ESTRO 36 Abstract Book

S522 ESTRO 36 _______________________________________________________________________________________________

R. Fabregat Borrás 1 , S. Ruiz-Arrebola 1 , E. Rodriguez Serafín 1 , M. Fernández Montes 1 , A. García Blanco 2 , J. Cardenal Carro 2 , J.T. Anchuelo Latorre 2 , M. Ferri Molina 2 , A. Kannemann 2 , D. Guirado 3 , P.J. Prada 2 1 Hospital Universitario Marqués de Valdecilla, Radiophysics, Santander, Spain 2 Hospital Universitario Marqués de Valdecilla, Radiation Oncology, Santander, Spain 3 Hospital Universitario San Cecilio, Radiophysics, Granada, Spain Purpose or Objective In vivo dosimetry (IVD) applied to HDR BT treatments allows to monitor real dose delivered to clinically relevant areas. MOSFET detectors are the most suitable devices for this task because of their tiny dimensions, which enables their introduction into identical needles to those used in treatments. However, these type of detectors show responses depending on source-to-detector angle and distance. Mathematical models describing these dependences can be obtained from a correct detector characterization. Applying these models on the measurements should minimize the impact of those dependences, improving precision and accuracy. The purpose of this study was to evaluate clinical data of IVD applied to HDR prostate BT using MOSFET TN-502RDM from Best Medical Canada with the Ir-192 Vr2 source contained in Flexitron aferloader (Nucletron-Elekta) and mathematical models describing those dependences obtained in a previous characterization work. Material and Methods Clinical data were taken from five patients suffering from prostate cancer. One to three measuring points were taken for each patient, where the MOSFET were positioned. Anatomical areas measured were neurovascular bundles, bulbourethral area and periurethral area. Nine measuring points were taken and evaluated. Real time ultrasound image guided technique was used to implant the treatment needles. An additional needle was needed for each measuring point. Oncentra Prostate 4.2.2.4. was used to calculate the treatment planning following a standard procedure. Subsequently, coordinates of measuring points and dwell positions were taken as well as dose contribution of each dwell position to each measuring point. After irradiation, mathematical models were applied on measured dose. Table 1 shows the three models considered, their parameters and the goodness of fit. TPS dose, direct MOSFET measured dose and calculated dose after applying the mathematical models on direct MOSFET dose were evaluated.

Material and Methods A decommissioned rectal retractor was modified by drilling a small hole to allow a microMOSFET to be inserted. The MOSFET was commissioned measuring energy dependence and angular dependence of response for the range of source-MOSFET positions expected in cervix brachytherapy treatments. Standard and conformal cervix plans covering the range of applicator sizes and geometries used in clinical treatments were delivered in a water phantom. The MOSFET was monitored during treatment delivery and measured doses compared to treatment planning system (TPS) calculated doses for the total plan and for ring and inter-uterine tube (IUT) individually. Results Corrections were applied for energy dependence response (6% variation between 1 and 8 cm source-MOSFET positions) and angular dependence of response (up to 8% under response for the largest polar angle of 170°). Total plan measurements agreed with TPS calculated doses within 3.1% - 7.8% for 30° and 60° applicators but measured 16% -24% high for 45° applicators (k=2 uncertainty was estimated as 14% for total plan measurements). Separate analysis of ring and IUT measurements similarly showed good agreement for all cases except the 45° IUT for which measurements were on average 55.3% higher than expected. For the 45° IUT the MOSFET position is directly in line with the source cable and longitudinal source axis based on the source positions assumed by the TPS (see figure). A combination of a small rotation of the source relative to the IUT axis and deviation of the actual source position from the centre of the IUT could explain the measurement difference. To verify this, treatments for the 45° applicator were re- measured with the MOSFET taped to the outside of the rectal retractor in a position that was not aligned to the IUT and measured doses agreed within 8%.

Conclusion In vivo dosimetry for cervix brachytherapy would be feasible if commercial rectal retractors were designed to allow a dosimeter to be inserted. However it is important to avoid dosimeter positions aligned with the source longitudinal axis as this is a region of high dose uncertainty. PO-0943 Evaluation of a recent in vivo dosimetry methodology for HDR prostate BT using MOSFET detectors

Results Figure 1 shows the results normalized to TPS d ose. All measurements suffer an approach to TP S dose after applying the mathematical models. The ave rage value of percentage difference between TPS dose and direct measured dose was 15% while the average percentage difference after applying the mathematical models decreases to 9% without any point exceeding 20%. Estimated global uncertainty associated to these corrected measurements were into the range 3.7-4.3%.