Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Chimeric antigen receptor (CAR) T cell therapy can achieve outstanding response rates in heavily pretreated patients with hematological malignancies. However, relapses occur and they limit the efficacy of this promising treatment approach. The cellular composition and immunophenotype of the administered CART cells play a crucial role for therapeutic success. Less differentiated CART cells are associated with improved expansion, long-term in vivo persistence, and prolonged anti-tumor control. Furthermore, the ratio between CD4+ and CD8+ T cells has an effect on the anti-tumor activity of CART cells. The composition of the final cell product is not only influenced by the CART cell construct, but also by the culturing conditions during ex vivo T cell expansion. This includes different T cell activation strategies, cytokine supplementation, and specific pathway inhibition for the differentiation blockade. The optimal production process is not yet defined. In this review, we will discuss the use of different CART cell production strategies and the molecular background for the generation of improved CART cells in detail.
Keywords: chimeric antigen receptor, CAR, CART, adoptive cell therapy, immunotherapy, T lymphocyte, CART cell production, T cell activation, cytokines
Modern cancer therapies are increasingly relying on immunotherapeutic approaches. In particular, immune checkpoint inhibitors and adoptive cell therapy (ACT), including tumor-infiltrating lymphocytes (TILs), T cell receptor (TCR)-modified T cells, and chimeric antigen receptor (CAR) T cells represent milestones in innovative strategies for cancer treatment. ACT showed limitations, as the therapy with TILs only achieved encouraging results in selected highly immunogenic cancer entities, such as malignant melanoma [1]. Human leukocyte antigen (HLA)-restricted antigen recognition limits the application of TCR-modified T cells. The downregulation of HLA expression can lead to tumor escape [2]. CART cells combine the dynamic of T cells with the antigen-specificity of an antibody. They can bind the tumor antigen without antigen processing and independent of HLA-mediated antigen presentation. CD19-specific CART cell therapy showed very promising results in B cell malignancies, including acute lymphoblastic leukemia (ALL), chronic lymphocytic leukemia (CLL), and Non-Hodgkin lymphoma (NHL) [3]. Recently, the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) approved Kymriah ® (Tisagenlecleucel) for the treatment of patients with relapsed/refractory (r/r) B cell precursor ALL [4] or diffuse large B cell lymphoma (DLBCL) [5] and Yescarta ® (Axicabtagene Ciloleucel) for the treatment of patients with r/r DLBCL and primary mediastinal B cell lymphoma (PMBCL) [6]. Additional tumor antigen targets are currently under development, such as B cell maturation antigen (BCMA), for the treatment of multiple myeloma [7]. In solid tumors, CART cells still have to overcome limitations in their therapeutic use [8].
Despite encouraging response rates, relapses occur and limit the efficacy of this promising treatment approach. Therefore, it is critical to understand the current limitations of CART cell therapy in order to utilize the full potential of this modern anticancer therapy [9,10]. The in vivo efficacy of CART cells is linked to their proliferative capacity and long-term persistence to sustain sufficient anti-tumor activity [11]. The in vivo expansion and persistence of CART cells is limited in certain patients and it prohibits long-term anti-tumor control. One approach for improving the activity of CART cells is the further development of CAR constructs and gene transfer systems. The reduced fitness and dysfunction of T cells in the applied final cell product of certain patients might be another reason for the impaired in vivo activity. Therefore, improving the mitochondrial fitness and biogenesis may enhance the therapeutic efficacy of CART cell therapy and other ACTs [12]. Another approach for improved therapeutic CART cell efficacy is the selection or modification of CART cell subpopulations and subsets. The cellular composition of the final cell product has a major impact on the proliferative capacity of CART cells and it is directly linked to in vivo efficacy [13,14,15]. Optimal T cell activation and cultivation strategies for CART cell generation are crucial in producing efficient CART cells with the preferred T cell immunophenotype and subsets. However, the manufacturing processes of CART cells are not yet standardized. In this review, different strategies for the generation of highly potent CART cells will be discussed.
The number of transfused CART cells was assumed to majorly determine the therapeutic success in an early stage of the CART cell therapy. However, above a certain threshold, the absolute number of transfused CART cells does not directly correlate with in vivo expansion and therapeutic success [3]. Consequently, other factors than the absolute number of transfused CART cells might be more important for CART cell efficacy. For example, the cellular composition and phenotype of the adoptively transferred T cells, including T cell subtypes and subpopulations, was identified as one of the most critical success factors for efficient immunotherapy [16,17].
Although cytotoxic CD8+ CART cells, in particular, mediate direct tumor cell eradication, CD4+ T helper cells (Th cells) were identified as a highly potent and clinically important T cell subset [18]. It was demonstrated that CD4+ CART cells possess cytotoxic capacities that are comparable to cytotoxic CD8+ CART cells [19]. In addition, a balanced ratio of CD4+ Th cells and CD8+ cytotoxic T cells can positively influence the product regarding tumor eradication [13]. It was reported that the treatment of B-ALL patients with a 1:1 ratio of CD4+ and CD8+ (CD4:CD8 ratio) CART cells could achieve high remission rates [15]. For example, Lisocabtagene maraleucel (liso-cel; JCAR017) represents an anti-CD19 CART cell product administered in a defined composition with a specific ratio of CD4+ Th CART cells and cytotoxic CD8+ CART cells [20]. The subsets must be isolated at the beginning of the production and separately modified in order to gain a defined CD4:CD8 ratio, leading to a more complex manufacturing process.
Moreover, the different Th cell subpopulations play an important role. The balance between TEff cells and regulatory T (Treg) cells can influence the success of adoptive immunotherapy [21]. The infiltration of CD4+ Treg cells into solid tumors can decrease the anti-tumor activity of CD28-CD3ζ signaling CART cells [22]. The deletion of the Lck binding moiety in the CD28 CAR endodomain of a CD28-CD3ζ signaling CAR can enhance the anti-tumor efficacy in the presence of Treg cells [22]. It was reported that CART cells with the inducible T cell costimulator (ICOS) intracellular signaling domain can stabilize the Th17 cell function and enhance the in vivo persistence of CART cells in mice bearing human tumor xenografts [23]. Additionally, CART cells with the ICOS and 4-1BB intracellular signaling domains showed enhanced efficacy in solid tumors when compared to the 4-1BB-based CART cells [24].
Beside the T cell subtypes, the differentiation status of CART cells also plays a crucial role for therapeutic success. Isolated and ex vivo expanded T cells provide intrinsic properties that have to be considered in cellular immunotherapy. T cells vary in effector function, phenotypic characteristics, and their appearance in peripheral blood (PB) of healthy donors and patients depending on age, previous antigen exposure, and applied cytotoxic therapies due to their differentiation status [25]. In ACT, terminally differentiated CD45RA+ CCR7− T effector-like cells (TEff cells) demonstrated enhanced in vitro anti-tumor activity, whereas in vivo T cell activation, proliferative capacity, and persistence were impaired [14]. These findings changed the approach and criteria for the selection of specific T cell subsets for ACT and set the focus on less differentiated T cells: naïve-like T cells (TN cells) defined as CD45RA+ CD45RO− CD95− T cells express the lymph node homing markers CCR7 and CD62L, as well as CD28 and CD27 [17]. In contrast, the CD45RA− CD45RO+ CD95+ memory T cells can be divided in CD62L+ CCR7+ T central memory-like cells (TCM cells) and in CD62L− CCR7− T effector memory-like cells (TEM cells) [17]. Stem cell memory-like T cells (TSCM cells) represent a recently described T cell subpopulation resembling TN cells in that they are CD45RA+ CD45RO− CCR7+ and they express memory associated markers, such as CD95, and thereby exhibit properties of stem cells, including high proliferative and self-renewal capacity [25,26]. TN cells and TSCM cells have the capacity to persist and proliferate long-term in vivo after administration to the patient and they can possibly lead to improved clinical outcome [16,26,27]. In particular, the ability of self-renewal and the capacity to differentiate in all memory and effector subpopulations enable TSCM cells to sustain a long-lasting anti-tumor activity by supplying the immune attack with more differentiated TEM cells and TEff cells and refresh the pool of T cells with new less differentiated TSCM cells and TCM cells [17]. Consequently, transfusion of a high number of less differentiated CART cells is favorable for therapeutic success. The potential of individual T cell subsets is well described in the literature [17]. However, descriptions regarding how the formation of a more favorable cellular composition and T cell phenotype in the final CART cell product can be achieved during the production process are sparse.
Inhibitory tumor microenvironment binding inhibitory receptors, such as PD-1, CTLA-4, LAG-3, and TIM-3 on T cells might also cause insufficient response rates of CART cells in certain tumor entities, and therefore impair the immune attack [28,29]. A high expression of fatigue-related inhibitory receptors PD-1 and TIM-3 on CD8+ T cells is associated with impairment of the T cell function [30]. The dysfunction of tumor-specific T cells is a dynamic process that leads to antigen-driven differentiation and it is initiated in an early stage of tumorigenesis [31]. Transcriptomic profiling demonstrated that the expression of memory-related genes was enriched in CART cells from CLL patients achieving complete remissions. In contrast, the analysis of CART cells from non-responders revealed an upregulation of genes that mediate T cell differentiation, glycolysis, exhaustion, and apoptosis [32]. Furthermore, a population of less differentiated CD8+ CART cells without PD-1 expression was identified to play a crucial role in tumor control [32]. In addition, lower expression of PD-L1, PD-1, LAG-3, and TIM-3 was observed in lymphoma patients responding to CD19-specific CART cells treatment. Whereas non-responders were expressing higher levels of immune-checkpoint ligands on tumor cells and receptors on immune cells [33]. CART cells can provoke a reversible antigen loss through trogocytosis by transferring the target antigen to T cells, leading to a decrease of target density on cancer cells [34]. Additionally, T cell activity is reduced through the promotion of exhaustion and fratricide T cell killing [34]. It was reported that CART cells encoding a single immunoreceptor tyrosine-based activation motif (ITAM) showed an improved persistence of highly functional CART cells [35]. Strategies that led to a disruption of the interaction between inhibitory T cell receptors and their ligands expressed on cancer cells may improve the therapeutic efficacy of cell-based therapies. The administration of a PD-1 blocking antibody increased the therapeutic efficacy of CART cells [36]. Additionally, it was reported that anti-PD-1 single chain variable fragment (scFv)-producing CART cells mediated potent therapeutic effects when compared to conventional CART cells in preclinical models [37]. While these strategies aim to optimize CART cell therapy in vivo after the administration to the patient, additional strategies are essential that improve the exhaustion status and, in particular, possibly reduce the expression of inhibitory receptors on CART cells during the manufacturing process.
Another challenge is the improvement of CART cell infiltration into the tumor site. The T cell homing is the consequence of multiple molecular interactions. The repression of the anti-tumor immune response of CART cells in the tumor site can be mediated by an immunologic barrier [38]. Different homing properties constitute another distinctive feature of the different T cell subsets. While TN cells, TSCM cells, and TCM cells tend to migrate into lymphoid tissue, the TEM cells and TEff cells prefer peripheral tissue [25]. A stronger expression of the lymphoid homing marker CD62L and CCR7 on less differentiated T cells is associated with increased anti-tumor activity in preclinical models of ACT and might be beneficial for CART cells [27]. T cell extravasation, homing, and persistence in the tumor microenvironment are essential aspects in overcoming current limitations of CART cell therapy in solid tumors. It was demonstrated that CD28 costimulation could reduce the inhibition of T cell proliferation mediated by the transforming growth factor β (TGFβ) [39]. The overexpression of CXCR2 can improve T cell migration into tumor sites [40]. The overexpression of CCR2b on mesothelin-specific CART cells [41] and GD2-specific CART cells [42] led to enhanced T cell tumor infiltration. It was reported that CD30-specific CART cells expressing CCR4 could mediate an enhanced tumor control in a xenograft model [43]. NKG2D-specific CART cells could recruit and activate endogenous antigen-specific cytotoxic CD8+ cells and CD4+ Th cells in the tumor site in a CXCR3-dependent manner, leading to improved tumor eradication [44]. The modulation and role of specific homing marker expression on CART cells has to be further examined in the future.
The major aspects of the CART cell manufacturing process are relatively standardized, whereas clear differences can be identified in every single manufacturing step ( Figure 1 ).