Abstract 1199: Discovery of subtype-specific and therapy associated effects on the tumor immune microenvironment in pediatric neuroblastoma

Abstract

Abstract Background: Neuroblastoma (NB) is the 3rd most common childhood cancer and accounts for 15% of all pediatric cancer deaths. Recently, immunotherapy using monoclonal antibodies targeting GD2 have improved survival rates for some patients with NB. Unfortunately, this response is not uniform, suggesting an incomplete understanding of the underlying immune biology. Large-scale sequencing of patient tumors have suggested that NB has diverse immune microenvironments (TMEs), which are associated with MYCN-amplification (A) and patient outcomes. These results need to be further validated, specifically to determine whether the infiltrating immune cells can interact with tumor cells and if the TME is impacted by therapeutic pressures. We hypothesized the TME is dynamic, changing with therapy, influenced by molecular subtype, and associated with patient outcomes. Methods: To better understand the NB TME, we obtained 93 clinically annotated tumors from 72 pediatric patients with NB, consisting of high-risk primary and metastatic tumors both pre- and post- chemotherapy. We designed multiplexed antibody panels targeting immune cells and performed either imaging mass cytometry (IMC) (n = 46) or NanoString GeoMx DSP (n = 47). Results: We identified 4 different immune cell types (CD8, CD4, myeloid, and NK cells) and confirmed that MYCN-non amplified (NA) tumors display higher frequencies of lymphocytes including CD4 (p < 0.003) and CD8 (p < 0.005) T-cells. Using nearest neighbors’ analysis, we found that CD4 cells are closer to tumor (p = 6.24E-16) and CD8 T-cells (p = 7.55E-11) in MYCN-NA tumors compared to MYCN-A. However, in MYCN-NA tumors, myeloid cells are also closer to tumor, (p = 1.71E-68), CD4 (p = 4.72E-58), and CD8 T-cells (p = 1.27E-70). Furthermore, we find that all immune cell subsets are further away from tumor cells independently of subtype following exposure to chemotherapy, and that there is a decrease (p = 0.0091) in the number of myeloid cells in MYCN-NA tumors, but not MYCN-A. Interestingly, we also observed an increased expression of the immune checkpoints CTLA-4 (p = 0.043) and the exhaustion marker TIM-3 (p = 2.05E-5), but not PD-1, PD-L1, or PD-L2, post chemotherapy, suggesting blockade of CTLA-4 or TIM-3 could improve response to therapy in these patients. Notably, high expression of CD56, a marker for both NK cells and NB, was associated with increased overall survival, indicating a potential role of NK cells in improving outcomes. Conclusions: Using two independent protein-based profiling methods, we investigated the TME in clinically annotated patient NBs. We discovered that the TME in NB varies with tumor subtype and changes with chemotherapy. We find that there is an upregulation of immune check point markers post chemotherapy. These results can inform future trials to optimize the timing and specificity of novel immunotherapeutic approaches for these high-risk patients. Citation Format: Katherine E. Masih, Zahin Islam, Paul Aiyetan, Ben J. Somerville, Igor B. Kuznetsov, William Bocik, Daniel R. Catchpoole, Jun S. Wei, Javed Khan. Discovery of subtype-specific and therapy associated effects on the tumor immune microenvironment in pediatric neuroblastoma [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 1199.