Although CAR-αβ T cell therapies have achieved remarkable success in hematologic malignancies, their clinical translation is limited by T cell exhaustion. Exhaustion, a state of functional decline under chronic antigenic stimulation, is characterized by reduced effector function, impaired proliferation, and upregulation of inhibitory receptors such as PD-1, LAG-3, and TIM-3. In contrast, CAR-γδ T cells may follow distinct exhaustion trajectories, attributable to their innate-like properties and adaptive mechanisms in chronic infections. γδ T cells exhibit phenotypic plasticity, maintaining multifunctionality under persistent stimulation rather than rapidly entering terminal exhaustion, offering enhanced persistence and resistance to functional decline in CAR applications (68–70).

Mechanistically, CAR-αβ T cell exhaustion is largely driven by MHC-independent and excessive TCR/CD3ζ signaling, leading to aberrant NFAT and AP-1 activation, induction of TOX and NR4A, and consequent upregulation of inhibitory receptors. Terminal effector memory RA (TEMRA) subsets are particularly susceptible. By contrast, CAR-γδ T cells recognize targets in an HLA-independent manner (e.g., phosphoantigens via BTN3A1/2A1 or stress ligands via NKG2D), reducing MHC-related overactivation. TOX and NR4A are transcription factors that function as “master regulators” of T-cell exhaustion in conventional CAR-αβ T therapy. Native γδ T cells naturally express much lower levels of these factors, helping them avoid the irreversible dysfunctional state seen in chronically stimulated αβ T cells. Their exhaustion is more influenced by tumor microenvironment (TME) factors, including TGF-β, PGE2, and metabolic stress. γδ T cells can also leverage innate NK-like receptor signaling to sustain function, and distinct subsets (Vδ1 vs Vδ2) display varying exhaustion profiles due to differences in tissue residency and proliferative capacity (31, 71–73).

At the signaling level, CAR-αβ T cell activation relies on CD3ζ and co-stimulatory signals (e.g., CD28), primarily engaging the PI3K/AKT pathway, and is susceptible to PD-L1 or TGF-β-mediated inhibition. CAR-γδ T activation, however, can be achieved via γδ TCR single signaling with minimal co-stimulatory dependency and can be further potentiated through BTN3A1/BTN2A1 complexes, conferring superior TME resilience (38, 69, 72, 73). Phenotypic analyses reveal that CAR-γδ T cells (e.g., anti-CD19, anti-GD2) maintain antigen cross-presentation and dual antitumor activity under chronic stimulation, whereas CAR-αβ T cells are prone to early exhaustion and poor persistence. In solid tumor models, δ1^+ γδ T cells exhibit prolonged in vivo persistence and reduced exhaustion. Functional heterogeneity in the TME is modulated by the PD-1/PD-L1 axis, with checkpoint blockade, metabolic modulators, BTN agonists, or genetic engineering further enhancing efficacy (74).

In chronic infections such as TB, HIV, and CMV, γδ T cells demonstrate adaptive features supporting their anti-exhaustion potential in CAR therapies. Unlike easily exhausted αβ T cells, γδ T cells (Vδ1^+, Vγ9Vδ2^+) maintain function via clonal expansion and memory-like responses. In TB, Vγ9Vδ2 T cells secrete IFNγ and IL-17 to clear pathogens while forming memory-like populations resistant to activation-induced cell death (AICD). In HIV, Vδ1^+ T cells lack CCR5 expression, avoiding viral cytotoxicity while producing IFNγ to control replication (68, 69). Functional adaptation and phenotypic plasticity allow γδ T cells to shift from IFNγ-dominant to IL-17-dominant states or maintain NK-like CD8^+ phenotypes to prevent functional decline (38, 68, 69). Non-peptide antigen recognition and innate immune features render γδ T cells more environmentally adaptable, a core advantage in CAR design (73). Moreover, γδ T cells exhibit “adaptive” and “trained” memory characteristics, as seen in BCG- or MMR-induced metabolic remodeling and heterologous stimulus enhancement, indicating cross-pathogen defense potential, though the underlying receptor and metabolic mechanisms require further investigation (25).

Based on these differences, CAR-γδ T cells may circumvent classical exhaustion pathways via infection-like adaptive mechanisms. Optimizing CAR constructs (e.g., incorporating DAP10 or IL-15 expression) can maintain stem-cell-like memory (T_SCM) states, enabling self-renewal and multipotent differentiation while avoiding terminal differentiation and functional decline typical of αβ T cells. In solid tumor models, this strategy enhances tumor infiltration and cytokine release, reduces GVHD risk, and facilitates the development of universal CAR therapies (75).

γδ T cells, particularly the Vδ2 subset (circulating, phosphoantigen-reactive), exhibit remarkable metabolic plasticity within the tumor microenvironment (TME), enabling sustained survival and effector function under nutrient-deprived and hypoxic conditions. Metabolic reprogramming of tumor cells creates competition for nutrients, which suppresses conventional immune cell function. In contrast, γδ T cells, leveraging their innate-like features and MHC-independent (e.g., sensing phosphoantigens and MICA/B), can remain active even in the absence of co-stimulatory signals, whereas αβ T cells rapidly undergo exhaustion under energy-limited conditions (76, 77). γδ T cells can flexibly switch between glycolysis and oxidative phosphorylation (OXPHOS): glycolysis supports IFN-γ-producing γδ T cells (γδIFN) for anti-tumor activity, whereas OXPHOS favors IL-17-producing γδ T cells (γδ17), allowing adaptation to pro-tumor niches. In plain language, conventional αβ CAR-T cells run almost exclusively on glycolysis (rapid sugar burning) and quickly fail when glucose or oxygen is scarce inside solid tumors. In contrast, γδ T cells can switch to mitochondrial respiration (OXPHOS) or even burn fatty acids (FAO), allowing them to stay active in the harsh, nutrient-poor tumor microenvironment. Approximately 49% of genes are differentially expressed between these subtypes, with γδ17 enriched for metabolic and environmental sensing pathways and γδIFN enriched for translation and TCR signaling genes. Molecules such as CD6 and Themis selectively promote γδIFN activation and cytotoxic anti-tumor responses, reflecting independent regulation and functional specialization of γδ T cell subsets (78).

This metabolic flexibility is particularly pronounced in Vδ2 T cells, enabling effective effector function across different energy states. Vδ2 T cells can switch between glycolysis and OXPHOS based on energy demand and, under glucose deprivation, rely on fatty acid oxidation (FAO) for energy (79). Upregulation of amino acid transporters enhances uptake, and de novo synthesis allows survival in arginine- or tryptophan-limited environments (80). Under hypoxia, HIF-1α promotes glycolysis and reduces oxygen consumption, while AMPK activation induces autophagy to maintain energy homeostasis (81). Coordinated mTOR–AMPK signaling optimizes metabolic efficiency and, via the PD-1/PD-L1 axis, fine-tunes functional recovery. Additionally, Vδ2 T cells maintain cytotoxicity under extreme nutrient deprivation through cytokine signaling and stress ligand recognition (82, 83).

By contrast, αβ T cells often experience mitochondrial damage and ROS accumulation under hypoxia and nutrient competition, leading to functional impairment (81). γδ T cells, however, can shift to OXPHOS or lipid metabolism under low glucose and continue proliferating and responding under hypoxic conditions (84). Recognition of stress ligands and metabolic intermediates (e.g., IPP, HMBPP) via receptors such as NKG2D further enhances tumor infiltration and persistence (76, 79, 83). This metabolic advantage provides a foundational rationale for CAR-γδ T cell therapy: glycolysis supplementation can boost γδIFN function, while targeting lipid/IDO/S2 pathways may restrain pro-tumor polarization. Agents such as nitrogen-containing bisphosphonates (N-BPs) and HIF-1α/COX-2 inhibitors further enhance Vδ2 T cell survival and therapeutic potential within nutrient-depleted TMEs (43, 85). The key differences and core advantages of CAR-γδ T cells compared to traditional CAR-αβ T cells are summarized in Table 1.

Native γδ T cells are intrinsically more exhaustion-resistant than αβ T cells due to their MHC-independent activation, flexible metabolism, and epigenetic programs observed in chronic infections and tumor surveillance. However, strong CD3ζ-based CAR signaling can impose tonic stimulation and reactivate TOX/NR4A-driven exhaustion pathways, partially overriding this natural resilience. Consequently, the proposed anti-exhaustion engineering strategies (DAP10-based signaling, γδTCR gating, IL-15 co-expression, checkpoint knockout) are designed not to create resistance de novo, but to preserve the intrinsic exhaustion-resistant phenotype of γδ T cells that would otherwise be compromised by conventional CAR constructs.

Within the programmable innate immune sentinel framework, CAR-γδ T cells leverage MHC-independent and TME remodeling to overcome the limitations faced by conventional CAR-αβ T cells in solid tumors. The TME is enriched with regulatory T cells (Tregs), M2-polarized tumor-associated macrophages (TAMs), and immunosuppressive networks, which typically dampen αβ T cell effector function. γδ T cells’ innate immune features allow them to resist these inhibitory mechanisms. First, γδTCRs directly recognize stress-associated ligands on tumor cells (e.g., phosphoantigens, lipid antigens), bypassing MHC-dependent suppression mediated by IL-10 and TGF-β, thus maintaining non-exhausted effector function in the TME (41, 69, 89). Second, γδ T cells can identify MICA/B expressed on M2 TAMs via NKG2D, enabling direct cytotoxicity and disruption of the tumor-promoting milieu (70, 90). Third, IFN-γ secreted by γδ T cells can polarize TAMs toward an M1 phenotype, activate myeloid immune cells, and restore anti-tumor immunity (39, 91, 92). Collectively, CAR-γδ T cells can reprogram the TME through innate mechanisms without additional “armoring,” offering a novel pathway for solid tumor immunotherapy (20, 93).

Tumor heterogeneity remains a major obstacle to durable immunotherapy responses. Conventional CAR-αβ T cells rely on single-antigen recognition and are vulnerable to antigen loss or downregulation. CAR-γδ T cells employ a dual mechanism—”engineered targeting + innate recognition”—to achieve broader tumor coverage. The CAR structure provides high specificity toward tumor antigens (e.g., HER2, GD2), enabling precise elimination of major tumor cell subsets (90). Concurrently, γδ T cells allow MHC-independent recognition of antigen-negative or stressed tumor cells. This complementary dual mechanism mitigates immune escape and reduces the risk of relapse (75, 94, 95). Preclinical models in prostate, breast, and bone metastases have confirmed superior broad-spectrum anti-tumor potential and simplified engineering feasibility compared with multi-target CAR-αβ T cells (90, 93, 96, 97).

Recent studies further validate the safety and efficacy of CAR-γδ T cells. mRNA-engineered CD5 CAR-γδ T cells achieved >70% cytolysis and enhanced IFN-γ secretion in acute lymphoblastic leukemia, while significantly reducing CRS and “fratricide” risks (98). In oral carcinoma models, mixed CAR-γδ T cells reached 60-85% lysis at low E:T ratios, outperforming conventional CAR-T and eliciting lower pro-inflammatory responses (99). In osteosarcoma mouse models, engineered CAR-γδ T cells achieved >50% tumor growth inhibition without GvHD or off-target toxicity. Clinically, ADI-001 (anti-CD20 CAR-γδ T) Phase I trial (NCT04735471) demonstrated preliminary responses in relapsed/refractory B-cell malignancies with minimal GvHD risk; ADI-270 (Vδ1 CAR-γδ T) Phase I/II trial showed favorable safety in renal cancer; and NKG2DL-targeted CAR-γδ T Phase I studies reported early efficacy signals (21). These results underscore CAR-γδ T cells as promising allogeneic candidates for solid tumor immunotherapy due to their MHC-independent and low GvHD risk (100).

CAR-γδ T cells possess intrinsic surveillance and repair functions, enabling dual regulation - “clearance of abnormal cells and restoration of tissue homeostasis.” Many chronic conditions, such as liver fibrosis, pulmonary fibrosis, and age-related tissue degeneration, share pathologies of abnormal cell accumulation and persistent inflammation (101–104). γδ T cells sense stress ligands (e.g., MICA/B, phosphoantigens) via NKG2D or γδTCR and rapidly eliminate aberrant fibroblasts or senescent cells in an MHC-independent manner. Engineering CARs to target fibrotic markers (e.g., FAP) enhances clearance specificity (69, 102, 105). Furthermore, secretion of IFN-γ and TNF-α remodels the inflammatory environment, promoting epithelial regeneration and ECM degradation, thereby reversing pathological fibrosis or tissue degeneration (106–109). These features position CAR-γδ T cells as a promising platform for non-oncologic therapies.

Effective anti-fibrotic therapy requires disrupting the “injury–inflammation–fibrotic deposition” cycle (110). Here, we put forward a hypothetical ‘Target–Eliminate–Repair’ tri-phasic model as a conceptual framework for future CAR-γδ T cell applications in fibrosis (Figure 2). While individual components (e.g., FAP targeting, IFN-γ–mediated remodeling, and reparative cytokine secretion) have been observed in related cell therapies, integrated validation of this full sequence in CAR-γδ T cells remains an important goal for future studies. CARs may enable precise localization to pathogenic cells expressing HSC surface markers such as FAP; γδ T cell activation releases perforin and granzymes to eliminate excessive ECM-secreting cells; finally, secretion of MMP-1/9 or HGF degrades ECM and promotes hepatocyte regeneration, achieving functional reversal (111–113). This approach has been extended to pulmonary, renal, and cardiac fibrosis (114). γδ T cells also modulate the Th17/Treg axis in pulmonary fibrosis, serving as potential biomarkers for disease progression and treatment response (115) (Figure 2). Recent transient expression strategies have demonstrated efficacy in ameliorating cardiac fibrosis and chronic inflammation, offering new avenues for systemic sclerosis and related disorders (116).

In autoimmune disorders, chimeric autoantigen receptor (CAAR) T cells provide a highly selective immune-reset strategy by incorporating the autoantigen itself (e.g., DSG3 or MuSK) as the extracellular domain, enabling antigen-specific recognition and apoptotic deletion of autoreactive B cells that express pathogenic autoantibodies as their B-cell receptor. This self-antigen-guided mechanism avoids the broad B-cell depletion and immunosuppression associated with conventional antibody therapies (12, 117–120).

Engineering CAAR constructs into γδ T cells further enhances safety and functional precision: γδ T cells possess a low risk of GvHD, inherent tissue-resident surveillance, and compatibility with allogeneic manufacturing, making CAAR-γδ T platforms promising off-the-shelf therapies currently undergoing Phase I evaluation (121). In systemic lupus erythematosus and rheumatoid arthritis, where pathogenic B-cell clones drive disease progression, CAAR-γδ T cells offer targeted elimination of autoreactive B cells while sparing normal counterparts.

Notably, the Vδ1 subset (predominantly tissue-resident) displays additional immunoregulatory features, promoting FoxP3/CD25 induction and secreting IL-10 and TGF-β1 to suppress autoreactive T-cell responses and support immune homeostasis (69, 122). Complementary preclinical evidence demonstrates that CD19-targeted CAR-γδ T cells can achieve profound B-cell depletion and durable remission in SLE models with minimal GvHD risk, reinforcing their potential as universal allogeneic immune-reset platforms (13).

In chronic infections such as HIV, achieving a functional cure requires effective clearance of latent viral reservoirs within resting CD4⁺ T cells, macrophages, and tissue sanctuaries. CAR-γδ T cells represent an attractive strategy due to their MHC-independent cytotoxicity, strong tissue infiltration, and intrinsically low risk of GvHD. γδ T cells can directly kill HIV-infected targets through NKG2D- and CD16-mediated pathways or via ADCC, while secreting antiviral cytokines including IFN-γ and TNF-α to restrict viral spread (123).

To enhance specificity and avoid the risks of self-infection associated with CD4-based CARs, preclinical programs have incorporated scFvs derived from broadly neutralizing antibodies (e.g., 3BNC117, VRC01) into CAR architectures. Viral resistance can be further strengthened by CCR5Δ32 knockout, and combining CAR-γδ T cells with latency-reversing agents (LRAs) enables a coordinated “shock-and-kill” approach to purge latent HIV reservoirs (124–126).

Recent studies using an HIV model demonstrate that mRNA-engineered CAR-γδ T cells achieve >70% killing of reactivated HIV-infected CD4⁺ T cells in humanized mice, without detectable cytokine release syndrome, reflecting improved safety and reduced immunogenicity (127–129). These findings highlight the rapid expansion of CAR-γδ T cell platforms from oncology into chronic infections, where their innate-like recognition and allogeneic compatibility offer distinct advantages for functional HIV cure strategies.

Conventional CARs rely on CD28 or 4-1BB domains, which may induce persistent activation and compromise expansion and durability (133–135). By contrast, CARs constructed with DAP10/DAP12 adaptors can integrate γδ T cell endogenous signals in a balanced manner: DAP12 provides ITAM-mediated activation, while DAP10 delivers YxxM co-stimulatory signaling, both highly expressed in resting and activated states (36, 136–138). These designs maintain antitumor efficacy while mitigating cytokine hypersecretion and exhaustion observed with CD3ζ-based CARs (35, 36, 139). NKG2D-based CARs incorporating these adaptors have demonstrated robust tumor clearance in solid tumor models and have been applied to γδ T-specific CAR engineering (136), highlighting their unique potential to reduce therapy-related toxicity (140, 141). KIR-based CARs further validate the feasibility of adaptor-based signal tuning (142).

γδ T cells can sense tumor stress via γδTCRs (143). CAR activation can be gated by γδTCR signals to form AND logic gates, initiating cytotoxicity only when both γδTCR stress recognition and CAR antigen engagement are satisfied (144, 145). An AND-gate CAR is a built-in safety logic: the engineered cell only becomes fully cytotoxic when two independent signals are received at the same time (e.g., the artificial CAR antigen plus a natural γδTCR stress signal), dramatically reducing the risk of attacking healthy tissue. γδTCR provides the first signal, and CAR provides the second signal, achieving dual verification and precise activation (35). This dual verification preserves MHC-independent cytotoxicity, enhances downstream CD3ζ signaling, and reduces excessive inflammatory responses (19, 70). Logic-gated CARs are increasingly explored to improve tumor specificity and minimize off-target effects (146–148), exemplified by γδ-enriched CAR-T combined with zoledronate in bone metastatic prostate cancer, achieving selective CAR-mediated killing (149). Controllable activation technologies, such as opto- or thermo-switches, support these logic-gated designs (98).

CRISPR/Cas9 has emerged as a key tool for γδT engineering, enabling selective knockout of inhibitory genes or insertion of function-enhancing modules to improve CAR-γδT cytotoxicity, metabolic fitness, and resistance to exhaustion. Typical strategies include knockout of KLRC1 (NKG2A), TGFBR2, CISH, and CD38 to enhance tumor targeting, as well as incorporation of proliferative genes such as IL-15 to boost expansion and persistence (19).

In solid tumor models, CRISPR-edited CAR-γδT cells reduce allogeneic antigenicity, lowering the risk of GvHD, and optimize DAP10/DAP12 signal integration to synergize with endogenous γδTCR signaling (150, 151). Non-viral CRISPR delivery allows for high-efficiency scFv knock-in, achieving higher CAR expression than conventional lentiviral transduction (152, 153). Recent studies also show that this strategy can mimic NK signaling pathways, enhancing metabolic reprogramming within immunosuppressive TMEs and avoiding mitochondrial dysfunction and functional decline observed in conventional CAR-αβT cells (148). These advances reinforce the principle of “empowerment rather than replacement” and provide a foundation for universal γδT cell products across diverse disease contexts.

ScRNA-seq provides molecular-level insight into γδT transcriptional heterogeneity and dynamic CAR responses, facilitating rational design. Integration with multi-omics data identifies γδT subset-specific states, predicts CAR-T responsiveness, and guides targeted gene modifications (154, 155). Pan-cancer analyses reveal γδTCR clonal diversity and tumor-specific expansions, informing multi-target CAR strategies to prevent antigen escape (156). ScRNA-seq can also track in vivo functional evolution of CAR-γδT cells, including cytokine production and exhaustion marker expression, and can be combined with CRISPR-mediated checkpoint knockout (PD-1/LAG-3) to enhance efficacy (48, 157). In non-malignant settings, scRNA-seq delineates γδT regulatory roles in autoimmune diseases, providing molecular evidence for conditional CAR designs (98). Overall, single-cell approaches strengthen signal-coordination strategies, improving CAR-γδT precision and predictability for personalized therapy.

Spatial transcriptomics and imaging preserve positional information, enabling analysis of γδT localization, interactions, and functional states within TMEs (158). Combined with single-cell data, spatial maps reveal compartment-specific infiltration and cell–cell networks. For example, Vδ1 γδT preferentially infiltrates fibrotic tumor regions and recognizes NKG2D ligands to enhance cytotoxicity (159). Spatial multi-omics can also identify immune resistance mechanisms, guiding CAR module design to target the TME. In tissue repair models, these techniques reveal anti-inflammatory roles of γδT in fibrotic microenvironments, supporting CAR-M or CAR-γδT strategies to promote M2 polarization and collagen degradation (38, 160). Spatially guided optimization of γδTCR-gated CARs ensures high-efficiency activation within diseased compartments, linking upstream signaling to downstream therapeutic outcomes.

Originally developed for cancer immunotherapy, CAR technology may enable precise recognition and killing of target cells (161). Its application has expanded into tissue repair and regenerative medicine, reflecting a shift from a “killer” to a “sentinel” paradigm (162). In non-malignant contexts, CAR cells not only target diseased cells but also initiate local reparative programs upon antigen recognition, releasing growth factors or immunomodulators to promote tissue remodeling and functional recovery. This approach offers novel interventions for chronic diseases characterized by fibrosis, inflammation, or tissue damage while avoiding systemic adverse effects (163).

Conventional CAR constructs typically comprise an extracellular antigen-recognition domain (scFv), a transmembrane region, and an intracellular signaling module (CD3ζ coupled with 4-1BB or CD28 co-stimulatory domains) (164). In the context of tissue repair applications, the design paradigm has shifted towards multifunctionality, which can be summarized in four key aspects. First, optimization of the targeting domain involves selecting disease-specific markers—such as FAP or uPAR, which are highly expressed in fibrotic lesions—to minimize off-target effects in healthy tissues. Second, dual functionality can be engineered so that CAR activation not only induces apoptosis in diseased cells but also triggers localized release of reparative factors (e.g., HGF, IL-10) through conditional expression systems driven by NFAT or IL-2 promoters, enabling a “sentinel-like” dynamic response. Third, expanding the repertoire of carrier cells, including CAR-macrophages and CAR-neutrophils, leverages their innate phagocytic or tissue-infiltrating properties to enhance repair outcomes. Finally, integrating safety switches—such as iCasp9 suicide genes or transient mRNA delivery platforms—allows precise temporal control of CAR expression, thereby mitigating long-term risks (110, 165–167).

In cardiology and oncology, CAR-based approaches have demonstrated remarkable potential in addressing tissue fibrosis. Cardiac fibrosis, often driven by fibroblast activation following myocardial injury, has been targeted using FAP-CAR-T and CAR-macrophage (CAR-M) strategies. For instance, intravenous administration of CD5-targeted lipid nanoparticles (LNPs) encapsulating FAP-CAR mRNA resulted in 17.5–24.7% of splenic T cells expressing CAR in AngII/PE-induced murine heart failure models. These CAR-T cells effectively cleared FAP^+ fibroblasts, reducing the fibrotic area by over 40% and improving left ventricular function, while the transient mRNA-mediated CAR expression (~1 week) conferred a favorable safety profile. Alternatively, bone marrow-derived macrophages (BMDMs) transduced with FAP-CAR containing a Megf10 domain (CAR-P) preferentially accumulated in the injured myocardium (GFP^+ cells representing 58% of CD3^+ T cells), eliminated FAP^+ myofibroblasts, and maintained an anti-inflammatory M2 phenotype. A nine-week follow-up revealed a 15–20% increase in left ventricular ejection fraction without detectable off-target cardiac toxicity (168–170).

In liver fibrosis, primarily driven by hepatic stellate cell (HSC) activation, uPAR-targeted CAR-Ms have been developed. These constructs—comprising uPAR-scFv, CD8 transmembrane domain, and CD3ζ signaling domain—were delivered via adenoviral transduction of BMDMs and administered to CCl₄-, DDC-, or HFCF diet-induced fibrosis models. CAR-M therapy efficiently eliminated uPAR^+ HSCs, reducing hepatic hydroxyproline content by 30–50% and serum ALT by ~40%. Additionally, CAR-Ms recruited neutrophils secreting MMP-9 for collagen degradation and induced M2 macrophage polarization (doubling the proportion of CD206^+ cells), thereby establishing a long-lasting anti-fibrotic microenvironment. In a 12-week CCl₄-induced cirrhosis model, this approach significantly decreased hepatic collagen deposition and increased serum albumin by 15% (171).

Renal fibrosis, associated with tubular interstitial fibroblast activation and local inflammation, has been targeted with dual-target CAR-M strategies. CARs designed against TNF and IL-4Rα may enable macrophages to activate the IL-4 anti-inflammatory pathway upon TNF engagement and adopt an M2 phenotype. In ischemia-reperfusion injury (IRI) models, CAR-Ms accumulated in injured kidneys (~8-fold fluorescence relative to healthy kidney), reduced neutrophil infiltration by ~60%, and promoted tubular epithelial proliferation (Ki67^+ cells doubled). In adriamycin-induced chronic kidney disease models, M2 CAR-Ms persisted for ~4 weeks, mitigating glomerulosclerosis and interstitial fibrosis while reducing proteinuria by ~50%. Moreover, FAP-CAR-Ms targeting renal FAP^+ fibroblasts in UUO models similarly decreased collagen deposition and improved renal function (163).

Beyond organ fibrosis, CAR technology shows broad potential in chronic non-malignant diseases, largely through modulation of the inflammation-fibrosis axis. In autoimmune disorders, FAP-targeted CAR-T cells have been shown to simultaneously alleviate fibrosis and modulate immune responses. In systemic sclerosis models, these CAR-T cells reduced fibrosis in skin and lung tissues while releasing IL-10 to rebalance Th17/Treg populations, decreasing fibrosis scores by ~25% (172, 173).

In age-related diseases, accumulation of senescent cells drives functional decline. NKG2D-L-targeted CAR-T cells efficiently eliminated senescent cells in cardiac and joint tissues and promoted regeneration via local HGF release. In non-human primates, this approach reduced local inflammation and suggested potential to extend health span (174, 175).

Chronic kidney disease represents another promising application, with anti-TNF CAR-Ms specifically recognizing inflammatory cues and activating IL-4 signaling to inhibit glomerulosclerosis and interstitial fibrosis. Preclinical studies reported a 20-30% improvement in serum creatinine, offering a novel therapeutic avenue (163).

In neurodegenerative diseases, CAR designs targeting activated microglia are under exploration. By incorporating BDNF release modules, these CAR cells can clear pathological microglia while providing neurotrophic support, representing a potential reparative strategy for Alzheimer’s disease (176). Collectively, these studies exemplify a shift in CAR applications in non-oncologic settings from cytotoxic “killer” functions toward immunoregulatory and reparative roles, highlighting their versatility across diverse chronic disease contexts.

In summary, the “δT-centric” engineering paradigm represents not merely an iteration of a single technology, but the establishment of an integrated technological ecosystem encompassing metabolic adaptation, signaling optimization, microenvironmental modulation, and multi-disease applicability (Figure 3). Centered on the intrinsic biological advantages of γδ T cells, this system achieves refined functional enhancement through metabolic-level modulation (glycolysis/OXPHOS switching and IL-15 insertion for sustained proliferation), signaling-level optimization (DAP10/DAP12 substitution, AND-gate control, and PD-1 knockout for precise activation), and microenvironmental reprogramming (macrophage polarization via IFN-γ and transient mRNA expression to reduce toxicity). Collectively, these synergistic strategies may enable differentiated therapeutic applications across solid tumors, fibrotic disorders, autoimmune diseases, and chronic infections. This “multi-dimensional synergy” design framework not only provides a technical foundation for the clinical translation of γδ T cell therapy across diverse disease contexts but also introduces a new paradigm for balancing personalization and universality in next-generation cellular immunotherapy.