This year was the initiation of the Next Generation CAR & T Cell Therapies conference, held in conjunction with the Next Generation Protein Therapeutics & Bioconjugates™ Summit and Cell Line Development & Engineering US event. With CAR & T cell therapies generating so much excitement for liquid cancer malignancies, many researchers are looking to this immunotherapy for solid tumors and also to broaden the scope of current CAR-T therapies by moving in an allogeneic direction. As the field matures, it is also evident that dialing in the CAR-T manufacturing process by incorporating PAT (process analytical technology) will be necessary to reduce costs and capitalize on novel analytical capabilities and equipment. This is by no means exhaustive coverage of the conference, but provides talk highlights on strategies to address key issues.
CAR-T in Solid Tumors
The first day of the conference focused on current developments for CAR-T in solid tumor modalities. It is well known that tumor cells have developed strategies to elude T cell-mediated immunity through events such as downregulation of MHC molecules and release of inhibitory signals in the tumor microenvironment (TME), which decreases their anti-tumor efficacy and activity. T cells require 3 signals for activation and proliferation:
- CAR or TCR engagement of peptide-MHC
- Co-stimulation between CD28 and 4-1BB
- Presence of inflammatory cytokines like IL-12 or type 1 IFN
In solid tumors, the cancer cells downregulate signal #1, lack signal #2 and secrete inhibitory cytokines into the tumor microenvironment (TME) that prevent T cell activation (weak or no signal #3). Many groups are looking to boost in vivo persistence and function of tumor-specific T cells using different strategies. It was evident throughout the day that likely a combinatory treatment approach to address these 3 signals will be the most efficacious for solid tumors.
For example, Cellectis is using Transcription Activator-like Effector Nuclease (TALEN) to modify endogenous immune pathways to increase T cell activity. They repurposed the TRAC (T Cell Receptor Alpha Constant) and PDCD1 (Programmed Cell Death 1) or IL2rα loci to express the CAR and activating cytokine IL-12 in a tumor-dependent manner (provide signal #3). They found the IL-12 expressing CAR-T cells to have improved anti-tumor activity in vitro compared to unmodified CAR-T cells. This multiplex gene editing technology is not limited to the CAR/TCR activation pathway and could be utilized to express other cytokines, which broadens the scope of this technology and allows for individualized approaches for specific indications.
At Refuge Biotechnologies, CRISPR/Cas9 sgRNA technology is used to simultaneously activate and repress multiple genes to increase activity of CAR-Ts. The Cas9 protein is mutated so it can no longer cleave DNA, creating a “dead Cas9” or dCas9 which acts as a carrier to specific regions of the genome, where it can deliver a transcriptional activator or repressor to turn on or turn off genes. In their clinical pipeline are the HER2 CAR/PD1 sgRNA T cells for HER2 breast and ovarian cancers.
Another approach as described by Cliona Rooney of Baylor College of Medicine is to target multiple tumorigenic antigens to overcome antigen escape by tumor cells, essentially boosting signal #1. While it is ideal to find the optimal target anti-tumor antigen in immunotherapy, the phenomenon of epitope spreading has been linked to superior clinical outcome. Therefore, targeting multiple antigens reduces the chance of immune evasion by antigen modulation by the tumor cells. In an effort to provide signal #3, T-cells modified to constitutively express IL-7 receptor were found to enhance T-cell survival and enable CAR-T cells to eliminate tumor in vivo in three different tumor models. Strategies to overcome the immunosuppressive TME were also discussed. Namely, overcoming TGFß, a potent T-cell inhibitor, expressed by tumor cells, with expression of dominant negative TGFß receptor (DNR). This DNR lacks an intracellular signaling domain eliminating the transmission of inhibitory signal when TGFß binds, thereby protecting the T-cells from inactivation by the TME.
CARs on other Immune Cells
While cytotoxic αβ T cells are potent immune cells, their safety profile and lack of persistence has led researchers to investigate alternative immune cells to arm with CARs. Carisma Therapeutics is looking at CAR-modified macrophage cells for solid tumor adoptive cell therapy. Macrophages are able to phagocytose tumor cells in a CAR-dependent manner. As the tumor cells are broken down, antigen processing and presentation to T-cells activates the adaptive immune system, broadening the immune response. These cells require a threshold concentration of antigen to activate safeguarding healthy cells from phagocytosis. For example, normal cells express low levels of HER2, therefore only tumor cells expressing high levels of HER2 would be selectively phagocytosed by the HER2-CAR macrophages. It was also noted that because macrophages leave the bloodstream to infiltrate tissues within 24hours of administration, CRS is not likely a concern. Since macrophages lack TCR, the potential risk of GvHD is also low and could allow for allogeneic CAR-macrophage approaches.
Improving CAR-T Safety
The incidence of deleterious effects of CAR-T therapies has led to researchers to find better ways to control the activity of these cells in vivo. Stephen Santoro of Kite/Gilead discussed molecular “on/off switch” technologies to improve the safety profile of CAR-T therapy. Early generation switch technologies (i.e. suicide genes) suffer from “leakiness” where there is an incomplete “off” when the switch is activated. Two next generation switch technologies were discussed:
- Heterodimerizing (HD) switch receptor technology (developed by Cell Design Labs) is based on ligand binding for activation. In this scenario, an FDA-approved orally administered small molecule drug acts as a ligand to mediate receptor dimerization on the surface of CAR-T cells to turn OFF the cells.
- “Concealed CAR” receptors are expressed on the T-cell surface only in the presence of ligand. This rapid cell membrane trafficking is reversible allowing for tight ON/OFF control of CAR-T activation.
By incorporating these “on/off switches” into next generation CAR products, physicians will have the potential to rapidly control and reversibly titrate the activity of CAR T-cells.
UNUM Therapeutics presented their antibody-coupled TCR (ACTR) platform as a unique strategy to improve safety of T cell adoptive therapies. Their ACTR technology relies on cancer-specific antibodies to direct the CAR-T modified cells to tumor cells. CD16 binding to the Fc region of the antibody allows selective targeting of the T cells. Preclinical studies showed that ACTR’s killing activity is based on the concentration of antigen on a target cell which avoids on-target/off-tumor toxicity commonly associated with traditional T cell approaches. The targeting of multiple tumor antigens through administration of antibody cocktails could overcome immunosuppressive tactics used by tumor cells, such as downregulation of certain antigens. In HER2+ solid tumors (ACTR707), in vitro cytotoxicity was observed with HER2-expressing mammary cell line MCF7 but not with a primary mammary cell line (normal tissue). This tumor specificity was not observed with CAR-HER2 which killed normal and cancerous mammary cells indiscriminately.
On day 2, many of the talks focused on allogeneic CAR-T approaches.
Celyad has developed CYAD-101, which is an allogeneic CAR T-cell therapy where T cells co-express TIM (TCR inhibitory molecule), which is a truncated form of CD3ζ and a NKG2D-based CAR. GvHD is triggered by the recognition of foreign HLA molecules expressed on the patient’s cells by the TCR of the donor lymphocytes. The TIM-mediated inhibition of TCR signaling is designed to prevent alloreactivity. CYAD-101 is currently being evaluated in the AlloSHRINK trial. Celyad is also using Horizon Discovery’s shRNA SMARTvector technology to knockdown the TCR complex in allogeneic CAR-T therapies. This technology has been used to generate non-gene edited allogeneic CYAD-200 series of CAR-T candidates. In vivo data demonstrate that shRNA targeting of CD3ζ effectively protects against GvHD comparable to CRISPR-Cas9 based knock-out. Preclinical tests show increased persistence of allogeneic T cells using shRNA targeting compared to gene editing technologies.
Ɣδ T cells
Several groups are looking to utilize a different T cell subset, the Ɣδ T cells (GDT cells) for allogeneic CAR-T therapy. As part of the innate immunity, GDT cells have intrinsic antitumor activity and recognize stress or transformed cells in an MHC-independent manner, thus avoiding the potential of GvHD making them promising allogeneic therapeutic candidates. Several companies are looking to harness this T cell subset for solid tumor targets.
1) Immatics presented work on their ACTallo® platform, which combines γδ T cells with their XPRESIDENT® target/TCR discovery platform. The ACTallo® process selectively expands Vγ9Vδ2 T cells that are retrovirally-transduced to co-express CD8 together with a tumor-specific αβ TCR. Testing showed these cells did not exhibit antigen exhaustion like αß T cells and had increased in vivo persistence post-expansion. Immatics is currently in preclinical stages of investigating their ACTallo® platform as “off-the-shelf” adoptive therapies with their IMA301 program.
2) Scotland-based TC Biopharma has several GBT cell therapies in their clinical pipeline. TCB001 is a phase II/III trial using autologous unmodified GDT cells (ImmuniCell) for treatment of solid tumors (late stage melanoma). The group is also making HLA-matched allogeneic GDT cell banks to manufacture next-generation ‘off-the-shelf’ GDT cell therapies. Around 15-18 biobanks would cover the genetic heterogeneity of GDT cells in the global population. TCB002, a Phase I clinical study initiated in January 2019, aims to treat Acute Myeloid Leukemia (AML) patients with allogeneic GDT cells from healthy donors, which are expanded and activated prior to infusion into patients. Additionally, CAR-modified GDT cells are being investigated for hematological malignancies (TCB003, TCB004) as well as solid tumors (TCB005, TCB006).
Alternatives to Viral Vectors
The viral vector bottleneck is a reality of the current paradigm of CAR-T manufacturing. Groups such as Precision Nanosystems are looking to address this with their non-viral nucleic acid delivery system. Their proprietary, chemically defined lipid formulations allow the encapsulation of siRNA, mRNA or DNA with no size limitation with high transfection efficiencies and minimal impact on cell viability. With an eye towards scalability, manufacture of lipid-encapsulated nucleic acids (hybrid LNPs) can be achieved with the suite of NanoAssemblr® microfluidics platforms for all stages of clinical development.
MD Anderson is looking at the non-viral Sleeping Beauty (SB) transposon system to develop therapeutics for both solid tumors and blood cancers. Compared with recombinant viral DNA delivery methods, SB is faster simpler and cheaper way to customize T cells. Using the SB platform, CAR-T cells co-express CD19-CAR, membrane-bound interleukin 15 (mbIL15), and a molecular safety switch, which can deactivate the modified T cells if needed. The mbIL15 increases T-cells persistence and prevent antigen exhaustion. Results from the first in-human trials has demonstrated safety, tolerability, disease response, long-term survival, and persistence of infused CD19-specific CAR-T cells. Currently, clinical trials are underway to evaluate the SB platform for a “Point-of-Care” approach where rapid personalized manufacture (RPM) – within two days – of CD19-specific CAR-T cells would occur at the treatment center (RPM mbIL15-CAR-T cells) with no ex vivo manufacturing of cells to allow a much faster “vein to vein” time for treatment.
The final day of the conference focused on manufacturing and analytics of CAR-T products. One key takeaway from many of the presenters is that a constant feedback loop from development through commercialization with your quality control and manufacturing teams is imperative to success. That way, as methods improve, any optimization or validation can be dealt with in real-time to prevent any delays in the timeline as products move through clinical phases.
For autologous CAR-T patients, vein to vein time is critical, but manufacturing methods are still highly dependent on skilled labor force, with little automation. Also, complex manufacturing and vector supply limitations complicates process control. Tom Elfante of Legend Biotechnologies discussed CMC (chemistry, manufacturing and control) pitfalls. One key point is that for cell therapy products, the quality of product is really defined by the customer, not by the process expert. Therefore, optimizing manufacturing needs to positively impact quality as defined by the customer (patient safety, product quality, on-time delivery, cost). And, in evaluating manufacturing procedures, one needs to implement Critical to Quality (CTQ) designations to balance waste and value for optimization. Activities can be divided into 3 categories:
- Value Added- procedures you want to optimize
- Value Enabling – procedures you want to automate
- Non-Value Added – procedures you want to eliminate
Andrea Rossi Moore, Director of Analytical Development from Tmunity Therapeutics gave an excellent talk on analytical testing for CAR-T cell products. One of the main challenges facing CAR-T therapies is the clinical pathways are not well defined for CAR-T making it difficult to apply traditional assay development strategies. She stressed the importance of reference standards, controls and reagent QC when looking to qualify analytical methodologies especially with patient-variable cells. For cell therapy products, reference standards are very important since there are no commercially available controls to use. Therefore, inventory management is key as the material is limited in supply.
Damian Marshall from the Cell and Gene Therapy Catapult discussed predictive analytics as it relates to PAT. Currently, during CAR-T manufacturing, the bulk of the testing is done on the finished product with little in process analytics data generated (descriptive analytics). The goal is to obtain predictive analytics where real-time in-process measurements allows for immediate changes in manufacturing. He described the use of Raman spectroscopy to quantify metabolite levels during manufacturing. This method used to be difficult to do but there are now more ‘plug and play’ formats making the technique more accessible. Each metabolite has its unique spectroscopy data (process fingerprint) so you can establish Raman profiles for various molecules. This allows for ‘adaptive manufacturing’ where the metabolite data drives process control allowing you to adjust your process in real time. Monitoring in-line at bioreactor level to check cell viability and other parameters but, there is a big data challenge. How does one handle all of the data points that are now being captured? That is the next big challenge for cell therapy manufacturing.
Overall, great strides are being made to address the safety of CAR-T, working on allo-CAR-T platforms and tackling solid tumor modalities more effectively. It was an intense few days of talks but definitely a great way to get a sense of the state of CAR-T research.
For more conferences on cell therapy, please see:
September 9-12, 2019, Boston Convention and Exhibition Center (BCEC), Boston, MA
December 3-6, 2019, RAI, Amsterdam
February 26-28, 2020, Westin Miyako, Kyoto, Japan