3D Simulation of Drug Concentration Profiles in the Brain Using a Multi-Injection Site Microcatheter

Abstract

Malignant brain tumors represent a global health challenge due to persistent high mortality rates following current treatment methods and despite significant progress in biomedical technologies. Ineffective tumor treatment is primarily attributed to the highly infiltrative nature of malignant cells, the selective permeability of the blood–brain barrier (BBB), and the limitations of conventional treatment methods, which are often unable to target residual malignant cells, leading to high recurrence rates. This challenge is further amplified by the fact that advanced drug delivery therapies remain at early stages of development and have not yet achieved substantial clinical efficacy. Convection-Enhanced Delivery (“CED”) is an alternative treatment method utilizing pressure-driven transport to directly infuse therapeutic agents to tumor regions. CED achieves this by bypassing the blood–brain barrier, thus yielding higher drug concentrations. Clinical implementation of CED remains limited due to the scarce availability of empirical evidence, which is hard to obtain due to ethical and practical considerations hindering brain experiments. Development of computational model helps to overcome this challenge by providing a testing platform for simulating brain processes. The present thesis investigates whether pressure-driven convection dynamics are more effective at achieving broader and more homogeneous drug distribution within the brain tissue compared to fluid diffusion mechanisms. A comparative analysis of the two mechanisms is conducted by simulating transport processes into a 3D brain structure model. The brain geometry is generated based on real-life patient MRI input using Python and simulation models are performed using the COMSOL Multiphysics program. The thesis also presents parametric analyses of the computational models and assesses the influence of key variables, including infusion pressure, diffusion coefficients, injected drug concentration, injection depth, number of injection sites, and tissue properties. Based on the simulation results, the thesis investigation demonstrate that convection-enhanced delivery substantially increases the volume of targeted tissue, exposed to therapeutically significant drug concentrations when compared against passive diffusion dynamics. Moreover, multi-site infusion methods achieve broader and more homogeneous critical area coverage, highlighting their potential to improve targeting of infiltrative tumor regions. While the thesis provides strong in-silico evidence on the effectiveness of multi-site CED, it does not intent to replace experimental or clinical validation. The thesis conclusions are subject to several constraints, most importantly, the simplified assumptions used as the framework for the simulations, which may not be representative of real-life cases. This is due to the fact that idiosyncratic attributes of each patient add an additional layer of complexity to the optimization of treatments, something that generalized models do not effectively capture. To this extent, the conclusions of this thesis investigation establish a framework for future research on optimizing CED models for patient-specific computational simulations.