ESTRO 36 Abstract Book

S448 ESTRO 36 _______________________________________________________________________________________________

Purpose or Objective Proton therapy, with or without breathing control techniques, may be used to reduce the cardiopulmonary burden in breast cancer radiotherapy. However, such studies typically assume a constant RBE of 1.1 for protons. This study aims to assess the impact of using a variable RBE in breast proton radiotherapy and to evaluate the sensitivity to respiratory motion when no breathing control is applied. Material and Methods Tangential photon IMRT and 3-fields proton IMPT plans for breast radiotherapy were generated for twelve patients, both on a free-breathing (FB) CT and on a CT using breath- hold-at-inhalation (BHI). 2 Gy(RBE) per fraction in 25 fractions were planned to the whole breast. The physical proton dose was optimized assuming a constant RBE of 1.1. Besides the constant RBE of 1.1, the variable RBE-model by Wedenberg et al. (2013), assuming an α/β of 3.5 Gy for the CTV/PTV and 3 Gy for the OARs, was used for plan evaluation. Subsequently, the FB plans were recalculated on the CT images of the two extreme phases (inhale and exhale) to evaluate the sensitivity of a treatment delivery without breathing control. Results All photon and constant RBE proton plans met the clinical goals with similar target coverage. The target conformity and homogeneity of the proton plans were superior to the photon plans. The plan quality was generally independent on whether the FB or BHI CT-scan was used. However, if the heart was close to the target, the BHI plan lowered the dose to the left anterior descending (LAD) artery in most cases. Applying the variable RBE-model resulted in an average of the mean RBE of 1.18 for the PTV and also increased the heterogeneity. The predicted RBE values in the OARs were also substantially higher than 1.1. However, due to the low physical doses, this is expected to have a minor impact. The dosimetric parameters for the BHI plans are shown in Table 1. The recalculation of the FB plans on the extreme phases generally resulted in minor differences for the CTV coverage and OAR doses for the proton plans. Small CTV volumes may, however, receive a slightly lower dose for the recalculated photon FB plans. The ranges of dosimetric parameters for the FB plan for one patient are shown in Table 2.

Conclusion The use of the variable RBE-model results in substantially higher predicted doses to the CTV compared to the constant 1.1, due to the low α/β associated with breast cancer. Substantially higher RBE values are also predicted for the OARs. This decreases the potential benefit with protons, but could probably be neglected in cases where the physical doses are low. However, if e.g. the LAD is close to the target this could lead to substantially higher predicted doses. The variable RBE could therefore be of importance in certain cases when employing a NTCP model based comparison between proton and photon plans. PO-0833 Reducing small bowel dose for cervical cancer using IMPT and target tailoring in treatment planning P. De Boer 1 , A.J.A.J. Van de Schoot 1 , H. Westerveld 1 , M. Smit 1 , M.R. Buist 2 , A. Bel 1 , C.R.N. Rasch 1 , L.J.A. Stalpers 1 1 Academic Medical Center, Radiation Oncology, Amsterdam, The Netherlands 2 Academic Medical Center, Gynaecology and Obstetrics, Amsterdam, The Netherlands Purpose or Objective Current radiotherapy standards for cervical cancer patients lead to irradiation of large bowel volumes and bladder during external beam radiotherapy (EBRT). Highly conformal techniques such as IMRT, arc-rotation therapy and image guided adaptive radiotherapy (IGART) have resulted in considerable reduction in volume to organs at risk (OARs), but there remains room for further improvement. We previously showed that cervical invasion into the uterine corpus assessed by MRI correlates well to pathological invasion [1]. In the present study we wish to investigate the potential clinical benefit from target tailoring by excluding the tumor free proximal part of the uterus during IGART. Furthermore, we compare this benefit with the advantage of an improved dose conformity by intensity-modulated proton therapy (IMPT). Material and Methods Diagnostic MRIs and planning-CTs from eleven patients with locally advanced cervical cancer were used; all previously had photon radiotherapy and a substantial (>4 cm) tumor-free part of the proximal uterus as visualized by MRI. IGART and robustly optimized IMPT plans were generated for both conventional target volumes (including the entire uterus), and MRI-based target tailoring (excluding the non-invaded proximal part of the uterus), which yielded four treatment plans per patient. For each plan, V 15Gy , V 30Gy , V 45Gy and D mean for bladder, sigmoid, rectum and bowel bag were compared. The clinical benefit of either and both approaches were estimated by calculating the normal tissue complication probability (NTCP) for at least grade II acute small bowel toxicity.