ESTRO 37 Abstract book

S1192

ESTRO 37

Mosci 4 , F. Valvo 5 , R. Orecchia 6,7 , M. Ciocca 1 1 CNAO Foundation, Medical Physics Unit, Pavia, Italy 2 Heidelberg Ion Beam Therapy Center - HIT, Medical Physics Unit, Heidelberg, Germany 3 CNAO Foundation, Clinical Radiotherapy Unit, Pavia, Italy 4 Galliera Hospital, Ocular Oncology Center, Genoa, Italy 5 CNAO Foundation, Clinical Directorate, Pavia, Italy 6 CNAO Foundation, Scientific Directorate, Pavia, Italy 7 European Institute of Oncology - IEO, Scientific Directorate, Milan, Italy Purpose or Objective Only few centers worldwide (around 10) treat intraocular tumors with proton therapy, all of them with a dedicated beamline. Our Centre is a synchrotron-based hadrontherapy facility equipped with fixed beamlines. Proton and carbon ion treatments are delivered with pencil beam scanning modality, with range in water from 3 to 32 cm. Recently, our general-purpose proton beamline was adapted to treat also ocular diseases. This work describes the design and the main dosimetric properties of this new proton eyeline. Material and Methods A 3 cm water-equivalent range shifter (RS) was placed along the proton beamline to shift the minimum beam penetration at shallower depths. Monte Carlo (MC) FLUKA code simulations were performed to optimize the position of the RS and patient-specific collimator, in order to achieve sharp lateral dose gradients. Lateral dose profiles were then measured with radiochromic EBT3 films to evaluate the dose homogeneity and lateral penumbra width at several depths. Different beam scanning patterns were tested. Fine adjustment of beam range was achieved using thin PMMA additional sheets. Eye- dedicated beam settings were implemented in the ocular energy range to decrease the treatment delivery (and gazing) time. The set of low energy proton beams was selected to have 1 mm Bragg Peak separation from 0 to 31 mm water-equivalent depth. Depth-dose distributions (DDDs) were measured with the Peakfinder system. To obtain uniform dose distributions, i.e. Spread-Out Bragg Peaks (SOBPs), the relative weights of each DDD were optimized simulating different beam penetrations and modulations. Absorbed doses were measured in water with an advanced Markus chamber. Neutron dose at the contralateral eye was also measured with passive bubble dosimeters. Results MC simulations and experimental results confirmed that maximizing the air gap between RS and collimator reduces the lateral dose penumbra of the collimated beam and increases the field transversal dose homogeneity. Therefore, RS and brass collimator were placed at about 98 cm (behind beam monitors) and 6 cm from the isocenter, respectively. The lateral penumbra ranged between 1.0 and 1.8 mm, in agreement with other proton therapy eye centers. Figure 1 shows examples of transversal dose profiles measured with EBT3 films at two different depths, compared against MC simulations. The distal fall-off of the DDDs ranged between 1.0 and 1.6 mm, comparable to the ones of most existing facilities. The measured SOBP doses were in very good agreement with MC simulations, as shown in Figure 2. The mean neutron dose at the contralateral eye was 68.8 ± 10.2 µSv/Gy. Beam delivery time, for 52 GyE prescribed dose in 4 fractions, was around 3 minutes per session.

Cause&Effect relationships were analyzed, identifying possible failure sources and potential alternatives for reducing the improbable occurrences. The lean instruments were considered for improving the process. The procedure was controlled over 6 months (Gen-Jun 2017). Results Applied shifts of 14948 consecutive fractions from 1353 patients were analyzed. No significant differences were observed over the 3 years (mean and SD were, respectively: 0.43±0.32cm, 0.43±0.32cm 0.45±0.33cm, showing an overall process stability. The major observed discrepancy was the monthly percentage of fractions with almost zero shifts (figure 1). Ishikawa fishbone method was adopted to recognize the artificial variability. Lack of confidence in applying shifts, reduced image quality, non-systematic use of automatic matching were considered as main con-causes of the discrepancy. Procedure harmonization to increase confidence in matching was implemented. In detail, (a) the DRRs parameters were personalized for the two projections (tangential and AP views); (b) the body and homo-lateral lung contours were exported with the DRR; (c) automatic pre-matching was performed, followed by manual fine tuning; (d) cases of corrections greater than 3mm were visualized aiming to help the RTT with the next shifts. In 2017, distribution symmetry improvement (Skewness moved from 1.4 to 1.1) and outlier reduction, verified by Kurtosis diminution, demonstrated a better “normali- zation” of the procedure after the LSSM (figure 2).

Figure 1 : Percentage of patients with shifts <0.1cm in function of month (before LSSM implementation).

Figure 2: Histogram of the shifts in function of year. Conclusion LSSM was successfully applied for the first time in a RT department, allowing the breast repositioning matching procedure to be redesigned. EP-2155 Dosimetry of the first synchrotron-based scanning proton beamline for the treatment of ocular tumors E. Mastella 1 , G. Magro 1 , A. Mirandola 1 , S. Molinelli 1 , S. Russo 1 , A. Vai 1 , D. Maestri 1 , A. Mairani 1,2 , M.R. Fiore 3 , C.