Lijun Ma1, Arjun Sahgal2, David A Larson3, Dilini Pinnaduwage3, Shannon Fogh3, Igor Barani3, Jean Nakamura3, Michael McDermott3, Penny Sneed3. 1. Department of Radiation Oncology, University of California, San Francisco, California. Electronic address: lijunma@radonc.ucsf.edu. 2. Department of Radiation Oncology, Sunnybrook Health Sciences Odette Cancer Centre, University of Toronto, Toronto, Ontario, Canada. 3. Department of Radiation Oncology, University of California, San Francisco, California.
Abstract
PURPOSE: To investigate how millimeter-level margins beyond the gross tumor volume (GTV) impact peripheral normal brain tissue sparing for Gamma Knife radiosurgery. METHODS AND MATERIALS: A mathematical formula was derived to predict the peripheral isodose volume, such as the 12-Gy isodose volume, with increasing margins by millimeters. The empirical parameters of the formula were derived from a cohort of brain tumor and surgical tumor resection cavity cases (n=15) treated with the Gamma Knife Perfexion. This was done by first adding margins from 0.5 to 3.0 mm to each individual target and then creating for each expanded target a series of treatment plans of nearly identical quality as the original plan. Finally, the formula was integrated with a published logistic regression model to estimate the treatment-induced complication rate for stereotactic radiosurgery when millimeter-level margins are added. RESULTS: Confirmatory correlation between the nominal target radius (ie, RT) and commonly used maximum target size was found for the studied cases, except for a few outliers. The peripheral isodose volume such as the 12-Gy volume was found to increase exponentially with increasing Δ/RT, where Δ is the margin size. Such a curve fitted the data (logarithmic regression, R(2) >0.99), and the 12-Gy isodose volume was shown to increase steeply with a 0.5- to 3.0-mm margin applied to a target. For example, a 2-mm margin on average resulted in an increase of 55% ± 16% in the 12-Gy volume; this corresponded to an increase in the symptomatic necrosis rate of 6% to 25%, depending on the Δ/RT values for the target. CONCLUSIONS: Millimeter-level margins beyond the GTV significantly impact peripheral normal brain sparing and should be applied with caution. Our model provides a rapid estimate of such an effect, particularly for large and/or irregularly shaped targets.
PURPOSE: To investigate how millimeter-level margins beyond the gross tumor volume (GTV) impact peripheral normal brain tissue sparing for Gamma Knife radiosurgery. METHODS AND MATERIALS: A mathematical formula was derived to predict the peripheral isodose volume, such as the 12-Gy isodose volume, with increasing margins by millimeters. The empirical parameters of the formula were derived from a cohort of brain tumor and surgical tumor resection cavity cases (n=15) treated with the Gamma Knife Perfexion. This was done by first adding margins from 0.5 to 3.0 mm to each individual target and then creating for each expanded target a series of treatment plans of nearly identical quality as the original plan. Finally, the formula was integrated with a published logistic regression model to estimate the treatment-induced complication rate for stereotactic radiosurgery when millimeter-level margins are added. RESULTS: Confirmatory correlation between the nominal target radius (ie, RT) and commonly used maximum target size was found for the studied cases, except for a few outliers. The peripheral isodose volume such as the 12-Gy volume was found to increase exponentially with increasing Δ/RT, where Δ is the margin size. Such a curve fitted the data (logarithmic regression, R(2) >0.99), and the 12-Gy isodose volume was shown to increase steeply with a 0.5- to 3.0-mm margin applied to a target. For example, a 2-mm margin on average resulted in an increase of 55% ± 16% in the 12-Gy volume; this corresponded to an increase in the symptomatic necrosis rate of 6% to 25%, depending on the Δ/RT values for the target. CONCLUSIONS: Millimeter-level margins beyond the GTV significantly impact peripheral normal brain sparing and should be applied with caution. Our model provides a rapid estimate of such an effect, particularly for large and/or irregularly shaped targets.
Authors: Danushka S Seneviratne; Laura A Vallow; Austin Hadley; Timothy D Malouff; William C Stross; Steven Herchko; Deanna H Pafundi; Daniel M Trifiletti; Jennifer L Peterson Journal: J Radiosurg SBRT Date: 2020
Authors: Jacob S Buatti; John M Buatti; Sridhar Yaddanapudi; Edward C Pennington; Dongxu Wang; Brandie Gross; Joël J St-Aubin; Daniel E Hyer; Mark C Smith; Ryan T Flynn Journal: J Appl Clin Med Phys Date: 2020-11-18 Impact factor: 2.102