ESTRO 37 Abstract book

S1061

ESTRO 37

EP-1950 Prescription isodose implications on SBRT lung treatments with VMAT. G. Pozo 1 , A. Prado 2 , M. Manzano 2 1 Hospital Universitario 12 de Octubre, Servicio de Oncología Radioterápica. Sección de Radiofísica., Madrid, Spain 2 Hospital Universitario 12 de Octubre, Radiofísica y Protección Radiológica, Madrid, Spain Purpose or Objective To compare lung SBRT plans with the same PTV locations and dose prescription if the prescription isodose is set to 95% or to 67%. Material and Methods Treatment plans were generated using Eclipse v.11 TPS (Varian Medical Systems, Palo Alto. California) utilizing a Varian Clinac iX with 6 and 15 MV energies. AAA (Analytical Anisotropic Algorithm) algorithm with 2.5 mm grid size were used. PRO v11.0.31 optimizer was employed for VMAT. All parameters related to the optimization cost functions were kept at same values when changing the prescription isodose. Three non- coplanar partial arcs were used from 30° to 180° with collimation angles were set to 30°, 330° and 30°, respectively. Couch angles were set to -10°, 0° and +10° so as to limit low dose extension. A cohort of 8 patients with laterally located right lung tumours was selected. 60 Gy in fractions of 7.5 Gy/fr was prescribed to every patient. Patient immobilization was achieved by a thermoplastic body mask. For plan evaluation several indexes described in ICRU 91 were used. Paddick conformity index (CI) and D 2% for PTV were considered, and D max and D avg were obtained for the main OARs. Gradient index (GI) was calculated so as to provide dose fall-off information. Average values were obtained for evaluated parameters. A two-tailed Student t-test was performed to elucidate whether the discrepancies are statistically significant or not. Results As a result of changing the prescription isodose from 95% to 67%, D avg of every studied OAR diminished as is shown in table 1. Variations were normalized to 95% prescription isodose value. Consequently, a negative value implies a decrement in the index considered with respect to 95% prescription isodose. Only for lung, bronchial tube and chest wall these differences were statistically significant (p<0.008). D max increased for lungs and chest wall (p<0.04) while it diminished for bronchial tube (p<0.005).

tangential fields, utilizing 6 MV for both internal fields and one external field and 15 MV for another external field. Once the 3DCRT plan was created, the same plan was reproduced employing an estimation of RM given by eq.1., where A i is the amplitude in the i direction; T, T inh and T exh stand for the complete respiratory cycle period, inhalation and exhalation phases periods, respectively.The original plan was divided into 10 subplans. In these plans the isocenter was shifted by a quantity of 1/5·A i , thus simulating different points of the respiratory cycle. MUs are proportional to the time in which the isocenter moves from one position to another.Finally, the plan which took into account the RM was the sum of the 10 isocenter-shifted plans. Values used for breast displacement due to RM were calculated as the mean values of several amplitudes found in the literature, given by (A LAT , A AP , A SI ) = (1, 2.3, 1.6) mm. The complete cycle period T is set to 3.5 s, T exh is calculated as 0.635 · T and T inh =1-T exh . The same procedure was performed multiplying by 5 the amplitudes to quantify the results on a quite unlikely scenario. A dosimetric analysis of the generated treatment plans was performed. For breast PTV homogeneity Index (HI), defined as (D 2% -D 98% )/D avg and Paddick Conformity Factor (FC) were evaluated. For contralateral (left) lung, heart and left breast D 2% was evaluated. For ipsilateral (right) lung, V 20Gy , V 5Gy and D avg were considered. Results Figure 1 summarizes the results for breast PTV. No differences were found in HI or CF considering RM. Moreover, no differences were found when multiplying by five the breathing amplitude.

Figure 1: Results obtained for breast PTV regarding homogeneity index (HI) and Paddick conformity factor (CF). Respiratory amplitudes equations.

Table 1: OAR results for 3DCRT, 3DCRT considering RM and 5 times the RM amplitudes. Table 1 results state that discrepancies between 3DCRT and RM plans were negligible, showing a subtle increase when comparing 3DCRT with a larger amplitude displacement (5A i ). Moreover, for amplitudes wider than 5A i discrepancies greater than 10% in OAR indexes could be found, although these values are too low compared with the organ limits. Conclusion Indexes evaluated do not differ significantly from the initial plan, where RM was not taken into account. These results bring to light the fact that tangential field breast treatments are quite robust with respect to RM. As a conclusion, RM has a negligible influence on 3DCRT breast treatments.

and ΔD avg

results for the main

Table 1: Evaluation of ΔD max

OARs.

Table 2: Relative variations of several indexes evaluated. An improvement in CI value was achieved by utilizing a 67% prescription isodose, hence creating a sharper dose distribution nearby the PTV (a GI reduction). These results are statistically significant for CI (p=0.008) and for GI (p=0.003). Conclusion The use of lower prescription isodose levels in laterally located lung tumours improves conformation to PTV, steepens the dose fall-off and significantly diminishes average dose in lungs, chest wall and bronchial tubes. However, maximum doses are increased in lung and chest wall. A good management of heterogeneities in GTV is desirable due to the stimulation of immunological response and tumour vascular epithelium damage linked