Abstract SY38-03: Microenvironment dictates differential responseof primary tumor versus metastases

Abstract

Abstract Metastasis is the leading cause of death in patients with advanced cancer. Activation of the metastatic cascade allows resilient cancer cells to adapt and progress in new organ microenvironments. Our work focuses on microenvironment-dependent mechanisms that control drug delivery, induce de novo resistance and promote acquired resistance to targeted therapies. By using multidisciplinary approaches with preclinical mouse models, including lymph node and brain metastasis, we gained compelling evidence revealing how stromal and cancer cell biology dictate the response to targeted therapies in the metastatic setting. Vascular and lymphatic vessels contribute to metastatic spread. We devised multiple strategies to normalize tumor vessels in order to enhance drug delivery and decrease immunosuppression (1). Preclinical and clinical studies suggest that lymph node metastases and primary tumors differentially respond to therapies. We used tumor models to address differential responses to the VEGFR2-3 inhibitors cediranib and vandetanib (2). Both compounds reduce primary tumor growth but failed to prevent progression of established lymph node metastases. Using multimodal imaging techniques, we observed that malignant cells in the lymph nodes do not promote angiogenesis but instead co-opt pre-existing vessels (3). This mechanism likely explains why anti-angiogenic therapies are ineffective in patients in the adjuvant setting. We are currently investigated novel co-option mechanisms in syngeneic models of glioblastoma. Brain tumors also display a unique vascular biology that dictates therapeutic response (4). The blood brain/tumor barrier (BBB and BTB) contribute to poor drug delivery in the central nervous system (CNS). Current therapies that control extracranial HER2+ breast cancer (BC) progression fail to inhibit brain metastasis in patients. To determine whether CNS vascular barriers are the sole cause of therapy failure, we treated mice bearing HER2+ BC brain metastasis with BKM120, a small molecular weight inhibitor of PI3K that bypasses the BBB/BTB (5). Despite similar drug distribution and target inhibition in brain and primary tumor lesions, only brain lesions progress under therapy. This phenomenon reflects de novo resistance as BC cells isolated from established brain metastases recover sensitivity to PI3K inhibition in vitro when compared to parental counterparts. In pursuit of potential resistance mechanisms in the brain tumor microenvironment, we found that 1) HER3 expression score is higher in CNS metastases when compared to extracranial metastases (including clinical samples); 2) Dual PI3K and HER3 inhibition significantly improves mouse survival when compared to monotherapies. Our data suggests that the brain tumor microenvironment unleashes multiple barriers, including poor drug delivery and de novo resistance, which prevent efficacy of targeted therapies. Collectively, our findings indicate that blocking the interactions between the tumor microenvironment and cancer cells is necessary to improve therapy regimens designed to treat and prevent metastatic progression.