Editors' ChoiceCancer

Modeling the impact of genetic heterogeneity on immunotherapy

See allHide authors and affiliations

Science Translational Medicine  25 Nov 2020:
Vol. 12, Issue 571, eabf7104
DOI: 10.1126/scitranslmed.abf7104


A panel of genetically engineered mouse models recapitulates genotype-driven responses to immunotherapy and uncovers a driver of immunotherapy resistance in ovarian cancer.

Responses to single-agent immune checkpoint blockade (ICB) in high-grade serous tubo-ovarian carcinomas (HGSOC) are below 15%, and combination immunotherapies are urgently needed to improve outcomes. Furthermore, HGSOC is characterized by extensive genomic heterogeneity, and the impact of this heterogeneity on the immune microenviroment is currently unknown. Tumor models that recapitulate genotype-driven effects of disease progression and chemotherapeutic response in the clinic are needed to enhance our understanding of tumor-immune interactions and uncover additional targets to design ICB-based combination therapies.

Iyer et al. developed and characterized a panel of genetically defined HGSOC mouse models to study the impact of tumor genotype on the tumor immune microenvironment and on immunotherapy responses. Mouse models with genomic variations in two commonly altered pathways, the homologous recombination (HR) pathway and the cyclin E1 pathway, were extensively characterized for DNA damage repair efficiency, in vivo growth, and ovarian cancer–specific markers. These models faithfully reproduced common drug responses of the corresponding human genotypes to platinum-based chemotherapy and poly (ADP-ribose) polymerase (PARP) or checkpoint kinase 1 (Chk-1) inhibitor treatments. Analysis of the immune microenvironment in vivo revealed differences in the proportion of cytotoxic T cells and myeloid cells between the HR-defective and cyclin E1–overexpressing models. In addition to characterizing immune cell infiltration in the immunosuppressive microenvironment, cytokines of tumor cells and ascites fluid were also profiled. In the HR-defective model, combination treatment of PARP inhibition with anti–programmed death-ligand 1 (PD-L1) treatment resulted in durable responses and animal survival that was also associated with high expression of an interferon γ–associated inflammatory signature. Treatment responses to triple combination of Chk-1 inhibitors with anti–PD-L1 and anti–cytotoxic T-lymphocyte–associated protein 4 (CTLA4) blockade in three independent tumor clones of the cyclin E1–overexpressing model demonstrated durable response only for a tumor clone that expressed low levels of follistatin. Interestingly, high follistatin expression was also selectively associated with shorter progression-free survival in HGSOC patients with cyclin E1 amplification.

The development of this diverse panel of transplantable immunocompent mouse models provides an excellent resource to elucidate mechanisms of immunosuppression in ovarian cancer and serves as a discovery platform of combination immunotherapies that are matched to patient tumor genotypes. Future studies are needed to evaluate mechanisms of response and resistance to PARP and Chk1 inhibitor–based combination immunotherapy across a wider panel of genetically diverse models.

Highlighted Article

Stay Connected to Science Translational Medicine

Navigate This Article