The Impact of Low-Density Media on Dose Calculations for Lung Stereotactic Body Radiation Therapy
Abstract
Purpose
To investigate dose calculations in low-density media and identify potential clinical impacts on lung stereotactic body radiation therapy (SBRT) patients.
Methods
Five patients previously treated with lung SBRT prescriptions of 4800 cGy in four fractions and dose calculated with AcurosXB were identified. Each patient’s computed tomography (CT) simulation, structure set, treatment plan, and planned dose were anonymized and exported. Planning target volume (PTV) voxel dose and HU correlation were compared using Spearman’s rank correlation coefficient for all patients combined, and each patient independently. Each patient’s CT simulation was copied four times. Voxels within the PTV with an HU below -200 were modified to -600 HU, -800 HU, -900 HU, and -975 HU. New 6X, 6FFF, and 10FFF treatment plans with a prescription dose of 4800 cGy in four fractions were created for each modified HU value scan using both the Anisotropic Analytical Algorithm (AAA) and AcurosXB. Plans were normalized such that 95% of the target volume receives 100% of the prescribed dose. Optimization reduced hotspots in the PTV below 130%. The mean dose to the ipsilateral lung excluding the PTV was recorded.
Results
There is a statistically significant (p < 0.01) non-linear, positive correlation between PTV voxel HU and dose in the combined patient data, and in four out of five patients independently, despite these plans being optimized for uniform dose to the PTV. The median percent change in the mean dose to the ipsilateral lung excluding the PTV from the -600 HU plans to the -975 HU plans calculated with AcurosXB are 28% (6X), 23% (6FFF), and 21% (10FFF). When calculated with AAA changes were within ± 2% on average.
Conclusion
Low-density media in a PTV absorbs less dose and requires an increased fluence to achieve prescribed goals. This additional fluence can increase lung dose.