Abstract
Brain cancers are notoriously difficult to treat, and the most malignant of these, glioblastoma, is particularly resistant to therapy. Current standard-of-care is immediate surgical resection, followed by rounds of radiotherapy and temozolomide chemotherapy. Few patients respond to this aggressive treatment, resulting in a median survival of 16 months. Glioblastoma is characterized by diffuse invasion of tumor cells from the primary bulk into the surrounding tissue, and it is these invading cancer cells that remain post-therapy to cause cancer recurrence. Many researchers focus solely on the invading cancer cells; yet, it is an emerging theme that the tumor microenvironment, or tissue surrounding the cancer, is important in promoting treatment resistance via multiple mechanisms. The understudied glioblastoma microenvironment is uniquely complex and includes many different components: glial cells, extracellular matrix, soluble factors and biophysical forces. Because this complexity can be difficult and costly to study in vivo, tissue-engineered models can provide a platform for incorporating defined populations of parenchymal cells as well as extracellular matrix to more realistically mimic the tumor microenvironment in vitro.
The overall objective of this dissertation was to use tissue engineering to build a patient-driven and physiologically relevant 3D in vitro model of the glioblastoma tumor microenvironment to study microenvironmental contributions and therapeutic response. Through development of quantitative techniques for histological analysis of the glioblastoma cellular microenvironment, we built statistical models for predicting patient survival from the cellular microenvironment makeup of their tumors. Based on the histological analyses of patient tumor samples, a 3D in vitro model, specifically mimicking the post-resection infiltrative edge of the glioblastoma tumor microenvironment, was designed and optimized. Taking advantage of the tunable nature of the tissue-engineered model, we then demonstrate assessment of multiparametric effects of the microenvironment in vitro, ability to identify microenvironment intercellular signaling targets, as well as therapeutic responses to standard of care, and to clinically relevant chemotherapeutics in vitro and in vivo. In all, this dissertation encompasses work to understand contributions of the cellular microenvironment to glioblastoma malignancy across the spectrum – from analysis of patient tumor samples to in vitro tissue-engineered modeling and in vivo xenograft studies.
