Abstract Book

S1058

ESTRO 37

EP-1947 Dosimetric comparison of four left breast treatment modalities with concomitant boost. A. Prado 1 , A. Milanés 2 , M. Manzano 1 1 Hospital Universitario 12 de Octubre, Radiofísica y Protección Radiológica, Madrid, Spain 2 Hospital Universitario 12 de Octubre, Servicio de Oncología Radioterápica. Sección de Radiofísica., Madrid, Spain Purpose or Objective To compare left breast treatment plans with concomitant boost employing 3DCRT, 4-field IMRT, 7-field IMRT and VMAT. Material and Methods Treatment plans were created in Eclipse v.11 TPS (Varian Medical Systems, Palo Alto. California) using a 6 MV Varian Unique. AAA (Analytical Anisotropic Algorithm) algorithm was used. DVO v11.0.31 and PRO v11.0.31 optimizers were employed for IMRT and VMAT. A cohort of 10 patients wasselected. The dose prescription was 40.05 Gy for breast PTV and 4.5 Gy for boost PTV in 15 fractions. For each case four different plans were created, one for each planning technique used. For the first three techniques the two main tangential angulations were chosen to be the same. 3DCRT included one reduced segment for each field to avoid hot spots. 4F IMRT included two more oblique fields angled 30º from the tangential angulations. 7F IMRT included five more fields located in an equally spaced angular distribution between the tangentials. Two partial arcs were used for VMAT. Start and finish angles coincided with those utilized for tangentials. Homogeneity index (HI) and Paddick conformity factor (CF) were evaluated for breast and boost PTVs. HI was defined as (D 2% -D 98% )/ D 50% . CF was defined as V c 2 /(V·V 100 ) where V c stands for 100% isodose volume contained inside the PTV, V is the PTV volume and V 100 is the 100% isodose volume. For both lungs D avg , V 5Gy and V 20Gy were compared. Regarding the heart, V 23Gy, D max and D avg were considered. D max and D avg were obtained for the contralateral breast. A two-tailed t-test was carried out to elucidate whether the discrepancies obtained were significant. Results Results obtained for OARs and PTVs are shown in tables 1 and 2. VMAT diminished heart V 23 and D max (p<0.022 and p<0.013). Ipsilateral lung V 5 and D avg were reduced using 3DCRT (p<3.4·10 -4 ), the improvement seen in V 20 by using VMAT being insignificant. Contralateral lung V 5 can be reduced by using 3DCRT or 4F IMRT (p<1.5·10 -4 ). 3DCRT achieved by far the lowest D max and D avg (p<0.0015 and p<3.35·10 -7 ). The best values obtained for HI and CF were achieved for 7F IMRT (p<6.6·10 -4 and p< 1.2·10 -5 ). The boost PTV CF is significantly better for 7F IMRT and VMAT (p<0.011) with respect to 4F IMRT and 3DCRT.

Table 2: Results obtained for HI and CF for breast and boost PTVs. Conclusion A dosimetric comparison based on well-known indexes was performed for ten breast patients with concomitant boost. For breast PTV a better CF was achieved with VMAT or 7F IMRT whether 3DCRT provided the lowest CF. The same thing was observed with respect to the boost PTV. HI was also better for 7F IMRT. For OARs sparing it was shown that 3DCRT provided the lowest values for contralateral breast and both lungs. VMAT spared the heart significantly better than other techniques. EP-1948 Craniospinal irradiation without geometrical beam overlap: An analytical solution. A. Ferrando 1 , A. Prado 2 , R. Díaz 1 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 In craniospinal irradiation it is mandatory to verify adjacent overlapping beams if over or underdosed regions are to be avoided. In this work a method for craniospinal treatment with an exact geometrical match of beam divergences is presented. Material and Methods The patient is set over the treatment couch in a cranio- caudal decubitus prono position and hyperextended neck. So as to immobilize the patients a thermoplastic mask is utilized. To design the geometry two treatment planning systems (TPS) were used: Varian Eclipse v11.0 and Elekta Oncentra ® v. 4.1. The cranial region was treated with two lateral 6 MV beams with gantry angles of 90º and 270º. The isocenter was placed in the center of the cranial region. The collimator was set to an angle θ 0 , the inferior jaw aperture was set to L 0 to avoid shoulder irradiation and the couch angle was set to RISO= arctg[L 0 /(cosθ 0 ·SAD)] (gantry=270º). The opposed field (90º) is adjusted using mirror angles to match beam divergences (figure 1).

Table 1: Results for OAR constraints evaluated for different techniques.