A predictive model for Gamma Knife intermediate dose spill: R50%Analytic-GK
Ivan L Cordrey1, Sare Kucuk2, Chester Ramsey3
1Thompson Cancer Survival Center, Cumberland Medical Center, Crossville, Tennessee, USA.
Journal of Applied Clinical Medical Physics
|November 29, 2024
Summary
A new model, R50%Analytic-GK, predicts intermediate dose spill in Gamma Knife (GK) stereotactic radiosurgery (SRS). This model helps assess treatment plans by providing a benchmark for dose spill, crucial for minimizing radionecrosis risk.
Area of Science:
- Medical Physics
- Radiation Oncology
- Neurosurgery
Background:
- Minimizing intermediate dose spill is critical in brain stereotactic radiosurgery (SRS) to reduce the risk of radionecrosis.
- R50% is a key metric for quantifying intermediate dose spill.
- A predictive model, R50%Analytic, exists for VMAT-delivered SRS.
Purpose of the Study:
- To extend the R50%Analytic model to Gamma Knife (GK) delivered SRS, creating the R50%Analytic-GK model.
- To predict intermediate dose spill in GK Icon treatments.
- To provide a benchmark for assessing SRS treatment plans.
Main Methods:
- Developed the R50%Analytic-GK model using phantom calculations on spherical targets (0.001–44 cm³).
- Validated the model against clinical data from 18 brain metastasis cases treated with GK Icon.
- Analyzed 35 targets (0.011–27.4 cm³) comparing predicted R50%Analytic-GK with clinically achieved R50%Clinical.
Main Results:
- R50%Analytic-GK generally predicted a lower bound for R50%Clinical.
- The median difference (R50%Clinical - R50%Analytic-GK) was 0.92, confirming the lower bound nature.
- The model accurately reflected intermediate dose spill characteristics in clinical cases, with outliers attributed to planning complexities.
Conclusions:
- The R50%Analytic-GK model successfully extends the R50%Analytic framework to GK-delivered SRS.
- It offers a reliable method for predicting intermediate dose spill in GK Icon treatments.
- The model serves as a valuable tool for evaluating SRS treatment plan quality.


