Tumor-Specific Antibody, Cetuximab, Enhances the In Situ Vaccine Effect of Radiation in Immunologically Cold Head and Neck Squamous Cell Carcinoma.

In head and neck squamous cell carcinoma (HNSCC) tumors that over-expresses huEGFR, the anti-EGFR antibody, cetuximab, antagonizes tumor cell viability and sensitizes to radiation therapy. However, the immunologic interactions between cetuximab and radiation therapy are not well understood. We transduced two syngeneic murine HNSCC tumor cell lines to express human EGFR (MOC1- and MOC2-huEGFR) in order to facilitate evaluation of the immunologic interactions between radiation and cetuximab. Cetuximab was capable of inducing antibody-dependent cellular cytotoxicity (ADCC) in MOC1- and MOC2-huEGFR cells but showed no effect on the viability or radiosensitivity of these tumor cells, which also express muEGFR that is not targeted by cetuximab. Radiation enhanced the susceptibility of MOC1- and MOC2-huEGFR to ADCC, eliciting a type I interferon response and increasing expression of NKG2D ligands on these tumor cells. Co-culture of splenocytes with cetuximab and MOC2-huEGFR cells resulted in increased expression of IFNγ in not only NK cells but also in CD8+ T cells, and this was dependent upon splenocyte expression of FcγR. In MOC2-huEGFR tumors, combining radiation and cetuximab induced tumor growth delay that required NK cells, EGFR expression, and FcγR on host immune cells. Combination of radiation and cetuximab increased tumor infiltration with NK and CD8+ T cells but not regulatory T cells. Expression of PD-L1 was increased in MOC2-huEGFR tumors following treatment with radiation and cetuximab. Delivering anti-PD-L1 antibody with radiation and cetuximab improved survival and resulted in durable tumor regression in some mice. Notably, these cured mice showed evidence of an adaptive memory response that was not specifically directed against huEGFR. These findings suggest an opportunity to improve the treatment of HNSCC by combining radiation and cetuximab to engage an innate anti-tumor immune response that may prime an effective adaptive immune response when combined with immune checkpoint blockade. It is possible that this approach could be extended to any immunologically cold tumor that does not respond to immune checkpoint blockade alone and for which a tumor-specific antibody exists or could be developed.

In situ Vaccine Plus Checkpoint Blockade Induces Memory Humoral Response.

In a syngeneic murine melanoma (MEL) model, we recently reported an in situ vaccination response to combined radiation (RT) and intra-tumoral (IT) injection of anti-GD2 hu14. 18-IL2 immunocytokine (IC). This combined treatment resulted in 71% complete and durable regression of 5-week tumors, a tumor-specific memory T cell response, and augmented response to systemic anti-CTLA-4 antibody checkpoint blockade. While the ability of radiation to diversify anti-tumor T cell response has been reported, we hypothesize that mice rendered disease-free (DF) by a RT-based ISV might also exhibit a heightened B cell response. C57BL/6 mice were engrafted with 2 × 106 GD2+ B78 MEL and treated at a target tumor size of ~200 mm3 with 12 Gy RT, IT-IC on day (D)6-D10, and anti-CTLA-4 on D3, 6, and 9. Serum was collected via facial vein before tumor injection, before treatment, during treatment, after becoming DF, and following rejection of subcutaneous 2 × 106 B78 MEL re-challenge on D90. Flow cytometry demonstrated the presence of tumor-specific IgG in sera from mice rendered DF and rejecting re-challenge with B78 MEL at D90 after starting treatment. Consistent with an adaptive endogenous anti-tumor humoral memory response, these anti-tumor antibodies bound to B78 cells and parental B16 cells (GD2-), but not to the unrelated syngeneic Panc02 or Panc02 GD2+ cell lines. We evaluated the kinetics of this response and observed that tumor-specific IgG was consistently detected by D22 after initiation of treatment, corresponding to a time of rapid tumor regression. The amount of tumor-specific antibody binding to tumor cells (as measured by flow MFI) did not correlate with host animal prognosis. Incubation of B16 MEL cells in DF serum, vs. naïve serum, prior to IV injection, did not delay engraftment of B16 metastases and showed similar overall survival rates. B cell depletion using anti-CD20 or anti-CD19 and anti-B220 did not impact the efficacy of ISV treatment. Thus, treatment with RT + IC + anti-CTLA-4 results in adaptive anti-tumor humoral memory response. This endogenous tumor-specific antibody response does not appear to have therapeutic efficacy but may serve as a biomarker for an anti-tumor T cell response.