ESTRO 38 Abstract book

S973 ESTRO 38

EP-1797 Skin dose in HDR brachytherapy for breast cancers: our in vivo dosimetry protocol and data analysis S. Fabiani 1 , M. Casale 1 , M. Italiani 1 , M. Muti 1 , E. Maranzano 2 1 “S.Maria” Hospital, Department of Radiation Oncology – Medical Physics Unit, Terni, Italy ; 2 “S.Maria” Hospital, Department of Radiation Oncology, Terni, Italy Purpose or Objective The aim of this work is to evaluate the skin dose in partial breast high dose rate brachytherapy (PB-BRT), developing a novel protocol for in-vivo dosimetry (IVD) and using two different dosimeter types. Material and Methods Currently there is not a general acknowledged constraint for skin dose in PB-BRT. Because of our center's experience (20 PB-BRT patients treated/year) we achieve a good cosmetic outcome if the TPS calculated skin dose is below 55% of the prescription of 32 Gy (4 Gy/fraction, twice daily). The skin dose, calculated by a TPS, is overestimated, assuming a homogeneous water medium and not accounting for the finite patient dimensions. IVD was performed on patients treated with multi-catheter brachytherapy, which involves the placement of 9-15 catheters through the breast. Both thermoluminescent dosimeters (TLDs) and MOSFET detectors were used in the IVD sessions and placed in five specific points on the skin: one close the nipple, two next to the entrance of the catheters (upper breast area) and two next to the exit (lower breast area). After a preliminary study using TLDs (IVD-TLD sessions), we performed MOSFET real time measurements to compare with TLD results and to further improve the estimation of the skin dose. We are currently implementing optical fiber real time measurements to test a new system based on the use of luminescent materials. Results The doses measured by both dosimeters were compared to those calculated by the TPS in the specific regions of interest. For a first group of three patients two IVD-TLD sessions were performed for each of them. A Gaussian fit of the percentage differences between measured and calculated doses yielded a mean overestimation value of 25% by the TPS. For a second group of seventeen patients IVD sessions were performed using MOSFET dosimeters, which were placed on the skin following the previous arrangement of TLDs. For each patient five IVD sessions were performed. Results showed a mean overestimation value of 28% by the TPS. The more accurate doses measured by MOSFETs indicated a different value when the dosimeter was placed on the upper or the lower breast area: the latter showed a smaller overestimation of the skin dose by the TPS. Conclusion We performed IVD on twenty patients, using two different dosimetry systems. We obtained a clear overestimation of skin dose by the TPS using both systems, because the TPS does not take into account the tissue-air interface. The different distribution obtained for MOSFETs placed on the lower breast area, is due to a possible reduction of the tissue-air interface. Preliminary measurements demonstrated the requirement to implement real time dosimeters in the in vivo sessions instead of using TLDs. Quantifying skin dose accurately does contribute to define a reliable constraint in PB-BRT, to achieve not only the tumor control but also a good cosmetic outcome to improve the quality of patient’s life. EP-1798 Calibration of the new RefleXion biology- guided radiotherapy unit in the context of the TRS-483 CoP L. Mirzakhanian 1 , R. Bassalow 2 , D. Zaks 2 , C. Huntzinger 2 , J. Seuntjens 1 1 McGill University, Medical Physics, Montreal, Canada ;

extracting the line dose was evaluated. Cross calibration on reference field (10x10 cm 2 ) and depth (10 cm) was applied on LA48-measured dose values. Results Figures 1 and 2 show Inverse Pyramid profiles for Elekta and Varian linac respectively. TPS beam modelling for head scatter conditions and MLC and jaw transmission are well within the criterion for all energies at depth of 3 cm. The modelling in Eclipse and Pinnacle show larger differences with measurements both at primary field dose region and scattered only regions for high energy and 1 cm depth. Beam modelling in RayStation shows agreement with the measurements within 3%/3mm for all regions, taken the spread-out of 0.5mm shift into account, whereas for the models in Eclipse and Pinnacle a 5%/5mm is met.

Figure 1 and 2.Inverse pyramid line dose profiles at 3- and 1-cm depths. Conclusion The higher electron contamination contribution in RayStation v6.1 modelling did not cause overdosage at shallow depths. Even for high energy and depths as low as 1 cm, the criterion of 3%/3mm was met. Extension of a QA protocol for IMRT beam model validation with dose calculation tests outside the fields at shallow depths is relevant. [1] Code of Practice for the Quality Assurance and Control for Intensity Modulated Radiotherapy -Report 22 of the Netherlands Commission on Radiation Dosimetry - June 2013 [2] Srivastava RP, De Wagter C. The effects of incidence angle on film dosimetry and their consequences in IMRT dose verification. Med Phys. 2012;39:6129-38