Abstract
Orthotopic glioblastoma xenografts are paramount for evaluating the effect of innovative anti-cancer treatments. In longitudinal studies, tumor growth (or regression) of glioblastoma can only be monitored by noninvasive imaging. For this purpose, bioluminescence imaging (BLI) has gained popularity because of its low cost and easy access. In the context of the development of new nanomedicines for treating glioblastoma, we were using luciferase-expressing GL261 cell lines. Incidentally, using BLI in a specific GL261 glioblastoma model with cells expressing both luciferase and the green fluorescent protein (GL261-luc-GFP), we observed an apparent spontaneous regression. By contrast, the magnetic resonance imaging (MRI) analysis revealed that the tumors were actually growing over time. For other models (GL261 expressing only luciferase and U87 expressing both luciferase and GFP), data from BLI and MRI correlated well. We found that the divergence in results coming from different imaging modalities was not due to the tumor localization nor the penetration depth of light but was rather linked to the instability in luciferase expression in the viral construct used for the GL261-luc-GFP model. In conclusion, the use of multi-modality imaging prevents possible errors in tumor growth evaluation, and checking the stability of luciferase expression is mandatory when using BLI as the sole imaging modality.