Introduction: At 52.5% identity in their extracellular domains (ECDs), mouse and human GITR (CD357) sequences are only modestly conserved. Thus transitioning promising murine anti-tumor activity data with the anti-GITR DTA-1 and 2F8 Mabs to clinical trials with the non-crossreactive anti-human GITR Mab TRX518 poses a considerable limitation. Here we investigated three elements of the GITR mouse and human antibodies to address this limitation: 1) the functional epitopes targeted by anti-human GITR (TRX-518) and anti-murine GITR (DTA-1 and 2F8) antibodies; 2) their biological activities on primary T cells; and 3) the role of anti-GITR Fc receptor engagement.

Results and Conclusions: Alanine scanning, amino acid substitutions and flow cytometry experiments were performed to define critical residues important for the binding of all three Mabs to their respective GITR targets. Swaps of 5-8 of these critical residues showed both the loss of murine and the gain of human directed Mab binding when residues critical for TRX18 binding were substituted into murine GITR. Likewise, the substitution of residues critical for DTA-1/2F8 binding into human GITR resulted in loss of TRX518 binding and the gain of 2F8/DTA-1 binding. We then mapped these critical residues to 3D structural models of the murine and human GITR ECDs built by contact-map guided fragment assembly simulations. Regions defined by these critical residues showed a high degree of overlap, indicating structural conservation even without amino acid identity. Thus the epitopes defined by both DTA-1 and 2F8 are homologous to those defined on human GITR by TRX518. In in vitro standard proliferation/suppression assays with effector T cells (Teff) and regulatory T cells (Tregs) isolated from human healthy donors and naïve mice respectively, TRX518, similarly to DTA-1, significantly increased proliferation, activation and pro-inflammatory cytokine production of human and murine Teff cultured alone or in the presence of Tregs. These observations indicate that Mabs directed to similar mouse and human GITR ECD epitopes result in similar biological activities. Lastly, we questioned the importance of a functional Fc domain in anti-GITR Mabs in order to translate these in vitro effects into anti-tumor activity in vivo. To this end, we compared the anti-tumor activity of DTA-1 and its aglycosyl version in the aggressive and poorly immunogenic B16 melanoma model. Both Fc and non-Fc binding DTA-1 were able to control tumor growth. This suggests that the aglycosyl-anti-GITR Mab TRX518 has the potential to be effective in the clinic despite the lack of Fc receptor engagement. These results align with our preliminary results in patients treated with TRX518 monotherapy showing reductions in Treg frequencies both in tumor biopsies and in the periphery. TRX518 would therefore appear to be well-poised for use in the treatment of solid tumor malignancies.

Note: This abstract was not presented at the meeting.

Citation Format: Roberta Zappasodi, Heidi Heath, Yang Zhang, Michael Haas, Min Yang, Christopher Mirabelli, Jedd D. Wolchok, Cyndi Sirard, Walter Newman, Taha Merghoub. GITR cancer immunotherapy: Epitope swapping of anti-GITR TRX518 to inform functional translatability from mouse to human [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2402.