Intermediate IMRT Dose Planning in Lung 4DCT Data Sets
R Lalonde1*, (1) D3 Oncology Solutions, Pittsburgh, PASU-E-T-660 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose:Recently, planning systems have included provision for intermediate IMRT optimization to correct for deficiencies in optimizer dose models, particularly near low-density heterogeneities. Our purpose is to quantify the effect of intermediate optimization on lung cancer planning, including realistic effects of target motion and setup error.
Methods:IMRT plans were generated in the Varian Eclipse system for 5 lung cancer patients, with and without the use of intermediate optimization. PTV volumes were expanded 0.5 cm radially and 1.0 cm S-I to account for internal target motion. Treatment plan complexity was evaluated by comparing total MUs in each plan, as well as average MLC leaf pair opening (ALPO). PTV dose uniformity index (UI) was calculated as D2%/D98%. Plans were then recalculated for simulated setup errors of up to 5 mm in all directions. For several patients, plans were also recalculated for all phases of a 4DCT data set.
Results:Intermediate dose plans generated more uniform doses for PTV volumes, at the expense of increased total MU and plan complexity. The UI decreased from an average of 1.25 in the original plans to 1.15 in the intermediate plans. Average total MU increased by an average of 54.5% in the intermediate plans, while ALPO decreased by an average of 33%. However, the PTV doses did not reflect the actual doses delivered to the CTV as it moved within low density lung tissue. Actual UI for the CTV volumes with motion averaged 1.11 for the original plans, and varied by only 1% with motion.
Conclusion:IMRT plans calculated on static CT images in lung give an inaccurate estimate of the true variation of dose with motion and setup errors. Modifying this incorrect dose by intermediate dose planning may be of limited value, and comes at the expense of higher MU and longer delivery times.