Glioblastoma spheroids produce infiltrative gliomas in the rat brainstem.

Purpose: Diffuse intrinsic pontine glioma (DIPG) is universally fatal without proven therapy other than radiation therapy for palliation. Representative animal models will play an essential role in the preclinical stage of future therapy development. To address the shortage of representative models, we created a novel infiltrative brainstem glioma model in rats based on glioblastoma spheroids.

Methods: Cells dissociated from glioblastoma spheroids grown from surgical specimens were implanted into the brainstem of NIH nude rats. Animals were serially assessed clinically and radiographically with magnetic resonance imaging (MRI). Tumors were further characterized using histology, immunohistochemistry, and cytogenetics.

Results: Tumor generation was successful in all animals receiving glioblastoma spheroid cells. The rats survived 17-25 weeks before severe symptoms developed. The tumors showed as diffuse hyperintense lesions on T2-weighted images. Histologically, they demonstrated cellular heterogeneity, and infiltrative and invasive features, with cells engorging vascular structures. The tumors were shown to comprise immature human origin glial tumor cells, with human epidermal growth factor receptor (EGFR) gene amplification and gain.

Conclusions: This study showed that cells from glioblastoma spheroids produced infiltrative gliomas in rat brainstem. The rat brainstem gliomas are radiographically and histologically accurate compared to DIPG. These tumors develop over several months that would allow sequential clinical and radiographic assessments of therapeutic interventions. This study demonstrated in principle the feasibility of developing patient-specific animal models based on putative cancer stem cells from biopsy or resection samples.