CAR T cell therapy in solid tumors: A review of current clinical trials

Abstract Chimeric antigen receptor (CAR) T cell therapy has made tremendous strides in the arena of hematological malignancies with approved therapies in certain leukemias, lymphomas, and recently myeloma with overall highly favorable response rates. While numerous clinical studies are still ongoing for hematological malignancies, research is developing to translate the feasibility of CAR T therapy in solid organ malignancies. Unfortunately, the majority of diagnosed cancers are primarily solid tumors. Thus, a highly unmet clinical need for further research and development exists in this field. This review article highlights currently active clinical trials and a few pertinent preclinical studies involving CAR T cell therapy in solid tumors while briefly discussing study outcomes and potential key targets that may allow for the feasibility of this therapy option. Finally, we mention critical challenges existing in the solid tumor environment and discuss developing strategies that may potentially overcome the existing barriers to CAR T cell progress in solid tumors.

While these few trials have rendered remarkable success in bringing these novel therapies to market, numerous other trials are continuing in the arena of hematologic malignancies. This potentiates the opportunity to explore the novelty of CAR T cell therapy in solid tumors. Current studies of CAR T cells in solid tumors primarily evaluate safety outcomes and report preliminary research findings thus far. Because the primary and secondary outcomes data continue to evolve for these trials, this article focuses on delivering a brief overview, reviewing essential targets, and discussing the current clinical trials of CAR T cell therapy in solid tumors.

BRAIN CANCER
Glioblastoma (GBM) is among the most common forms of malignant primary brain tumors. Current treatment options typically consist of surgery followed by chemotherapy or radiotherapy with a median 2-year patient survival rate of less than 30% [5]. In addition, because of its complexity, current treatments do not provide adequate disease control for patients [5]. Immunotherapy with CAR T cells is being studied as a novel option for this disease [6]. In part to investigate potential targets for CAR T cell therapy in GBM, clinical trials are exploring a variety of immunotherapeutic strategies, one of which is the target IL13Ra2, a commonly expressed membrane-bound protein in over 75% of GBMs that is associated with activating the mammalian target of rapamycin (mTOR) pathway favoring tumor growth [7].
One ongoing phase 1 clinical trial investigates the safety, efficacy, and feasibility of IL13Ra2 as a potential CAR T-cell target [8] in patients with recurrent or refractory malignant GBM due to its specificity for GBM tumor cells and limited expression on normal brain cells (NCT02208362). The findings from a patient case report related to this trial demonstrated a transient complete response when given intraventricular CAR T cells targeting IL13Rα2, with a clinical response sustained for 7.5 months after therapy initiation and improvements in quality of life [9].
Another considerable CAR T cell target is the human epidermal growth factor receptor 2 (HER2), a tyrosine kinase receptor that is overexpressed in GBM and many other human cancers [10]. artery. Yet, the data recorded from trials investigating this approach is limited [13].
A patient case report observes how anti-CEA CAR-T cells were infused via the hepatic artery using pressure-enabled drug delivery (PEDD) technology and was not associated with any serious or ontarget off-tumor adverse events. Following the CAR-T treatment, a complete metabolic response within the liver was sustained for 13 months revealed by positron emission tomography and normalized serum tumor markers with an abundance of CAR+ cells found within post-treatment tumor specimens [14]. Further studies will investigate this unique delivery method in the treatment for patients with liver metastases.
Another phase 1 trial is assessing the safety and tolerability of CYAD-101, a CAR-T receptor encoding natural killer group 2D (NKG2D) receptor within its intracellular domain. Data from preclinical models determined NKG2D to be a commonly over-expressed target in CRC. In this trial, among fifteen patients with unresectable metastatic CRC receiving three doses of CYAD 101 cells after standard chemotherapy, two patients had a partial response and nine were of stable disease [15].

PANCREATIC CANCER
While current immunotherapy with antibodies targeting PD-1, PD-L1, and CTLA-4 are used as treatments in pancreatic cancer, CAR T cell therapy is a particularly appealing and emerging therapy con- has become a key target for CAR T cell therapy trials [16], [17] (NCT03323944).
In a phase I study of HER2-directed CAR T cells in advanced pancreatic cancers [18], measured outcomes include off-tumor toxicities of HER2, partial response rate, achievement of stable disease, and ther- Another potential antigen target in hepatocellular carcinoma (HCC) for CAR T is glypican-3 (GPC3). Two prospective phase I studies in adults with advanced GPC3+ HCC used infusion of GPC3 targeting CAR T cell after cyclophosphamide and fludarabine-based lymphodepletion [22]. Out of 13 patients, nine experienced CRS and no patients experienced grade 3 or grade 4 neurotoxicity. The overall survival rates at 3 years were 10.5%, 1 year was 42%, and 6 months was 50.3%. One patient from the trial with the sustained stable disease was alive after 44.2 months. In addition, clinical trials are also investigating the use of GPC3 CAR T in combination therapy with checkpoint inhibitors, particularly with PD-L1-positive HCC [23].
A different phase I study used CD133 targeting CAR T in advanced metastatic solid tumors [24]. While enrollees that qualified had a vari-

BREAST CANCER
As One open-label phase I investigated the use of regional delivered autologous mesothelin-targeted CAR T with pembrolizumab for MPM [31].

Mesenchymal tumors
Despite a lack of specific targetable molecules, osteosarcoma is Another phase I trial included a pediatric patient with refractory metastatic rhabdomyosarcoma that was featured in a case study discussing the child's response to multiple cycles of HER2 CAR T [35]. CAR T regimen given the maintained HER2 expression. Pembrolizumab was also started 2 weeks after the second HER2 CAR T administration and was given every 3 weeks to promote CAR T function. The patient was able to obtain a second remission that has lasted.

CHALLENGES AND OPPORTUNITIES
Contrasting with the successes of CAR T treating hematologic malignancies, the development of CAR T therapy in solid tumors has progressed at a slower pace. Challenges unique to solid tumor settings arise in the form of tumor histopathological characteristics, lack of tumor-specific antigens, immunosuppressive tumor microenvironments (TME), and on-target, off-tumor toxicity that can be lifethreatening [36][37][38][39].

Antigen specificity
The lack of the specificity of antigens to target tumor cells is a critical issue leading to on-target, off-tumor toxicity in solid tumors. Tumorassociated antigens (TAA), antigens overexpressed on tumor cell surfaces, were initially thought to be an excellent target for the CAR T but utilization led to damage of normal healthy tissues throughout the body where these antigens were also present ( [39], [40]). These events can be fatal. For example, one case arose when a patient treated with anti-HER2 CAR T for metastatic colon cancer died five days later after the CAR T cells attacked healthy HER2 expressing lung epithelial cells [41]. Another case occurred when patients with neuroblastoma were given high affinity-GD2 CAR T that attacked healthy brain tissue expressing low levels of GD2, causing fatal encephalitis [41]. Both cases highlight the detriment of on-target off-tumor toxicity while also demonstrating the need for better targeting antigens.
Another differing factor between hematologic and solid malignancies is the homogeneity of antigens presented. Hematologic malignant cells tend to express homogeneous TAA, but solid tumors display antigen heterogeneity between tumor types and the primary versus metastatic stages of individual tumors ( [40], [42]). This means that one group of tumor cells expressing the antigen used for targeting would be destroyed, but other groups of tumor cells lacking the same antigen would escape and continue proliferating.

Tumor microenvironment
Vital to CAR T cells eliminating solid tumors is the proper trafficking of CAR T cells to the surface of the cancer so that they may bind to the target protein, but the TME impedes this transit. Solid tumors produce chemokines like CXCL1, CXCL12, and CXCL5 within the TME, preventing T cells from reaching the tumor cells. An example specific to CXCL12 revolved around a study in pancreatic cancer. CXCL12 is produced by carcinoma-associated fibroblasts (CAF) expressing fibroblast activation protein (FAP) [43]. Tumor cells have high concentrations of CXCL12 suspected to be due to the overexpression of high  [44].
Other factors within the microenvironment including excessive blood vessels, fibroblasts, and myeloid cells producing extracellular matrix, also serve to impede CAR T trafficking ( [38], [40], [42]). Other immune repressor cells are recruited to the TME such as regulatory T cells (Treg) and tumor-associated macrophages that act similarly to MDSCs by preventing cytotoxic cells from killing tumor cells.

Future directions
Each of these hurdles present in solid tumors contributes to the limited success of CAR T against solid tumors. Still, attempts are being made to overcome some of these obstacles, mainly through the design of CAR T agents. One strategy being investigated to overcome the issue of on-target, off-tumor toxicity from TAAs is the engineering of CARs targeting glycopeptide epitopes from mutations creating glycosylation present on tumor cells. However, this area of interest still requires extensive testing to ensure damage does not occur in healthy tissues [37]. Another idea is to use a strategy called Boolean AND-gate logic where multiple receptors are engineered on the T cells so that activation requires specific combinations of signals that will not be present beyond the TME [45].
In a similar vein to Boolean AND-Gate logic, AND-NOT logic is another strategy where CAR specificity could be increased by triggering T cell activation only in the presence of a TAA and not in the presence of a second antigen expressed on healthy cells. This can be achieved through engineering CARs to express a zipCAR, a universal receptor, consisting of a leucine zipper ectodomain fused to the transmembrane and intracellular signaling domains. Split, universal, programmable CAR T products (SUPRA CARS) utilize zipCARs [45].
ZipCARs lack ligand-binding domains and must be reconstituted with exogenous zipFv proteins, single-chain variable fragment adaptors, to activate T cells and bind with TAAs. It is thought that the second class of zipFv molecules could be administered to a patient to compete against zipCAR for binding, effectively preventing CAR-T activation in normal tissues.
These methods are not without drawbacks. There is the possibility that solid tumor cells may escape detection due to the various stages of mutation a tumor may be in preventing recognition by CAR-T products requiring combinations of antigens for activation. Mutations for glycosylation may also be absent eliminating one method of differentiation between normal and solid tumor cells [37]. To avoid convoluted engineering practices, the simplest way to overcome a poor antigen choice is to select a better antigen specific to the target, but this too poses its problems. An example is EGFRvIII found in glioblastomas.
Studies initially showed that it might serve as a good target, but heterogeneity in expression and tumor response to downregulate EGFRvIII leads to marginal growth effects on the tumor [37].

CONCLUSION
While the investigation of CAR T cell therapy in solid tumors is relatively new, an extensive amount of ongoing research will be necessary to assess the safety and feasibility of their place in treatment in solid tumors. Most of the current clinical studies discussed in this article are phase 1 trials that mainly investigate CAR T products' safety and efficacy in solid tumors with several in the pipeline.
Indeed, the need to better understand these therapies will continue to exist and develop over time. CAR T therapy has great potential to impact the entire landscape of solid tumor malignancies just as it has for hematological cancers. Perhaps, the success of these current trials will lead to the progress of future phases 2 and 3 trials. In addition, the new designs for CAR T products emerging may serve to circumvent challenges posited by the solid tumor microenvironment. As research expands in CAR T therapies, further challenges and opportunities will continue to emerge. Thus, scientific research and development will continue to grow, guide, and impact its potential in a positive direction.