Modeling gamma knife radiosurgical toxicity for multiple brain metastases

Area of Science:

• Radiation Oncology
• Neurosurgery
• Medical Physics

Background:

• Radiation oncology protocols for radiosurgery rely on predicting the risk of radionecrosis using the 12 Gy normal brain volume (V12).
• Accurate prediction of V12 is crucial for setting appropriate dosing criteria and minimizing treatment toxicity.

Purpose of the Study:

• To evaluate tumor surface area (SA) and a simple power-law model using pre-treatment variables for estimating and minimizing radiosurgical toxicity.
• To develop a predictive model for V12 that does not require specialized planning software.

Main Methods:

• Retrospective review of 245 patients with 1217 brain metastases treated with Gamma Knife radiosurgery.
• Univariate, multivariable linear regression, and power-law models were used to identify predictors of V12.
• A power-law model (V12 ~ Rxn^1.5 * LAD^2) was developed and validated on a separate cohort.

Main Results:

• Tumor surface area was the best univariate predictor of V12 (adjR^2 = 0.770).
• The power-law model explained 90% of the variance in V12 across 1217 lesions and 245 patients.
• The model demonstrated strong predictive capability in an independent validation cohort.

Conclusions:

• Tumor surface area is a highly accurate univariate predictor of V12 for metastatic brain lesions.
• A novel pre-treatment power-law model effectively estimates V12, aiding in radionecrosis risk assessment and dose prescription for brain metastases.
• This model facilitates safe dose escalation strategies without reliance on planning software.