09.04 Secondary resistance to immunotherapy is associated with death and de-differentiation of activated T cells. (21st September 2022)

• Record Type: Journal Article
• Authors: Qing, C; Ghorani, E; Foster, K; Amann, M; Uddin, I; Beattie, G; Solomon, I; Arce-Vargas, F; Peggs, K; Quezada, S
• Is Part Of: Journal for immunotherapy of cancer. Volume 10(2022)Supplement 1
• Page Start: A5
• Page End: A6
• Publication Date: 2022-09-21
• Subjects: Cancer -- Immunotherapy -- Periodicals; Cancer -- Immunological aspects -- Periodicals; Tumors -- Immunological aspects -- Periodicals; Immunotherapy -- Periodicals; 616.99406105
• DOI: 10.1136/jitc-2022-ITOC9.9
• Languages: English
• ISSNs: 2051-1426

Abstract

Abstract : Background: Immunotherapies have transformed the care of patients with multiple tumor types, but the majority who respond eventually progress after a period of stabilization. The reasons for this are not well known. We set out to explore mechanisms of immunotherapy failure in this setting, using a murine model of melanoma treated with a regulatory T cell (Treg)-depleting antibody combined with a cancer cell vaccine. Materials and Methods: C57BL/6 mice were injected subcutaneously with B16 cells. Treatment with a mouse IgG2a depleting αCD25 antibody was given on day 5, followed by a B16 tumor vaccine (GVAX) given intradermally on days 6, 9, and 12. Tumors were harvested for single-cell RNA sequencing (scRNAseq) or flow cytometry using panels designed to interrogate the activation and differentiation landscape of infiltrating T cells. Results: Combined αCD25/GVAX therapy resulted in three different clinical response patterns - no response, partial response, and secondary resistance, with characteristic immune phenotypes. Amongst partially responsive tumors, about 90% of them relapsed between day 35–45. Reasoning that loss of immune control precedes clinical progression, we characterized the evolution of the immune landscape in pre-relapse tumors. We analyzed stable, partially responding tumors on days 28, 35, and 47. Over time, we found a decrease in the abundance of 4–1BB+TIM-3+TCF7- CD8+ effector memory T cells and Ki67+ CD4 effector cells (Teffs). In parallel, non-activated TCF7+ T cells rose in abundance. Treg abundance also recovered over time. ScRNAseq and scTCRseq analyses of pre-relapse and relapse tumors revealed that non-activated CD4 Teffs accumulating at relapse were transcriptionally equivalent to their activated counterparts at pre-relapse except for the expression of activation related genes. Overlaps were found in CDR3 usage between CD4 activated and non-activated populations at both pre-relapse and relapse, suggesting that the accumulating non-activated CD4 cells had been deactivated. In contrast, little overlap in CDR3 usage was found between CD8 activated and non-activated populations at relapse, indicating that the accumulating non-activated CD8 cells had been replaced by new, non-reactive clones. Additionally, we observed that in pre-relapse tumors, the percentage of Fas+ cells in activated Teffs is higher than that in non-activated Teffs. Blocking Fas/Fasl interactions with an αFasl antibody synergized with αCD25 on stable tumors to prevent relapse. Conclusions: Combined Treg depletion/whole tumor vaccine therapy is effective in a poorly infiltrated B16 model. Most mice that achieve partial response eventually relapse, mimicking what is often seen in human disease. By characterizing the evolution of the immune landscape within partially controlled tumors, we revealed that progression is associated with a loss of immune fitness characterized by deactivation and death of activated infiltrating Teffs. Disclosure Information: C. Qing: None. E. Ghorani: None. K. Foster: None. M. Amann: A. Employment (full or part-time); Significant; Roche. I. Uddin: None. G. Beattie: None. I. Solomon: None. F. Arce-Vargas: None. K. Peggs: None. S. Quezada: None.