Comparison of Dose-Volume and Dose-Mass Inverse Optimization in NSCLC
I Mihaylov1*, E Moros2, (1) Rhode Island Hospital / Warren Alpert Medical, PROVIDENCE, RI, (2) H. Lee Moffitt Cancer Center, TAMPA, FLTH-C-137-12 Thursday 10:30AM - 12:30PM Room: 137
To compare dose-volume (Dvh) and dose-mass (Dmh) optimized inverse treatment plans for lung cancers.
A retrospective investigation on fourteen NSCLC cases was performed. For each case two IMRT plans were generated. The plans used the same beam angles, maximum number of segments per plan, minimum segment area and MUs per segment. The only difference was in the Dvh and Dmh quadratic objective cost functions used for inverse optimization. Both plans were normalized such that 95% of the PTV was covered by the same prescription dose. After prescription was achieved doses to organs at risk (OARs) were iteratively lowered until the standard deviation of the dose across the PTV was ~ 3.5%. Plan quality was evaluated by several dose and mass indices (DIs and MIs). A DI/MI is represented by the dose to be delivered to certain volume/mass of an anatomical structure. Tallied DIs/MIs were for cord1%, heart33%, lungs20%, lungs30%, and esophagus50%. Statistical significance in the differences among those indices was quantified with two-tailed paired t-tests.
On average Dmh optimized plans performed better than Dvh plans. Statistically significant dose reduction in heart DI33, lung DI20, lung DI30, esophagus DI50, and cord D1 were 7%, 5%, 6%, 26%, and 27% respectively. The statistically significant differences for heart MI33, lung MI20, lung MI30, esophagus MI50, and cord MI1 were 8%, 6.3%, 6%, 25%, and 27% respectively.
The findings herein indicate that on average Dmh based optimization results in better OAR sparing than Dvh optimization. The results however are case dependent, where Dmh does not always outperform Dvh. Those preliminary studies on a cohort of lung patients suggest that the existence and the degree of OAR sparing with Dmh optimization is location and tumor size dependent on patient-by-patient basis.
Funding Support, Disclosures, and Conflict of Interest: Supported by NIH grant R01 CA163370