Advanced models of lobular breast cancer metastasis capture clinical organ tropism, endocrine response, and bone remodeling

Abstract

Patients with invasive lobular carcinoma of the breast (ILC) are at high risk of long-term recurrence and metastatic progression with poor prognoses due to delayed detection and treatment-refractory disease. Unfortunately, few models are available to investigate metastatic ILC (mILC) and understand the unique metastatic patterns and phenotypes, including abdominal metastases, leptomeningeal disease, and mixed osteosclerotic/lytic bone metastases. Therefore, we expanded upon the previously established mammary intraductal (MIND) cell line xenograft model by supplementing mice with low-dose estradiol to promote disease progression. We observed spontaneous multi-organ spread from the mammary gland to common and mILC-specific tissues, with micro-metastatic disease as early as 12 weeks post-engraftment and macro-metastatic disease in 24-30 weeks, without the need for primary tumor resection. Primary and metastatic tumors remain highly endocrine responsive, allowing for the evaluation of novel therapeutics in the setting of disseminated metastasis. Derivative cell lines were isolated from various metastatic lesions, a total of 13 derivates from 7 sites across three hosts, and were found to have shared gene expression changes related to metabolism and intercellular signaling. Focusing on bone-derived variant cells as bone is the most common site for mILC to present, we found that bone-derived variant lines maintain multi-organ metastatic potential upon rechallenge by MIND or intratibial injection, despite increased aggressiveness and maintained endocrine response. Notably, bone lesions from either challenge route showed mixed osteosclerotic/lytic features characteristic to clinical ILC. Accordingly, we found that conditioned medium from ILC cells and the mILC bone-derived variants induce osteoblast differentiation and suppressed osteoclast differentiation in vitro, consistent with their effect on bone remodeling in vivo and in clinical disease. Together, the models developed herein can be utilized to understand the unique metastatic processes of mILC, and to investigate new therapeutic combinations in the setting of endocrine-responsive primary and metastatic ILC.