SU-E-T-644: Incorporating High Dose Inhomogeneity into SBRT Treatment Planning

By encouraging greater dose inhomogeneity in the PTV, SBRT allows higher doses to be delivered to the target, and thus improves local control and treatment outcome. This work aims to incorporate the high dose inhomogeneity constraint into SBRT treatment planning in safer manner.

Two approaches are proposed to achieve high dose inhomogeneity: (1) implicitly reduce dose uniformity by increasing the error tolerance for the PTV; (2) explicitly add specific localized information by introducing the structure, located at 5 mm to 1 cm inside the PTV, and prescribe a higher dose. We formulate the inverse planning problem in a total-variation minimization framework to reduce the complexity of the fluence maps. The optimization problem was solved with a fast and efficient first-order method called TFOCS. For validation, 15-beam IMRT plans were generated for a lung SBRT case, where normal tissue constrains are directly taken from the RTOG lung SBRT protocol 0915. The inhomogeneous plans generated from the two approaches were compared with a homogeneous plan in terms of dose distributions.

Both inhomogenous plans delivered higher doses to the target compared with the homogenous plan, with a ∼10% increase in max dose. The implicit approach increased the high dose spillage (outside PTV) to 0.52%, leading to more tissue damage. On the other hand, the explicit approach maintained the high dose spillage at a lower level of 0.18%, similar to that of the homogenous plan. Moreover, with local information of the PTV, the explicit approach allocated the hot spots around the center of the PTV, ensuring the safety and effectiveness of SBRT.

The proposed method using specific localized information inside the PTV can safely and effectively deliver high doses to the target and lead to improved local control.