Program Information

Dosimetry for Small Moving Lung Tumors: Considerations for Treatment Planning of Free Breathing Lung SBRT

B Guo*, S Hadley, The University of Michigan, Ann Arbor, MI

SU-E-T-563 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

Purpose: Dosimetry for free breathing lung SBRT is challenging due to tissue heterogeneity and tumor motion. This study aimed to quantify the effects of these two factors on target coverage and correlate them with the mean HU in PTV. The validity of a common approach for lung SBRT planning: assigning a uniform HU to PTV was also investigated.
Methods: Seven free breathing lung SBRT plans were generated in Eclipse treatment planning system using AAA algorithm and the average CT images for dose calculation. 50 Gy was prescribed to periphery of PTV in 5 fractions. For each plan, the (1) planned dose (AAA-3D) was compared with (2) Acuros-3D: dose calculated on the average CT images using Acuros algorithm and (3) 4D: dose calculated by copying the plan to each individual phase, calculating doses on each phase using Acuros algorithm and rigidly registering the doses to end-inhale phase. The 4D dose was considered as the true dose the tumor received.
Results: For most cases, PTV coverage was overestimated by planning using AAA algorithm and underestimated by planning using Acuros algorithm. Mean difference in PTV D95 was 0.7 Gy between AAA-3D and 4D (range -1.5 Gy to 2.6 Gy) and -1.4 Gy between Acuros-3D and 4D (range -4.3 Gy to 1.0 Gy). Mean HU in PTV has a moderate correlation with the difference in PTV D95. Lower mean HU in PTV results in larger differences between 3D and 4D doses. Assigning a uniform HU to PTV will increase the difference.
Conclusion: For small moving tumors in free breathing lung SBRT, planning using AAA algorithm results in overestimate of the coverage to the target and planning using Acuros algorithm will underestimate the coverage. Assigning a uniform HU within PTV will further increase the differences so it is not recommended for lung SBRT planning.

Contact Email: