Immunotherapy has emerged as an effective cancer treatment, particularly with immune checkpoint inhibitors (CPI) and adoptive cell therapies. However, limitations remain, including inefficient immune cell infiltration in some tumors, tumor heterogeneity, and immune escape mechanisms. Therapeutic cancer vaccines and oncolytic virotherapy offer potential solutions, and combining these modalities may enhance efficacy. The KISIMA vaccine platform is a chimeric recombinant protein with three key elements: a Multi-Antigenic Domain (Mad) containing multiple tumor antigens, a Cell Penetrating Peptide (CPP) for antigen delivery, and a TLR agonist (TLRag) for self-adjuvanting properties. This platform addresses potential antigen loss and induces cytotoxic T cells with a high frequency of memory precursors. Oncolytic viruses (OVs) exert therapeutic effects through various mechanisms, including direct tumor cell lysis and the release of tumor-associated antigens. However, their efficacy varies among tumors due to differences in intratumoral inflammation and immune activation. Arming OVs with tumor antigens can further enhance the tumor-specific T cell response. VSV-GP, a chimeric vesicular stomatitis virus (VSV) pseudotyped with LCMV-GP, is a potent oncolytic agent with reduced neurotoxicity and broad tumor tropism. It induces strong innate and adaptive immune responses and serves as a vaccine vector. A heterologous prime-boost regimen, combining a non-viral cancer vaccine with an oncolytic vaccine, might overcome the limitations of monotherapies. Alternating these platforms expressing shared tumor antigens could shift the immune response from antiviral to antitumor dominance. While this concept has been explored preclinically and in early clinical trials, a detailed mechanistic understanding remains limited. This study investigates the combination of KISIMA-TAA and VSV-GP-TAA in a heterologous prime-boost regimen, examining its impact on the tumor microenvironment (TME) and the quality and quantity of antitumor immune responses across various murine tumor models. The models represent different antigen classes (model antigen Ovalbumin or OVA, neoantigens Adpgk and Reps1, and oncoviral antigen HPV-E7). The study aims to dissect the induced immune components and assess their immunogenicity and efficacy.

The research builds upon previous work demonstrating the promise of both KISIMA and VSV-GP as single-agent therapies. KISIMA's effectiveness in inducing robust CD4+ and CD8+ T cell-mediated antitumor immunity has been established, as has its ability to enhance antitumor immune responses through optimized combinations with adjuvants. Similarly, the oncolytic potential of VSV-GP, including its ability to induce strong innate and adaptive immune responses in permissive tumors and to boost immune responses in homologous prime-boost settings, has been demonstrated. The literature also highlights the potential of heterologous prime-boost regimens combining non-viral cancer vaccines with oncolytic vaccines, suggesting a synergistic effect in overcoming monotherapy limitations and shifting the immune response towards a predominantly antitumor profile. However, mechanistic insights into these combinations have been limited, prompting the current investigation.

The study utilized various murine tumor models (E.G7-OVA, B16-OVA, TC-1, MC-38) expressing different tumor-associated antigens (OVA, Adpgk, Reps1, HPV-E7). Mice were immunized with KISIMA-TAA and VSV-GP-TAA using different administration schedules and routes (subcutaneous, intramuscular, intravenous). The optimal schedule, determined using an OVA model, involved priming with subcutaneous KISIMA-TAA followed by intravenous VSV-GP-TAA boost, and a subsequent KISIMA-TAA boost (KVK regimen). Immune responses were assessed through flow cytometry, measuring the frequency and number of antigen-specific CD8+ T cells in peripheral blood, spleen, bone marrow, and tumor infiltrates. The phenotype and functionality of these T cells (cytokine secretion, degranulation) were characterized. Transcriptome analysis (NanoString technology) of tumor tissue was conducted to investigate changes in the TME following vaccination. Immunohistochemistry was used to visualize immune cell infiltration in tumors. The therapeutic efficacy of the heterologous vaccination, alone and in combination with checkpoint blockade (anti-PD-1 or anti-PD-L1 antibodies), was evaluated by monitoring tumor growth and survival in the various tumor models. The study also included re-challenge experiments to assess the development of immunological memory. In vitro experiments examined the susceptibility of tumor cells to VSV-GP-induced oncolysis in the presence and absence of interferon.

The KVK heterologous prime-boost regimen resulted in significantly higher frequencies and numbers of circulating antigen-specific CD8+ T cells compared to homologous vaccination with either KISIMA-TAA or VSV-GP-TAA across various tumor models. This superior response was observed for both model antigens (OVA) and tumor-associated antigens (HPV-E7, Adpgk, Reps1). The KVK regimen induced a larger pool of poly-functional and persistent antigen-specific cytotoxic T cells. In the TC-1 model (immunologically 'cold'), KISIMA priming followed by VSV-GP boosting resulted in significantly higher numbers of HPV-E7-specific CD8+ T cells in the periphery, while both regimens induced comparable infiltration into the tumor. However, KVK vaccination resulted in a less exhausted phenotype of tumor-infiltrating CD8+ T cells, with higher functionality (cytokine secretion, degranulation) compared to homologous VSV-GP vaccination. Transcriptome analysis revealed a profound reshaping of the TME in KV-vaccinated TC-1 tumors, with significant upregulation of genes associated with immune activation (cytotoxic T cells, DCs, cytokines, chemokines, antigen processing and presentation) and downregulation of genes involved in cancer progression. A marked influx of CD8+ T cells and a decrease in immunosuppressive cells (M2-like TAMs, MDSCs) were observed in KV-vaccinated TC-1 tumors. In tumor models resistant or only partially responsive to oncolytic VSV-GP monotherapy, the heterologous vaccination significantly delayed tumor growth and increased median survival. The combination of KVK vaccination with checkpoint blockade further enhanced therapeutic efficacy, resulting in long-term survival in several models. Re-challenge experiments demonstrated the establishment of an effective memory response in long-term survivors.

The findings demonstrate the superiority of the heterologous KISIMA-TAA/VSV-GP-TAA prime-boost regimen over homologous vaccination in inducing potent and durable antitumor immunity. The combination addresses key challenges in cancer immunotherapy, namely inducing high-quality tumor-specific T cells and overcoming TME immunosuppression. The profound remodeling of the TME, characterized by increased immune cell infiltration and a shift towards an immunostimulatory environment, is crucial for enhancing antitumor responses. The observed synergy with checkpoint blockade highlights the potential of this approach to overcome resistance in immunologically 'cold' tumors. The varying responses across different tumor models underscore the importance of considering the tumor's susceptibility to oncolytic viruses and the interplay between the oncolytic and immunogenic effects of the combined therapy. The data provide compelling evidence supporting the clinical translation of this combined vaccine approach.

This study provides strong preclinical evidence supporting the combined use of KISIMA-TAA and VSV-GP-TAA in a heterologous prime-boost regimen for cancer immunotherapy. This approach effectively enhances both the quantity and quality of antitumor T cell responses, profoundly reshapes the TME, and overcomes resistance to checkpoint blockade in multiple tumor models. Given the ongoing clinical trials for KISIMA and the preclinical safety data for VSV-GP, clinical translation of this combined regimen is warranted. Future research should explore further optimization of the regimen, investigate the specific roles of different immune cell subsets, and evaluate its efficacy in diverse cancer types.

The study was conducted in murine models, and the findings may not directly translate to human patients. The sample sizes in some experiments were relatively small, potentially limiting the statistical power. Although several tumor models with varying degrees of susceptibility to viral oncolysis were used, further investigations in a broader range of tumor types are necessary to fully assess the generalizability of the results. While the study explored the effects of checkpoint blockade, additional mechanisms of resistance might need to be considered.