A multimodal in vivo MRI study of an intracerebral model of brain metastasis in the rat

S├ębastien Serres, Christopher J. Martin, Manuel Sarmiento Soto, Claire Bristow, James Larkin, Alexander Khrapitchev and Nicola R. Sibson

Poster at Aegean conference: Tumor Microenvironment and Cellular Stress: Signaling, Metabolism, Imaging and Therapeutic Targets, Crete, Greece (2012)


Abstract

Background: It is estimated that approximately one in five cancer patients will develop brain metastases, resulting in significant impairment of brain function and poor patient survival. Our aims here were (i) to establish a clinically relevant model of focal brain metastasis in the rat that would enable serial multimodal magnetic resonance imaging (MRI) and functional MRI (fMRI) studies, and (ii) to establish the progression of MRI-detectable changes in this model together with the associated underlying histopathology. Subsequently, we used this model in a proof-of-principle study to establish the effect of localised brain metastases on functional activity in the whisker-barrel pathway.

Methods: BD-IX rats were injected intracerebrally with different numbers of ENU1536 cells (an N-ethyl-N-nitrosourea-induced mammary adenocarcinoma) in the ventroposterior medial nucleus of the thalamus (VPM), one of the nodes of the barrel cortex pathway. For a 3-week period after intracerebral injections, animals underwent multimodal MRI at 7 Tesla acquiring T1-, T2-, and diffusion-weighted (DW) images, and magnetization transfer ratio (MTR) and regional cerebral blood volume (rCBV) maps. Post-gadolinium T1-weighted images were acquired to assess blood-brain barrier integrity. For functional MRI, animals underwent T2*-weighted imaging to measure the BOLD response in the cortex, a surrogate marker of brain function, to electrical stimulation of the whisker pad.

Results: T2 and post-gadolinium T1-weighted images revealed the presence of tumours in the VPM, with associated tissue damage and increased water diffusion as detected by DW MRI. Interestingly, the necrotic core of the tumour mass could be differentiated from surrounding active metastases owing to a reduction in tissue water diffusion and hypointensity on T2 weighted images. Increased rCBV in the area of the tumour revealed blood vessel dilation probably associated with greater metabolic activity in the tumour environment. MRI detectable changes associated with brain metastases were correlated with histolological changes. At day 21 after injection, the area of functional activation in the cortex, measured by fMRI, in response to whisker pad stimulation was significantly reduced for the hemisphere affected by the tumour compared to the control hemisphere. The degree of reduction in this functional response appeared to be dependent on the size of the tumour in the VPM.

Conclusion: In this study, we have established (i) a clinically relevant model of brain metastasis, (ii) new MRI biomarkers of brain metastasis reflecting changes in the tumour microenvironment, and (iii) the effect of localised brain metastases on functional activation of the whisker-barrel pathway.

Perspectives: Future work will use the model and approaches described here to better understand the metabolic microenvironment of brain metastases and changes associated with tumour progression. Ultimately, these methods may provide sensitive means of following therapeutic outcome.