Abstract
In murine models, patient derived orthotopic xenografts (PDOXs) generated from neurosphere cultures of human glioblastomas (GBMs) are unmatched in representing the molecular heterogeneity of the disease. However, intra-axial tumors with blood-brain barrier (BBB) breakdown visible on magnetic resonance imaging (MRI) are rare in these models and, thus, phenotypic recapitulation of human tumors is typically absent. We hypothesized that this absence lay in the interaction between tumor and stroma and that manipulation of the tumor stromal composition might produce a consistent BBB breakdown. Thirty-two athymic nude rats were implanted intracerebrally with one of two neurosphere cell lines (HF2303 or HF3016) derived from the tumors of two GBM patients demonstrating post-gadolinium enhancement on MRI. Neurospheres (3x105 cells in 10μl) were co-injected with a preparation containing either 5μl Matrigel with high (M+) or low (M-) growth factors, or PBS (M0). When the lesion reached 4 to 5 mm in diameter on imaging, MRI with contrast was used to study BBB breakdown. Two animals with only extra-axial tumors were excluded from analysis, and the intra-axial tumors were graded for radiological evidence of BBB breakdown on a scale of 0 to 4, with 0 being no evidence, and 4 being florid enhancement. A two-way ordinal logistic regression model was fit with treatment factors (three levels: M0, M+ and M-) and cell line (two levels: HF2303 and HF3016). A Wilcoxon Rank Sum test analysis within each cell line demonstrated a significant BBB breakdown effect with Matrigel in HF3016 (N=16, p=0.004), but not for HF2303 (N=16, p=0.328). By producing phenotypic conditions closer to the human glioma pathology, a reliable BBB breakdown in PDOX models represents a significant advance in assessing experimental therapies. However, as demonstrated herein, the effect of stromal enhancement is cell-line dependent, and the biological correlates that accompany BBB breakdown still need to be described.