Left hippocampus sparing model for glioblastoma radiotherapy by utilizing knowledge-based planning and multi-criteria
Shima Y Tari1,2, Amr Heikal1,2, Connie Le3
1Department of Medical Physics, Cross Cancer Institute, Edmonton, Alberta, Canada.
Journal of Applied Clinical Medical Physics
|February 15, 2025
Summary
This study developed a RapidPlan (RP) model with multi-criteria optimization (MCO) to reduce radiation dose to the left hippocampus in glioblastoma patients. The new model significantly decreased the mean hippocampal dose while preserving treatment effectiveness and sparing other organs.
Area of Science:
- Radiation Oncology
- Medical Physics
- Neuro-oncology
Background:
- Radiotherapy for glioblastoma is associated with neurocognitive decline, potentially linked to the dose delivered to the left hippocampus.
- Minimizing radiation dose to critical structures like the hippocampus is crucial for preserving cognitive function post-treatment.
Purpose of the Study:
- To develop and validate a novel treatment planning strategy using RapidPlan (RP) and multi-criteria optimization (MCO) to spare the left hippocampus in glioblastoma patients.
- To assess the dosimetric improvements and target coverage achieved by the hippocampus-sparing model compared to standard clinical plans.
Main Methods:
- A left hippocampus sparing model (HCS1) was configured using RP based on 97 volumetric modulated arc therapy (VMAT) plans.
- The HCS1 model was further optimized using MCO to create a refined model (HCS2), exploring trade-offs between hippocampal dose reduction and other planning objectives.
- The models were validated on 50 independent VMAT plans for newly diagnosed glioblastoma patients.
Main Results:
- The HCS2 model achieved a 26% reduction in the mean dose to the left hippocampus compared to standard VMAT plans (24.04 Gy vs. 32.65 Gy).
- Target volume coverage (V95%) was maintained at clinically acceptable levels across all plan types (98.80% ± 1.48% for HCS2).
- Mean dose to the brainstem (0.1 cc) was also improved with the HCS2 model (39.29 Gy vs. 45.91 Gy).
Conclusions:
- The combination of RP and MCO is effective in significantly reducing the mean dose to the left hippocampus for glioblastoma patients.
- This approach allows for substantial hippocampal dose sparing without compromising target coverage or the sparing of other organs-at-risk.
- The developed planning strategy offers a promising method to mitigate neurocognitive deficits associated with radiotherapy in glioblastoma treatment.
