Search

Brain metastases (BrM) remain an intractable, deadly complication for advanced melanoma patients and efficient therapeutic strategies are desperately needed. The tumor microenvironment (TME) plays an important role in response to therapy. However, studies addressing the contribution of the TME to therapy efficacy for BrM are lacking, mostly due to limited access to human samples and scarcity of appropriate preclinical models. Here, we describe two novel isogenic immunocompetent BrM models generated by intracardiac injection of UV-induced mouse melanoma cell lines, representative of mutant-RAS human melanoma subtypes. We used these models to test immune checkpoint blockade (ICB) therapy and to interrogate the role of the TME in therapeutic efficacy. To evaluate response, we developed and applied a new machine-learning method to quantify metastatic burden. We investigated the TME by high-parametric flow cytometry and single-cell RNA sequencing (scRNA-seq). We showed that the models have distinct metastatic behaviors, with BR1 being mostly brain tropic and BR3 displaying widespread metastases. Notably, BR1 BrM were sensitive to ICB with a better response to anti-PD-L1/anti-CTLA-4 combination therapy as compared to monotherapies. In contrast, BR3 BrM were resistant to both mono- and combination therapies. Interestingly, we found that ICB efficacy on extracranial BR3 metastases is organ-dependent. Characterization of the BrM immune microenvironment before and after treatment revealed dramatic differences between the models. Untreated BR1 BrM showed significant recruitment of T cells, dendritic cells, and natural killer cells, while neutrophils were enriched in untreated ICB-resistant BR3 BrM. Moreover, we uncovered phenotypically distinct microglia populations exclusively present in ICB-sensitive BR1 BrM that positively correlated with T cell infiltration. Consistent with this finding, scRNA-seq showed upregulation of genes encoding for T cell-attracting chemokines and antigen presentation uniquely in the BR1-associated microglia. Post-treatment analysis of the brain TME highlighted beneficial changes induced by ICB in the responsive BR1 model, including increased recruitment of CD8 T cells with an activated phenotype, while a mild recruitment of exhausted T cells was observed in the resistant BR3 model. Altogether, our data emphasize the importance of interrogating the BrM TME to understand therapeutic response. Our unique BrM models, mirroring the diversity of ICB response observed in patients, provide a robust platform for the much-needed mechanistic studies to optimize BrM therapy. Deciphering the contribution of the newly identified BR1 BrM-associated microglia to ICB efficacy will be crucial to the identification of novel therapeutic targets. Citation Format: Amélie Lopès, Jessica Rappaport, Eva Pérez Guijarro, Quanyi Chen, Emily Wu, Isabella Church, April Huang, Jessica Bridge, Sung Chin, Cari Smith, Charli Gruen, Khiem C. Lam, Romina E. Araya, Antonella Sassano, Chi-Ping Day, Glenn Merlino, Romina S. Goldszmid. Interrogating the role of the immune microenvironment in the response of brain metastases to immunotherapy using new preclinical melanoma models. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5188.

Neutrophils play an important role in cancer progression and both pro- and antitumorigenic functions have been described. However, their role in response to therapy and whether environmental signals modulate their function in the tumor microenvironment (TME) remain unclear. Here, we show that neutrophil content in the TME defines the response to chemo- and immunotherapy. Neutrophil depletion or recruitment blockade impaired the response, while overexpression of neutrophil-attracting chemokines rendered non-responder tumors susceptible to therapy. Importantly, we demonstrate that tumor neutrophils are a heterogeneous and dynamic population that differ from those found in blood or bone marrow, and we identified a neutrophil subset that correlates with therapy response. Additionally, we found that mice lacking microbiota have impaired neutrophil recruitment and function associated with reduced therapy efficacy. Mechanistically, administration of a microbiota-derived NOD2 ligand to mice lacking microbiota was sufficient to restore neutrophil dynamics and response to therapy. We confirmed tumor neutrophil heterogeneity in cancer patients and showed that while total neutrophil content had no predictive value for patient’s overall survival, two specific neutrophil subsets associated with better outcome for pancreatic cancer patients. Remarkably, the subset enriched in non-responder germ-free tumors in our preclinical studies associated with a similar poor outcome in colorectal cancer patients. Collectively, our findings highlight the importance of characterizing neutrophil heterogeneity in the TME and the contribution of microbiota in shaping these cells to predict a successful therapy outcome. Supported by NIH Intramural Program

Gut microbiota impacts antitumor immunity and recent studies have shown that manipulating microbiota via fecal transplant can rescue immune checkpoint blockade (ICB) response in refractory melanoma patients. However, not all patients benefited from this approach. We hypothesize that tumor-intrinsic factors (e.g. neoantigen load and immune microenvironment) contribute to the microbiota effect on anticancer response. To test this, we performed preclinical microbiota manipulation studies using dietary intervention. Mice fed high-fiber diet (FD) had improved spontaneous tumor control and response to ICB. Mechanistically, FD-induced changes in microbiota composition skewed the tumor innate immune repertoire towards an antitumor profile marked by increased dendritic cell infiltration. Downstream of the innate compartment, FD reduced the proportion of intratumoral Tregs and exhausted CD8+ T cells. We further assessed the effects of FD across 8 tumor models including EL4 lymphoma, MC38 colon carcinoma, and 6 melanomas (M1–M6) that mimic the variety of human melanoma subtypes. Although FD had a beneficial effect overall, not all models responded to the same extent, and were accompanied by different changes in the tumor immune profile. FD delayed tumor initiation for EL4 and M4, reduced growth rate for M3 and M5, both delayed initiation and reduced growth rate for MC38, M2, and M6, but had no significant effects for M1. Additionally, we found that increasing tumor immunogenicity resulted in a synergistic effect with FD. Our data suggests that tumor-intrinsic factors influence the beneficial effect of microbiota on antitumor immunity. Identifying these factors will help refine the use of microbiota-based therapies in the clinic. This research was supported by the Intramural Research Program of the NIH (CCR-NCI). This research was supported by the Intramural Research Program of the NIH (CCR-NCI).