γδT cells are the first T cell lineage to develop in the thymus and can be observed in humans as early as 12.5 weeks of gestational age. However, once generated, these cells will expand and mature extrathymically, and their gene repertoire changes in response to age (33). γδT cells derive their name from their TCRs, which are made up of gamma and delta chains. Like αβT cells, γδT cells undergo somatic V(D)J rearrangement, a process that generates diverse TCRs to respond to a wide range of antigens (34). However, in contrast to αβTCRs, γδTCRs allow cross-reactivity with multiple ligands and each combination is associated with different functional avidities (35). Despite the fact that V(D)J rearrangement of γδT cells generates less diversity than αβT cells, TCR δ chains have a higher potential of diversity at the complementarity-determining region 3 (CDR3) junction and can provide information on a person’s unique history of infection (33, 36, 37).

γδT cells target real and perceived immunological insults through the production and release of soluble factors. One example of this is when γδT cells recognize pathogen specific antibodies and stress-induced antigens. In response, γδT cells will produce Th1 cytokines including IFN-γ and TNF-α. Subsequently, γδT cells also release cytotoxic granules containing perforin and granzyme, further promoting pathogen degradation (38). Additionally, there is evidence in literature suggesting that Vγ9Vδ2 cells–a subset of γδT cells (discussed in the next section) can act as sensors of a dysregulated isoprenoid metabolism that target specifically cancer cells (39). Moreover, several recent studies have indicated that different subsets of γδT cells may have remarkably different functions in targeting tumor cells (40–43). Therefore, it is important to understand the structure and subsets of γδT cells, which we describe in the next section.

This section will focus on two main subsets of γδT cells: Vδ1 and Vγ9Vδ2.

The successful clinical application of γδT cell-based immunotherapy must address several challenges, starting with the selection of appropriate sources ( Figure 2 ). Inconsistent effects of autologous γδT cells have prompted investigators to design standardized cell products. Because HLA-matching is not required, fully allogeneic mismatched or haplo-identical γδT cells sourced from healthy donors have emerged as an appealing approach with a commendable safety profile (75, 76). A thorough investigation into the donor’s infection history can also benefit patient outcomes when used as a screening criterion. For instance, the reactivation of cytomegalovirus (CMV) in patients receiving HSCT can potentially induce the expansion of Vδ2^neg γδT cell clones, which exhibit dual reactivity to CMV and acute myeloid leukemia (AML) (77–79). Another challenge lies in determining which γδT cell subset will be more effective for a specific tumor considering their differing characteristics, particularly their chemotaxis ability and tumor cytotoxicity. Up to now, the main sources of γδT cells include cord blood, peripheral blood, skin, and inducible pluripotent stem cells (iPSCs).

The developmental trajectory of γδT cells reveals that Vδ1+ cells constitute the predominant population (approximately 50%) of γδT cells in cord blood at birth, while Vδ2+ cells typically represent 25% (80). Over time, Vγ9Vδ2 T cells emerge as the predominant subset (over 75%) of the γδT cell population in peripheral blood by adulthood, with less than 10% being Vδ1+ (80). Therefore, cord blood has been explored for its predominant expansion of Vδ1+ cells, or occasional viable expansion of Vδ2+ cells (81–83). However, there are several challenges associated with the in vitro expansion of γδT cells from cord blood, including a low number of γδT cells (less than 1% of cord blood lymphocytes), phenotypically and functionally immature γδT cells, and a poor response to IL-2 and phosphoantigen stimulation (80). In contrast, γδT cells isolated from peripheral blood mononuclear cells (PBMCs) are predominantly Vγ9Vδ2 (84). Due to their relative convenience and availability, PBMCs provide easy and stable access for expanding Vγ9Vδ2 T cells and viable Vδ1+ cells such as Delta One T (DOT) cells. Additionally, owing to natural tissue tropism of Vδ1+ cells, human tissues such as skin also provide an alternative source of Vδ1+ cells through enzymatic digestion or other methods (85, 86). Despite the roles of skin γδT cells in the cutaneous malignances such as melanoma, complex skin γδT cell subsets necessitate a thorough investigation for therapeutic strategies (87).

In addition to γδT cells derived from donors, these cells can also be generated from iPSCs (88, 89). Two companies, Century Therapeutics and CytoMed Therapeutics, have developed platforms that enrich γδT cells from healthy donor leukapheresis and then reprogram them into T cell-derived iPSCs (TiPSCs). TiPSCs are engineered with CAR expression, followed by directed differentiation into γδ CAR-T cells ( Figure 2 ) (90). During this process, the genome characterization of a single CAR-TiPSC clone enables the production of a highly uniform clonal γδ CAR-T cell bank (> 95% CAR expression) and minimal DNA mutation caused by engineering (90, 91). This off-the-shelf platform provides an appealing source of γδT cells with several benefits: overcoming quantitative limitations of γδT, reducing the wait time for ex vivo expansion of γδT cells, and not relying on the ex vivo expansion efficiency of PBMC-derived γδT cells (92). Importantly, TiPSC-derived γδT cells retain cytotoxicity to solid and blood tumor through both γδTCR and NKG2D (92). However, this complex manufacturing process is time-consuming and needs more evaluation on potential risks.

Overall, most of the research is adopting PBMC and cord blood as the primary source of γδT cells. In contrast, investigations into skin-derived and iPSC-derived γδT cells are still in the preclinical stages. Our current understanding of the migration and colonization of γδT cells in peripheral tissues primarily relies on research conducted in mice. Further studies involving humans will significantly advance our comprehension of tissue-specific γδT cells, potentially expanding the applications of Vδ1+ cells in immunotherapy.

The clinical-scale manufacturing of γδT cells requires robust and highly reproducible expansion methods that meet good manufacturing practice (GMP) standards. Current approaches mainly include cytokine only, synthetic p-Ag and bisphosphonate (BP) stimulation, antibody-based expansion, and feeder cell-based strategies as summarized in Table 1 . Undoubtably, cytokine combinations strategies simplify the manufacturing process but often produce insufficient expansion.

p-Ag or BPs have been recognized as the most established approaches to selectively expand Vδ2+ γδT cells (9). Zoledronic acid (ZOL), a BP, has been widely used to numerically expand Vγ9Vδ2 T cells in vivo and ex vivo. ZOL can be used alone or in combination with IL-2 to achieve these effects (115). ZOL (5 uM) and IL-2 (1000IU/ml) administration over 14 days has been reported to initiate an over 4000-fold proliferation and expansion of γδT cells (mainly Vγ9Vδ2) from PBMCs of both healthy donors and patients with advanced non-small cell lung cancer (116). However, the expansion folds and purities of γδT cells vary in different published results.

Current protocols for expanding Vδ1+ T cells in vitro primarily rely on mitogenic plant lectins such as phytohemagglutinin (PHA) or concanavalin-A (ConA), which induce AICD in Vγ9Vδ2 T cells (93, 117). To transition from the laboratory to the clinic, more efforts have been made to avoid potentially hazardous components. Almeida et al. first developed a clinical-grade two-step method through combination of cytokines (IL-1β, IL-4, IL-21, and IFN-γ) and anti-CD3 mAb (clone: OKT-3) to achieve the expansion of Vδ1+ T cells (94). This method enables large-scale expansion (up to 2,000-fold) of Vδ1+ T cells known as DOT cells (94). GDX012, based on DOT cells, has been granted orphan drug designation by FDA for AML treatment and is currently undergoing evaluation in a phase I trial (NCT05001451). Recently, Ferry et al. also apply only anti-CD3 mAb and IL-15 to stimulate αβTCR- and CD56-depleted PBMC, resulting in robust Vδ1 cell expansion (97).

The feeder cell-based method utilizing artificial antigen-presenting cells (aAPCs) has been explored to provide γδT cells with a sustained activation and costimulation signal. K562, a human chronic erythroleukemic cell line lacking MHC expression, is primarily used as aAPCs. These cells are engineered with costimulatory molecules (like CD80, CD86, CD137) and antigens (e.g., CMV antigen-pp65), allowing for the targeted expansion of specific γδT cell subsets (108, 118). Deniger et al. first activated and propagated polyclonal γδT cells utilizing K562-based aAPCs as irradiated feeders (108, 118). This method requires the additional labor-intensive manufacturing process of culturing feeder cells, yet it mitigates the AICD effects in γδT cells associated with prolonged antigen exposure. Additionally, methods of removing all residual feeder cells before infusion remains a hurdle to clinical implementation of this approach. To address this, several solutions have been proposed, such as gamma-irradiation of aAPCs and the transduction of aAPCs with an inducible suicide gene (107). The ex vivo aAPC expanded donor-derived γδT cells are under evaluation of safety and cell dose in a phase I/II trial (NCT05015426) in patients with high-risk acute leukemia (104).

In the future, efforts should focus more on eliminating the use of xenogeneic serum and feeder cells and integrating GMP/pharmaceutical-grade reagents into the expansion process. An example of such a method is the protocol proposed by Bold et al. in a recently published article, which has shown better outcomes in terms of expansion and purity (119). Further efforts can be directed towards enhancing the rate of γδT cell expansion, optimizing the procedure, and lowering manufacturing costs. Besides assessing quantity, evaluating the quality of expanded γδT cells–such as memory and exhaustion phenotypes, is crucial for maximizing therapeutic efficacy and requires further investigations.

CAR-T therapy, with its potential for HLA-independent tumor antigen recognition, has found its place as a key player in cancer immunotherapy. Traditionally, αβT cells have been the main candidates for CAR development (150). However, despite their effectiveness, these cells present several limitations. They are susceptible to GvHD, can cause severe and potentially lethal toxicities, contribute to the development of cytokine release syndrome (CRS), and pose issues related to antigen escape (150). These challenges have spurred an interest in alternative solutions, with γδT cells showing potential to offset these limitations.

Given the wealth of limitations associated with αβT cells, γδT cells are garnering interest as an alternative for CAR-T therapy. These cells do not instigate GvHD, curb antigen escape resulting in decreased relapse rates, and retain beneficial traits such as a less differentiated phenotype with enhanced antigen presentation capacity (125, 151). With these advantages, γδ CAR-T cells may have the potential to overcome the obstacles that have historically troubled conventional αβ CAR-T therapy.

The primary goal of CAR design is producing extracellular domains capable of targeting unique tumor cell antigens while sparing healthy tissues (150, 152). Owing to the deficit of tumor-specific antigens, lineage-specific antigens have been a key focus in CAR T cell development. Under investigation are promising candidates like CD19 (153, 154), GD2 (130, 155), GPC-3 (128), CD123 (131, 156), CD5, CEA, CD20 (10), B7H3 B7H3 (157), and PSCA (151) ( Table 2 ). While CD19-targeting CAR-T products have earned FDA approval for treating B-cell lymphoma and leukemia, they carry risks, like CRS, neurotoxicity, and B-cell aplasia, primarily due to on-target off-tumor toxicities (152, 153). Interestingly, γδ anti-CD19 CAR-T cells have been reported to produce fewer inflammatory cytokines compared to their αβ counterparts, suggesting a potential decrease in cytokine-mediated side effects (90).

However, the optimization of CAR for highly specific antigen recognition remains vital. Recent studies have investigated the incorporation of ligands like NKG2DL and inhibitory receptor programmed cell death ligand 1 (PD-L1) into CAR constructs to improve safety or efficacy (158). Some attempts have even added T cell antigen coupling (TAC) components to γδT cells, thereby redirecting them to target tumors with reduced off-tumor toxicity compared to conventional CAR-T cells (159, 160) ( Figure 3D ). Adicet Bio is working on CAR designs that target tumor intracellular antigens using their TCR-Like monoclonal antibodies (TCRLs) technology (91).

Clinical trials are in progress for CAR-γδT cells targeting various antigens such as CD19 (NCT02656147, NCT05554939), CD20, NKG2DL, CD7, CD33, CD28, and CD123 ( Table 3 ). While many of these trials have yet to disclose their results, some promising preliminary findings have been reported. For instance, Adicet Bio, Inc. is testing an allogeneic CD20 CAR+ Vδ1 γδT cell called ADI-001, designed for patients with refractory B cell malignancies (NCT04735471). Their early report shows a 71% overall response rate and 63% complete response rate among patients with aggressive B-Cell non-Hodgkin lymphoma, all without the presentation of GvHD (91).

One challenge with CAR-T therapy is its potential ineffectiveness in tumors exhibiting heterogeneity or low antigen expression. Dual-specific CARs, which target two antigens concurrently, are proposed as a potential solution, although this requires further investigation (161). Other research focuses on fine-tuning other CAR components, including the intracellular signaling and transmembrane domain, with construction of Boolean logic gates for combinatorial antigen sensing. Balancing the DNA length of dual-CAR plasmids and transduction efficiency necessitates further study.

Recent innovative extracellular designs aimed at enhancing safety also include the development of ON/OFF switches like the masked CAR. Here, the antigen-binding site of CAR is coupled with a masking peptide through a protease-sensitive linker. Activation of masked CAR-T cells occurs when tumor microenvironment proteases cleave this linker, causing the masking peptide to detach, and revealing the antigen-binding site (162) ( Figure 3C ). In essence, this provides a level of control, reducing risks associated with unregulated CAR-T activation (162).

To sum up, while there are promising advancements in the development of γδT cell-based CAR-T therapies, it is critical to continue fine-tuning these interventions for increasing specificity and safety. A combination of innovative design strategies and rigorous clinical trials may bring forth the next generation of cancer immunotherapies. The hope is for these novel treatments to cure more patients, more reliably, with fewer side effects, revolutionizing the approach to cancer treatment.

Cell engagers and bispecific antibodies have become an increasingly attractive immunotherapeutic method for enhancing the anti-cancer activity of γδT cells. Bispecific T cell engagers (bsTCEs) are specially designed antibodies, each having two separate binding areas aimed at individual components like tumor-associated antigens (TAAs) and the TCR complex (Vδ2 or Vγ9) (176). The flexibility of bsTCEs allows for varied applications, such as MHC-independent targeting of TAAs by γδT cells, immune checkpoint modulation, and controlling inflammatory and other signaling pathways (176). These functionalities provide several unique advantages, including their small molecular size and high versatility, eliminating the need for additional co-stimulatory signals for T cell activation, low picomolar range for the half-maximal effective concentration (EC50), effectiveness against both blood-borne and solid tumors, excellent safety profile, and efficient and cost-effective production (177). Most frequently, cell engagers incorporate a fragment-based design or lgG/lgG-like formats (136, 137). Fragment-based designs principally modify constructs such as scFv (178), Fab (135), or single-domain antibodies (sdAbs, also known as V_HH) (176) into their binding regions ( Figures 4A-C ). sdAbs, originating from the variable domain of heavy-chain-only antibodies, have attracted attention because of their unique features, including small size, target specificity, and minor immunogenicity (179). Currently, cell engagers can be applied both as stand-alone therapies and in partnership with allogeneic γδT cells to generate readily available products.

The first CD3-targeting bsTCEs, exemplified by blinatumomab and Tebentafusp, yielded significant positive outcomes in B-cell malignancy and melanoma patients during clinical trials (NCT03070392) (180). However, adverse effects like CRS and immune effector cell-associated neurotoxicity syndrome (ICANS) constrained their clinical usage (177). Further, CD3-targeting bsTCEs may unintentionally activate other CD3+ T cell subsets, which could depress tumor-specific immune responses (137). As a category of innate T cells, γδT cells present a logical choice for engagement to reduce CRS and off-tumor toxicity.

The successful usage of bsTCEs in LAVA Therapeutics’ Gammabody platform, employing tandem single-domain antibodies (V_HHs) ( Figure 4D ), exemplifies their potential. These include EGFR-Vδ2, CD1d-Vδ2, CD40-Vδ2, and PSMA-Vδ2 bsTCEs ( Table 3 ). EGFR-Vδ2 bsTCEs have displayed compelling activation of Vγ9Vδ2 T cells which induce cytotoxicity against EGFR+ tumor cells (137, 154). The CD1d-Vδ2 bsTCE, or LAVA-051, has shown anti-tumor potential against hematological malignancies expressing CD1d in preclinical models (176). Its specificity for NKT and Vγ9Vδ2-T cells, alongside low-nanomolar range EC50 values in vitro, further demonstrates its potential (176).

Bispecific antibodies (bsAbs) comprise a class of engineered antibodies with two distinct binding sites, setting them apart from traditional antibodies (181) ( Figures 4D-F ). These antibodies, as exemplified by anti-Vγ9/CD123 bsAbs, selectively rally Vγ9+ γδT cells, promoting cell conjugate formation between γδT cells and AML cells (136). As such, these cell engagers can enhance Vγ9Vδ2+ T cell cytotoxicity against B-cell lymphoma, particularly when accompanied by a co-stimulatory signal pair (178). ImCheck Therapeutics’ humanized anti-BTN3A antibody, ICT01, serves as another example. It operates by recognizing three distinct BTN3A forms and prompting their activated conformation, thereby selectively activating Vγ9Vδ2 T cells in an antigen-independent manner (182).

In a phase I/II clinical trial, ICT01 showed tolerable safety profile and increased infiltration of Vγ9Vδ2 T cells into tumor tissue in patients with advanced solid tumors (NCT04243499). ( Table 3 ) Besides, LAVA-1207 (PSMA-Vδ2 bsTCEs) has shown a favorable safety profile and clinical symptom improvement (decreased PSA level) in a Phase 1/2a clinical trial involving metastatic castration-resistant prostate cancer (mCRPC) patients (N=20, NCT05369000) (137).

Notably, as cell engagers depend on the activation and migration of the patient’s inherent γδT cell pool, initial Vγ9δ2T cell counts could be a useful predictor for clinical outcomes. Take, for instance, a melanoma patient with a high baseline count of circulating Vγ9Vδ2 T cells who showed considerable tumor infiltration of Vγ9+ T cells post ICT01 administration (182). Cell engagers can also be combined with γδT cell-based therapies to develop easily available TAA-targeting γδT cell products (148).

Acepodia’s technology, for instance, conjugates antibodies to cells to create products like ACE1831, which is the CD20-targeting γδT cells (148). This product is currently under phase I trial for patients with relapsed/refractory B-cell lymphomas (NCT05653271). Other products, ACE2016 (EGFR-targeting γδT) and ACE1708 (PD-L1-targeting γδT), are in the preclinical exploratory stage (183).

In conclusion, while cell engagers and bispecific antibodies present significant potential compared to CAR-T therapy, their definitive superiority is yet to be determined. Like CAR-T therapy, cell engagers also encounter hurdles such as immune escape owing to loss of target antigen expression and an immunosuppressive tumor microenvironment. Further research is needed to modify cell engagers specifically for γδT cells, paving the way for effective treatments in the future.

Harnessing natural receptors through the engineering or transfer of T cell receptors (TCRs) serves as an alternative approach to the use of synthetic ones. The transduction of cancer-specific TCRs is an appealing strategy for generating large volumes of readily available, antigen-specific T cells. Transferring cancer-specific αβTCR engenders T cell specificity, simplifying procedures compared to isolating specific T cell subsets. However, the transgenic transfer of αβTCRs to other αβT cells runs the risk of triggering TCR competition and mispairing. Recognizing these limitations, γδT cells are appreciated as safe and ideal carriers for antigen-specific effector cells because TCR-α and -β chains can’t pair with TCR-γ and -δ chains (138, 184). To produce cytotoxic γδT cells capable of attacking tumor cells and secreting cytokines via αβ and γδTCR-dependent activity, one can isolate tumor antigen-specific αβ CD8+ cytotoxic T lymphocytes and clone their TCR αβ genes (138) ( Figure 4I ). However, a notable reduction in γδTCR expression post αβTCR transduction was observed, likely due to competition for limited CD3 molecules (138).

Van der Veken et al. demonstrated that αβTCR -transduced γδT cells display sustained in vivo endurance and can elicit a recall response (139, 184). More so, infusing αβTCRs from invariant natural killer T (iNKT) cells into γδT cells can create bi-potential T cells with NKT cell functionality (143) ( Figure 4J ). Other research endeavors are concentrated on transferring γδTCR to αβT cells to leverage the superior understanding of their effects and memory function mechanisms (185). One product, GDT002, which contains Vγ9Vδ2TCR-expressing αβT cells, allows αβT cells to detect augmented phosphoantigens in stressed or malignant cells (185). An ongoing phase 1/2 study is investigating GDT002’s safety and tolerability in patients with multiple myeloma. Furthermore, strategies for engineering TCRγδ involve fusing with single-chain variable fragments (scFv) or Fab fragments from antibodies. For example, one study used CRISPR/Cas9 to fuse an anti-PD-L1 scFv to the TCRγ or δ chain in activated γδT cells, creating scFv-γ/δ-TCRγδ cells that showcased anti-tumor capacity akin to traditional CAR-T cells (186) ( Figure 4G ). Alternatively, the Fab domain of an antibody can be connected to the C-terminal signaling domain of the γ and δ chains of the TCR, creating an antibody-TCR construct (187) ( Figure 4H ). The use of the TCR alongside endogenous costimulatory molecules can lower co-stimulation input compared to CAR constructs, thus diminishing cytokine release and mitigating the exhaustion phenotype (187). Anti-CD19 Fab – TCR-γδT cells or ET019003, for instance, have displayed similar anti-tumor actions against B-cell lymphoma as CAR-T cells in vivo (187).

Promisingly, a phase I clinical trial (NCT04014894) indicates that, aside from showing agreeable safety profiles, ET019003 has achieved an impressive clinical response rate (87.5%) among patients with relapsed or refractory diffuse large B-cell lymphoma (188). However, TCR gene transduction or engineering research has somewhat stagnated in recent years, possibly due to complex manufacturing processes involved. In summary, the exploration of novel therapeutic approaches incorporating γδT cells continues to expand, with significant potential for future cancer treatment innovations.

CAR-T cell therapy has proven extremely promising for treating hematologic malignancies. However, distinct issues related to the immunosuppressive microenvironment of solid tumors require further refinement and personalization of this approach. A potential solution could be combination therapies that adequately address the complexity of solid malignancies.

The concurrent usage of CAR-T/bsTCE therapies and immune checkpoint inhibitors is recognized as a potentially effective strategy to overcome immune system suppression. Exhaustion status, marked by the upregulation of inhibitory receptors, can potentially compromise the therapeutic efficacy of CAR-T cells (189). In a murine model of bone metastatic prostate cancer, γδ CAR-T cells persisted in the tumor-bearing tibia for approximately 21 days post-infusion. However, these cells exhibited an upregulation of PD-1 expression while simultaneously losing expression of activation markers (151). Consequently, the combination of therapies such as ICT01 and pembrolizumab, an anti-PD1 antibody, exhibited favorable safety profiles in a phase I clinical trial (NCT04243499). This suggests that the co-administration of CAR-T/bsTCE therapy with anti-PD-1/PD-L1 antibodies could potentially boost treatment benefits (151).

Chemotherapy and radiotherapy, owing to their immune-sensitizing attributes, are plausible options for combination therapy with immunotherapy (190). Temozolomide (TMZ), a chemotherapy mainstay for glioblastoma (GBM), transiently heightens the expression of stress-associated antigens such as NKG2DL on tumor cells. Engineering γδT cells to express the methylguanine DNA methyltransferase (MGMT) can thus potentially confer TMZ resistance, enabling the engineered cells to operate efficiently despite the presence of therapeutic concentrations of chemotherapy. The amalgamation of TMZ and MGMT-modified autologous γδT cells, or drug resistant immunotherapy (DRI), showed improved survival outcomes in a model of high-grade gliomas compared to monotherapy (124).

In a phase I clinical trial, INB-200 (an example of DRI) displayed a favorable safety profile, extended progression-free survival (PFS), and presented no dose-limiting toxicities, CRS, or neurotoxicity in glioblastoma multiforme patients (NCT04165941). As a result, autologous DRI- γδT cells (INB-400) have proceeded to a phase II clinical trial, and MGMT-modified allogeneic γδT cells (INB-410) are currently undergoing a phase Ib clinical trial (NCT05664243). The product INB-400/410, developed by IN8bio, has been granted FDA Orphan Drug Designation for the Treatment of Newly Diagnosed Glioblastoma.

As it stands, most approved combination immunotherapies largely rely on a combination of immune checkpoint inhibitors (ICIs) and have emerged as first-line treatments for several major cancer types (191). The future of combination immunotherapies with γδT cells likely extends beyond ICI-based approaches, aiming for control and eradication of established tumors. Further research in this area will be instrumental in harnessing the full therapeutic potential of γδT cells.