Tracking glioblastoma progression after initial resection with minimal reaction-diffusion models.

Area of Science:

• Computational biology
• Mathematical oncology
• Biophysics

Background:

• Glioblastoma multiforme is an aggressive brain cancer.
• Modeling tumor progression is crucial for effective treatment strategies, especially after recurrence.
• Current models require refinement to accurately predict tumor behavior.

Purpose of the Study:

• To develop and validate reaction-diffusion models for glioblastoma multiforme growth.
• To simulate tumor progression in patients with recurrent glioblastoma.
• To assess model parameter variability and predictive accuracy.

Main Methods:

• Utilized Fisher-Kolmogorov and a novel 2-population reaction-diffusion model.
• Simulated models on 3D brain geometries from MRI scans of 10 patients.
• Employed Taguchi sampling for parameter variability analysis across 144 parameter sets.

Main Results:

• Successfully identified model parameters for 9 out of 10 patients.
• Simulated tumors matched at least 40% of observed tumor volume in successful cases.
• Demonstrated significant tumor progression over 1-3 months in patient data.

Conclusions:

• Preliminary models show promise in simulating recurrent glioblastoma growth.
• Model parameterization and initialization require further investigation for improved accuracy.
• This approach offers a foundation for personalized glioblastoma treatment planning.