INTRODUCTION: Breast cancer (BC) research encompasses not only the study of tumor behavior, but also requires understanding the innumerable interactions that occur between breast cellular components and the extracellular matrix (ECM) in the development and progression of the disease. Despite extensive research, BC continues to be a leading cause of morbidity and mortality for women across the globe. The lack of three-dimensional models that are able to accurately replicate the tumor microenvironment and tumor-associated stroma has resulted in the unsuccessful translation of most preclinical studies into the clinic. It has been suggested that adipose tissue is a drug-metabolizing organ that protects tumor cells from chemotherapeutic agents. We have developed a tissue engineered, three-dimensional, high-throughput, patient specific, biomimetic model of BC that incorporates patient-derived adipocytes, adipose stromal cells, breast-duct organoids, and BC cells, and studied the effect of treatment with the chemotherapeutic agent doxorubicin on BC cell survival. METHODS: Under an approved IRB, breast tissue was acquired from patients and differentially processed to isolate mature adipocytes, stromal cells, and breast organoids which were subsequently co-cultured with cancer cells in a 0.6% type I collagen matrix. Mixtures containing fluorescently tagged MDA-MB-231 or MDA-MB-468 triple negative BC cells at a concentration of 200,000 cells/mL were plated onto 96 well plates. BC cells in plain collagen and biomimetic hydrogels without BC served as controls. Groups composed of three replicates of both Biomimetic and collagen only controls were treated with one time with serially diluted doses of doxorubicin at 0, 0.001, 0.01, 0.1, 1, and 10 uM, in cell culture media and fixed after 3 days in culture, stained, and fluorescently imaged using confocal microscopy. Tissues from three patients were used to test each cell line. Analysis was performed using Imaris™ software. RESULTS: After 3 days of doxorubicin treatment, MDA-MB-231 and MDA-MB-468 cell lines showed increased survival at 10mM doxorubicin (p<0.05) in the biomimetic platform derived from patient tissue, when compared to BC cells cultured in the collagen only controls. Moreover, fluorescent confocal microscopy demonstrated doxorubicin uptake by adipocytes that was directly proportional to increasing doxorubicin concentrations. No increased doxorubicin fluorescence was detected within the collagen-only controls. CONCLUSION: When cultured within a tissue engineered biomimetic platform comprised of patient tissue components, we observed significantly increased survival of two different triple negative breast cancer cell lines (MDA-MB-231 and MDA-MB-468) when compared to cultures in collagen alone. These data clearly demonstrate that the tumor microenvironment and neighboring tissue components modulates the effect of doxorubicin therapy on two different BC cell lines at least partially through adipocyte sequestration of doxorubicin. The ability of our platform to be personalized for individual patients may allow us to not only elucidate how the microenvironment affects BC cells, but also how patient specific tissue impacts the effects of various chemotherapeutic agents.

INTRODUCTION: Breast cancer (BC) is the most common non-skin cancer in females, affecting 12.5% of women throughout their lifetime. Partly due to a lack of models that accurately mimic the tumor microenvironment of individual patients, many preclinical successes fail translation into the clinic. Current two-dimensional models fail to replicate the cellular behaviors and interactions that occur in vivo, and newer three-dimensional systems, though promising, all lack vasculature which is crucial to understanding the mechanisms of tumor cell invasion and metastasis. We have engineered an advanced three-dimensional biomimetic platform derived from patient specific tissues that contains all components of the breast tumor microenvironment (glandular organoids, adipocytes, stromal vascular fraction (SVF)) surrounding engineered vascular structures that can be precisely positioned at predetermined distances from BC spheroids (BCS) allowing for highly novel ex-vivo investigations of the interactions between human tumor cells and blood vessels. METHODS: Polydimethylsiloxane (PDMS) wells were created using 3D-printed molds that include stages for localization of BCS, and putative vascular channels (VC). Type I collagen was neutralized at 0.3% and 0.6% w/v. Red-fluorescent MDA-MB-231 cells were mixed with 0.6% collagen at a density of 40,000 cells/uL; 1uL of the collagen/cancer cell mix was plated on stages of the PDMS molds and allowed to nucleate. A biomimetic platform made with BODIPY stained adipocytes and all other patient-derived tissue components mixed within both concentrations of collagen was plated in the pre-designed well, surrounding the BCS and VC. Twenty-four hours after plating, fluorescently labeled endothelial cells (EC) and smooth muscle cells (SMC) were seeded within the channel at a concentration of 5 million cells/mL. Control constructs were made by generating vascular structures and BCS within a collagen-only matrix, and by creating full biomimetic platforms with vascular channels in the absence of cancer cells. Constructs were cultured for 7 days, formalin-fixed, counterstained with DAPI, and analyzed with confocal microscopy. RESULTS: After 7 days in culture, confocal microscopy revealed successful fabrication of biomimetic platforms with a type I collagen (different concentrations) extracellular matrix containing patient-derived adipocytes, SVF and breast duct organoids. Patent VC lined with fluorescently labeled SMC and EC were visualized within the platform, with vascular walls located within 1mm of the red fluorescent, triple-negative MDA-MB-231 cancer foci. Invasion of BC cells into the surrounding tissue was identified by the presence of red fluorescent cells within the biomimetic platform in constructs containing type I collagen at both 0.3% and 0.6% w/v. Decreased vascular integrity was observed in constructs containing BCS when compared to those without BC cells. CONCLUSION: With the aid of three-dimensional printing technology, we have successfully engineered an advanced, patient-specific, biomimetic platform of the breast cancer microenvironment that not only replicates patient tissue characteristics, but also includes vascular structures and cancer foci that closely resemble early tumors. Observed BC invasion into the surrounding microenvironment, and the platform’s ability to mimic patient specific tissue with extremely high fidelity, make this platform a highly versatile and powerful tool that holds significant promise for diagnostic and therapeutic applications in the study of breast cancer.