Computational simulation of convection-enhanced drug delivery in the non-human primate brainstem: a simple model predicting the drug distribution

Abstract

Objectives: Convection-enhanced delivery (CED) is a technique that delivers therapeutic agents directly and effectively into the brain parenchyma. Application of CED is now under investigation as a new treatment for various diseases. Diffuse brainstem glioma is one of the important candidates that could be targeted with CED. Especially when targeting brainstem lesions, prediction of drug distribution prior to CED will be necessary. This study evaluated the computational simulation of CED in the primate brainstem using a simplified model.

Methods: Three in vivo experiments infusing gadolinium solution into the non-human primate brainstem were analyzed. T1-weighted magnetic resonance (MR) images were acquired during infusion of a total of 300 μl gadolinium solution. Computational simulation reconstructed the surface geometry of the brainstem from the MR images. The volume of the whole structure was meshed by grid generating software. Under the assumptions that the brainstem surface was rigid and the interior was filled with cerebrospinal fluid, the equations of continuity and Darcy’s law were solved within a computational fluid dynamics package using a finite volume method. The results of computational simulations were compared with those of the in vivo experiments.

Results: The distribution volume (V_d) in the simulations corresponded well with the in vivo experiments. Under the condition without massive ‘catheter back flow’, computational simulations predicted almost 70% of the V_d of the in vivo experiments.

Conclusions: The simplified computational simulations were consistent with the experiments in vivo. The methodology used in this study can be applied to predict convective drug distribution in the primate brainstem.

Keywords:

• Brainstem
• Computational fluid dynamics
• Convection-enhanced delivery
• Magnetic resonance imaging

This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology in Japan to T.T. Computational simulation was performed using the supercomputer system at the Advanced Fluid Information Research Center, Institute Fluid Science, Tohoku University.