Title : Comparing the efficacy of treatment regimens for glioblastoma multiforme with real-life MRI data using the COBWEB simulation software
Abstract:
Glioblastoma multiforme (GBM), a Stage IV brain cancer, poses a significant clinical challenge, with less than 5% of patients surviving five years after diagnosis despite aggressive therapies. This poor prognosis is driven by factors such as angiogenesis, the blood-brain barrier, and rapid treatment resistance, which complicate treatment and drive recurrence. The standard Stupp Protocol offers modest survival improvement by 2.5 months, but recurrence remains inevitable, emphasizing the importance of comparing existing and novel therapy regimens. As such, we wanted to compare a new therapeutic approach—Stupp Protocol with localized Temozolomide and anti-CD47 antibody—against traditional therapies. Our work employed agent-based modeling using the COBWEB simulation software to mimic GBM progression and treatment in a virtual brain environment, simulating 400 randomized controlled trials in four treatment groups: control, Stupp Protocol, Stupp Protocol with localized Temozolomide, and Stupp Protocol with Temozolomide-antiCD47 hydrogel. This model accounted for genetic variables, immunological responses, and medication combinations into account, providing insights that typical clinical trials do not. Furthermore, MRI data from the LUMIERE Dataset was utilized to assess our trials, ensuring the robustness of our findings. The results showed that the Stupp Protocol combination with TMZ and anti-CD47 exhibited the highest percentage of healthy cells, and was inversely correlated with tumor-affected areas. The predicted healthy cell proportion (91.6%) was lower than observed (98.7%), suggesting the need for further refinement of the predictive model for more targeted clinical decision-making. Combining TMZ and anti-CD47 hydrogel with the Stupp Protocol dramatically decreased tumor-affected regions compared to other therapies (p<0.05), supporting clinical trial data and demonstrating the effectiveness of this method above current regimens. Our study's novel application of COBWEB simulation and MRI data validation provides useful insights into comparative treatment efficacy, particularly the potential of the Temozolomide-anti-CD47 hydrogel in reducing GBM symptoms, which is critical for improving outcomes and patient quality of life in GBM. This study underscores the transformational potential of innovative therapies in the fight against GBM, emphasizing the importance of personalized treatment strategies informed by advanced computational methods and real-world data validation.
Audience Takeaway Notes:
- Audience members will gain valuable insights into leveraging COBWEB simulations for advancing glioblastoma (GBM) research and treatment. Researchers will discover how COBWEB can overcome traditional study limitations, such as follow-up difficulties, and offer new avenues for designing and expanding research on GBM progression and treatment efficacy in greater detail. For clinicians and oncologists, the study provides a model that can be used to simulate novel, innovative treatment combinations, such as TMZ and anti-CD47, to refine personalized treatment plans, potentially improving patient survival and quality of life. Our findings from 400 simulated patients demonstrate that this chemoimmunotherapy combination not only controls tumor size but reduces symptoms, offering a more effective approach overall. Educators will see the advantages of integrating COBWEB into curricula to enhance students’ computational modeling skills and understanding of complex biological systems. Healthcare policymakers will understand how incorporating COBWEB simulations can streamline treatment development and reduce reliance on costly animal models and human trials with difficult follow-up. The study model’s high accuracy and adaptability, demonstrated through validated MRI data comparisons, make it a powerful, ethical tool for accelerating research by allowing rapid testing of multiple treatment variables in a controlled environment, optimizing treatment strategies, and bridging interdisciplinary gaps between biology, medicine, and computational sciences.
