Chimeric Antigen Receptor T cells (CAR T) are effective in hematologic malignancies yet remain largely ineffective in solid tumors. We recently engineered a synthetic vaccine to boost CAR T cells through the chimeric receptor directly in vivo. Vaccine boosting enhanced CAR T expansion, polyfunctionality, and anti-tumor efficacy in multiple immunocompetent solid tumor models. Importantly, vaccine boosting of CAR T cells triggered robust antigen spreading to circumvent antigen-negative tumor escape and more effectively treat solid tumors with pre-existing antigenic heterogeneity. Thus, vaccine boosting provides a clinically-translatable strategy to enhance CAR T cell therapy against solid tumors.

An exciting, potentially new therapeutic approach to cancer, neoantigen-based personalized cancer vaccines (PCVs) have demonstrated promising signals in single-arm and more recently in randomized Phase 2 study. What do these results mean for patients and the field, and where will neoantigen-based PCV go from here? This discussion will review findings to date discuss the broad potential of this novel approach to drive improved outcomes across the spectrum of solid tumors.

This discussion will highlight recent results from a promising Phase 1 trial of a new lymph node-targeted vaccine modality targeting mutant KRAS. The preclinical and clinical data will be reviewed including discussion of the AMP platform technology, ELI-002 product design, most recent FIH Phase 1 clinical trial results including immunogenicity (mKRAS-specific T cell responses and personalized antigen-spreading responses), anti-tumor biomarker results, and association with clinical outcome (RFS and OS). This will provide an example of an off-the-shelf vaccine approach targeting a common public driver mutation with promise for application in numerous settings of unmet medical need.

Here we demonstrate the proof of concept of a bispecific optimised HLA-Fc fusion conjugated to a SIRPa protein which shows potent in vitro anti-tumor efficacy through its multimodal binding of LILRB1, LILRB2 and CD47 receptors. This novel modality of HLA/SIRPa Bispecifics has the unique characteristic of targeting myeloid cells (positive for LILRB1/2) and the ability to be directed to tumor sites by targeting the CD47 checkpoint receptor expressed on cancer cells.

Here, I discuss recent advances in the use of high-throughput genomics and single-cell-based technologies that can expedite the discovery mechanisms to improve the safety and efficacy of adoptive cell therapies.

Despite its promising potential, CAR T cell therapy faces challenges due to the tumor microenvironment (TME), which harbors immunosuppressive factors and physical barriers that hinder CAR T cell efficacy. To overcome this resistance, strategies are being developed to enhance CAR T cell fitness, engineer them to resist the TME, and target multiple tumor antigens. By addressing these challenges, CAR T cell therapy holds the promise of achieving more durable responses and better patient outcomes.

To broaden the application of IEC therapy and reduce the significant resource demands required to facilitate CAR-T, newer strategies for the prevention of toxicity are required. Trials using prophylactic strategies such as tocilizumab to prevent CRS resulted in higher rates of high grade ICANS and deaths. We have yet to identify optimal management for prevention. New approaches utilizing novel immunomodulatory agents have given insight into potential mechanisms of toxicity and have provided paths forward to deliver safer and more effective cellular therapies.

Genome annotation, protein structure prediction, and other biological subfields have been revolutionized by artificial intelligence (AI) and machine learning (ML). Advances in DNA synthesis and high-throughput experimental techniques enable construction and screening of large libraries of modular cell therapy constructs. AI and ML algorithms trained on library screening data can accelerate the development of these cell therapies by generating predictive models, design rules, and improved designs.

Patient responses to engineered cell therapies are variable and unpredictable. And while hundreds of potential gene targets, CAR sequences, and engineering strategies to improve T cell function have been proposed, no method exists to rapidly examine different classes of genetic manipulations simultaneously or in combination. We have developed CRISPR-All, a unified genetic perturbation language able to arbitrarily and combinatorially examine genetic perturbations across perturbation type and scale in human immune cells.

Tumor-targeting immunotherapies including ADCs have great potential for the treatment of solid tumors, but their progression in the clinic can still be limited by off-tumor toxicity, where the antibody therapeutic binds to healthy tissues and induces healthy-cell killing or payload delivery. Developing selective therapeutics for the tumor or TME is critical to reduce therapeutic adverse events (TAE). We show that we can design a bispecific HER2-targeting TCE by machine learning, to be wholly selective against HER2+ tumor cells while avoiding healthy tissue engagement, generating an antibody that only functions above a certain tumor-antigen density threshold.

Mesothelin is an attractive target for cell therapy of solid tumors given its limited expression on normal human tissues and high expression in most solid tumors. In general, anti-mesothelin CAR T cells have had limited efficacy in patients. We have developed a highly active CAR T cell that binds to the membrane proximal region of mesothelin close to the cell surface and is associated with increased persistence in the tumors.

B7-H4 is an attractive target for breast cancer with robust expression on cancer cells and low expression in normal tissues. We have engineered EVOLVE-106, a B7-H4-targeting T cell engager that integrates affinity-tuned anti-CD3 binding and CD2 co-stimulation in a single biotherapeutic molecule to optimize anti-tumor responses. Following antibody engineering, we have identified an EVOLVE-106 lead candidate with superior biophysical properties, lack of peripheral immune activation and robust in-vitro tumor killing.