(F-233) Interstitial Fluid Flow as a Predictor for Glioblastoma Invasion, Progression, and Recurrence.

Introduction:: Glioblastoma (GBM) remains universally incurable in spite of aggressive standard of care therapy. Specifically, the highly invasive nature of the GBM cancer cells results in disease relapse in more than 80% of cases due to cancer cells being left behind after surgical resection and radiation. Using standard of care dynamic contrast-enhanced MRI (DCE-MRI), we developed a multiplexed quantitative image analysis approach to investigate interstitial fluid flow in glioblastoma, which has been identified as a potential mechanism for glioblastoma cell invasion. We aim to identify the features of interstitial fluid flow that are predictive of the location of cell invasion, progression, and post-surgery recurrence in GBM mouse models and human patient datasets.


Materials and Methods:: Fluid flow features are calculated using the Lymph4D computational tool that

computationally solves for the fluid flow velocity field that best explains the observed change in MRI signal intensity obtained using DCE-MRI. Three in-vivo datasets were analyzed to identify predictive fluid flow features in three GBM invasion scenarios. 1) Individual cancer cell invasion: C57Bl/6 mice were injected with GFP fluorescent GL261 cancer cells. 15 days after injection, a DCE-MRI was obtained and 16 days after injection, the mice were sacrificed and the brains were sectioned and imaged for the fluorescent GL261. These two imaging modalities were then co-registered for correlation between fluid flow dynamics and IHC identified cancer cell invasion. 2) Progression of primary disease: C57Bl/6 mice were injected with GFP fluorescent GL261 cancer cells and DCE-MRI imaging was performed on day 6, 8, and 10 post-injection. The correlation between fluid flow dynamics and regions of radiologic progression over time was determined. 3) Recurrence of post-surgery human GBM: Using MRI data from the Ivy Glioblastoma Atlas Project the correlation between pre-surgical fluid flow dynamics and regions of radiologic recurrence was determined.


Results, Conclusions, and Discussions:: Our analysis identified multiple fluid flow metrics that are predictive of the location of invading cells identified on registered IHC sections. We found that interstitial fluid flow velocity is significantly higher, and the diffusion coefficient is significantly lower in the invasive regions of glioblastomas compared to non-invasive regions. Additionally, we showed that a novel metric identifying interstitial regions that experience interstitial fluid flow that originated within the tumor volume is a predictor of invasion. Initial results for the longitudinal studies and human dataset analysis suggest that these features are also predictors of in-vivo tumor progression and can predict the general location of recurrence in post-resection GBM human patients.



Our study provides a broad analysis of interstitial fluid flow in glioblastoma invasion using DCE MRI with multiplexed quantitative image analysis in three in-vivo disease scenarios. Together, these studies suggest that data available from clinical standard imaging could guide physicians to aggressively target certain areas with surgery or anti-tumor chemotherapies, potentially delaying or preventing the fatal recurrence of disease.






Acknowledgements (Optional): :

References (Optional): :