The Circuitous Journey of Cancer Cells to the Leptomeninges

Abstract

Cancer cell metastasis to the leptomeninges (LM) is a rapidly fatal disease complication. Despite recent advances in treating brain parenchymal metastasis, standard of care for leptomeningeal disease (LMD) has remained essentially unchanged for decades. The paucity of targeted molecular therapies to treat LMD is attributed to our poor understanding of the molecular mechanisms governing LM invasion. We previously showed that acute lymphoblastic leukemia (ALL) cells can invade the LM by abluminal migration along emissary vessels that connect vertebral/calvarial bone marrow to meninges. This process is dependent on ALL cell integrin α6 engagement with vascular basement membrane laminin. To study whether bone metastatic breast cancer could invade the LM along this pathway, we developed a new syngeneic mouse model of breast cancer LMD and also utilized the xenograft 1833 breast cancer model in which intracardiac engrafted mice develop bone metastasis and LMD. We used these models in combination with cutting-edge techniques including intravital microscopy (IVM) of the LM and high-resolution microCT to study LMD at a single-cell level. Integrin α6 deletion in breast cancer cells (BCCs) resulted in prolonged survival and a decrease in LMD burden in both a xenograft and syngeneic mouse model. Histologic analysis, tissue immunofluorescence, and microCT of tumored mice revealed BCCs in transit along the abluminal surface of emissary vessels connecting the bone marrow and LM. IVM of LM also revealed that >95% of BCCs co-localized with CSFR1+ meningeal macrophages. We found evidence of crosstalk between BCCs and LM macrophages, as breast cancer invasion led to a two-fold upregulation of GDNF, a key neurotrophic factor released by macrophages in response to neuronal injury. In support of an important functional role for BCC-macrophage interactions in tumor cell survival, specific ablation of this CSFR1+ population in vivo markedly reduced immune cell derived-GDNF and resulted in prolonged LMD-free survival. In our previously published models of ALL we found that emissary vessel trafficking is dependent on PI3Kδ regulation of integrin α6 expression and activation of cellular migration pathways; however, ALL cells did not rely on PI3Kδ signaling for growth. We therefore explored the effect of targeting multiple PI3K isoforms with the pan-PI3K inhibitor, copanlisib. Using multiple mouse models of ALL we found that copanlisib treatment induced a growth arrest in ALL cells thus reducing systemic disease burden in mice. Moreover, we found that copanlisib sensitized ALL cells to chemotherapy and reduced systemic and CNS disease burden by inhibiting PI3K/Akt-dependent survival pathways activated upon cellular stress. Taken together, our data suggest that BCCs co-opt neuronal pathfinding mechanisms and resident macrophages to efficiently invade and thrive within the LM niche. Our data also reveal the promising, multifaceted potential of pan-PI3Ki for ALL CNS prophylaxis, systemic disease control, and chemosensitization.