT cells, all of which exhibit innate or non-MHC-restricted activity reducing risks associated with CAR-T therapies ^{32,61,65,66,211,250,261,262}. Notably, hiPSC-derived NK cells overcome several limitations associated with hiPSC-derived CAR-T cells, such as protracted time for differentiation, CAR persistence, and multiple genetic edits to avoid GvHD. In addition, NK cells can mediate anti-tumor killing independent of antigen presentation ^{73,107,114,263–265}. Preclinical and clinical studies have reported an improved anti-tumor efficacy of autologous CAR-Ms against solid tumors without obvious CRS; however, GvHD in allogeneic settings remain to be explored. Despite significant progress, several challenges persist in the development of iPSC-derived immune cells, including macrophages, NK cells, and T cell subsets. Key issues involve improving differentiation efficiency, achieving scalability, ensuring stable CAR expression, and enhancing in vivo persistence and expansion (potentially through targeted gene knockouts, mutations, and/or exhaustion). Furthermore, the use of lentiviral and retroviral vectors poses risks of insertional oncogenesis, as highlighted by recent reports in patients receiving genetically modified cells ^266,267. These risks can be mitigated by site-specific insertion of the entire CAR cassette into a defined genomic locus via homologous recombination, followed by Southern blot screening and whole-genome sequencing, a strategy feasible at the single-cell level in iPSCs ^{53,97,268,269}. Additionally, identifying optimal signaling domains for CAR constructs tailored to each immune cell type remains an active area of investigation. Future studies should focus on overcoming tumor immune evasion, optimizing function, and enabling large-scale production to establish iPSC-derived CAR-innate immune cells as an attractive alternative to present-day autologous CAR-T therapies.