Are established glioblastoma models suitable to mirror the radiobiology of patient tumors?

Are established glioblastoma models suitable to mirror the radiobiology of patient tumors?

Dietrich, A.; Jakob, A.; von Neubeck, C.; Fursov, A.; Tillner, F.; Baumann, M.; Krause, M.; Bütof, R.


Introduction: Suitable in vivo modelling of Glioblastoma multiforme (GBM) is essential for developing new radio-oncological treatment strategies. On the translational axis from bench to bedside, it is important to have a model which closely reflects the clinical situation. However, it is also essential to investigate clinically relevant endpoints and use models which are reasonable regarding costs and feasible regarding statistically necessary animal numbers. Established cell lines are comprehensively characterized and can be efficiently engrafted in large cohorts of animals. In this project, a panel of five human GBM cell lines (U 87 MG, U 251 MG, A7, LN 229, HGL21) is characterized after subcutaneous and orthotopic xenograft transplantation (take rate, radiosensitivity, histology, putative stem cell markers (SM)) to investigate their potential as suitable GBM models.

Methods: Limiting dilution assays were performed using subcutaneous injection of decreasing cell numbers. Intrinsic radiosensitivity and effectiveness of combined Radiochemotherapy was studied by irradiation of subcutaneous tumors with different dose levels. Take dose 50% (TD50) and tumor control dose 50% (TCD50) were calculated. For orthotopic transplantation, mCherry- or luciferase-positive cell variants were used. Intracranial transplantation was performed with a stereotactic frame. Tumor growth was assessed weekly via optical imaging and contrast-enhanced magnetic resonance imaging. After excision, tumors were analysed histologically (Haematoxylin/Eosin, SM).

Results: Both, TD50 and TCD50 values are low for the five GBM models. One of four investigated models showed no signs of orthotopic tumor growth. Two models grew within 30-60 days to end size but the histological phenotype showed weak analogy to GBM patients. Stem cell marker heterogeneity was high between the models and also differed between the orthotopic and heterotopic transplantation site (e.g. Nestin positivity in U-87 MG ortho: 100%, hetero: ~80%).

Conclusion: The low TD50 values indicate high amounts of cancer initiating cells. However, the surprisingly low TCD50 values are in contrast to the remarkable radioresistance of GBM in patients. Additionally, no investigated model shows a GBM-like histology after orthotopic transplantation. Although xenograft models from established cell lines of other entities very closely mirror the clinical situation, this remains questionable for GBM.

  • Lecture (Conference)
    19th Annual Meeting of the Society for Biological Radiation Research - GBS, 26.-28.09.2016, Erlangen, Deutschland