Combining Oncolytic Vaccinia Virotherapy with Adoptive T Cell Therapy,

Immunotherapy with cytotoxic T lymphocytes (CTLs) has proved safe and effective for the treatment of post-transplant viral re-activation and lymphoproliferative diseases. T cells specific for non-viral tumor antigens (TA) have been effective for certain tumors but their efficacy is limited. To improve T cell activity and circumvent tumor evasion T cells have been genetically modified with chimeric antigen receptors (CARs), that consist of extracellular variable domains of antibodies specific for tumor cell surface markers linked via a transmembrane domain to the z chain of the T cell receptor. Although the persistence of CAR-T cell can be improved by the inclusion of signaling domains from various costimulatory molecules, these rarely produce sufficient proliferation and persistence in vivo. An alternative way to promote T cell expansion in vivo is by vaccination. If the T cells carrying the transgenic CAR are naturally specific for a vaccine antigen, then the vaccine could be used in vivo to enhance proliferation and anti-tumor activity.

The oncolytic vaccinia virus (OVV) JX-594 produces reduction in tumor bulk, but complete remissions have not been published to date. JX-594 is tumor selective due to its dependency on increased EGFP-Ras pathway and sensitivity to IFN and it potently activates innate immunity by transgenic expression of human GM-CSF and TLR stimulation. We hypothesize that JX-594 could enhance the antitumor efficacy of vaccinia virus (VV)-specific T cells expressing tumor-specific CARs while at the same time reducing tumor mass allowing increased T cell penetration and survival.

To identify VV vaccine antigens that induce T cell proliferation in vivo, we selected 6 different VV antigens and asked first if they could activate memory T cells from vaccinated individuals and second if the frequency of these T cells in vivo increased in response to vaccination both in healthy donors and in patients with hepatocellular carcinoma (HCC) who had received OVV treatment. We then determined if VV-specific T cells could be induced to express a CAR specific for human epidermal growth factor receptor 2, HER2 that is expressed on a range of solid tumors.

PBMCs were obtained from healthy donors and patients who had received intratumoral JX-594 injections at the University of California in San Diego, USA, and McMaster University Medical Center, Canada. PBMCs were stimulated with overlapping peptide libraries (20aa overlapping by 15aa) spanning the entire protein sequences of the A10L, D8L, H3L, G5R, B22R, and D8L antigens of VV. Three days after stimulation, cells were transduced with retroviral vector encoding HER2.CAR and then expanded as for non-transduced T cells. After 9 days T cells were tested for their dual specificity and function in ELIspot and cytotoxicity assays.

We activated VV-specific T cells from 20/21 healthy donors vaccinated from 1 month to over 40 years previously. VV-specific T cell lines recognized a median of 4 of the 6 VV antigens (range 0 to 6). The frequency of VV-specific T cells increased in response to vaccination in healthy seronegative donors and in patients receiving multiple OVV injections. Transduced VV-specific T cells expressed the HER2.CAR in 40 to 50% of cells and CAR(+) T cells secreted g-IFN in response to stimulation with VV peptides in intracellular cytokine assays, validating their dual specificity. HER2.CAR-transduced VV-specific T cells killed both VV peptide pulsed activated T cells and HER2-expressing tumor cells in an HLA-independent manner.In most donors we were not able to detect VV-specific T cells without prior stimulation. However VV-specific T cells from two healthy donors could be detected in blood and expanded 120 and 90 fold, respectively, after one stimulation. In all the other donors, VV-specific T cells expanded to comparable frequencies at the end of the first stimulation, range 0.1% to 5% (with a median of 1.5%) of the final T cell population as measured by ELISpot assay.

We have identified 6 VV antigens that consistently reactivate and expand T cells after vaccination. T cells specific for these antigens will therefore be suitable hosts for tumor-specific CARs that can be reactivated and expanded in vivo after adoptive transfer until the tumor is eliminated. We will test our combined OVV & adoptive T cell strategy in an immunocompetent murine model prior to evaluation in clinical trials.

No relevant conflicts of interest to declare.