recognized experts, more than 50 abstract presentations, hundreds of scientific posters, receptions and networking opportunities among colleagues, exhibits of latest technologies from vendors around the world, and a site visit to the new Cincinnati Children's Hospital/University of Cincinnati Medical Center Proton Therapy Center. Target Audience * Physicians, Radiation Oncologists, Oncologists, Surgeons, Residents, and Fellows * Medical Physicists, Dosimetrists, and Radiation Therapists * Radiation Biologists, Accelerator Engineers, and Scientists * Registered Nurses and Clinical Researchers * Health Care Policy Makers, Insurance Executives, Industry Personnel, and Hospital Administrators Particle Therapy Cooperative Group (PTCOG) Steering Committee Jay Flanz, Chair Project and Technical Director at the Harvard Massachusetts General Hospital's (“MGH”) Francis H. Burr Proton Therapy Center Tadashi Kamada, Vice-Chair Director General of Clinical Research Cluster, National Institute of Radiological Sciences, QST Marco Durante, Vice-Chair Director, Trento Institute for Fundamental Physics and Applications (TIFPA), National Institute of Nuclear Physics (INFN) Martin Jermann, Secretary/Treasurer Consultant, Directorate Staff, Paul Scherrer Institut (PSI) Particle Therapy Cooperative Group 2018 (PTCOG 57) Conference Committee John Perentesis, MD, FAAP, Co-Chair Directo, Division of Oncology, Cincinnati Children's Hospital Medical Center Anthony Mascia, PhD, DABR, Co-Chair Adjunct Assistant Professor, UC Medical Center Proton Therapy Center, Physician, UC Health. John Breneman, MD, Co-Chair Professor of Radiation Oncology and Neurosurgery, UC College of Medicine, Physician, UC Health. Oral Presentations O 001: New Range Margin Recommendations Based on Dose Calculation Uncertainties J. Schuemann^1, H. Paganetti^1 ^1Massachusetts General Hospital & Harvard Medical School, Rad. Onc., Boston, USA The distal falloff of proton beams offers sharp dose contrasts to form highly conformal dose distributions. However, due to range uncertainties, margins are applied to ensure target coverage, resulting in unnecessarily irradiated tissue and avoidance of certain beam configurations. One of the main contributions determining range margins is uncertainty from analytical dose calculation algorithms (ADCs). Paganetti (2012) estimated that range margins (neglecting setup uncertainties) should be between 2.7% and 4.6% when using standard clinically used ADCs, depending on the heterogeneity of the patient geometry. Schuemann et al (2014) came to a similar conclusion, estimating that range margins could be reduced for some sites to 2.8% while they should be as high as 6.3% for others; however, that analysis did not focus on target coverage. We analyzed 1920 treatment fields to provide new recommendations for range margins. Six treatment sites (liver, prostate, breast, head-and-neck, lung and medullo) were sub-divided into 12 categories. For each category, the range margins were estimated for target coverage, assuming a perfect match of dose and range of delivered and ADC fields in water phantoms. We found that range margins can be reduced to 2.5% for prostate, pure liver and brain treatments. However, when not correcting for loss of dose due to small field sizes, scattering, etc., 3.5% should be used. Prostate fields could also use a constant range margin of 2.5 mm. For heterogeneous geometries, (see table and figure) generic range margins of 5.3% may be necessary. O 002: Shortening Delivery Times for PBS Proton Therapy by Reducing the Number of Proton Spots without Compromising Dosimetric Plan Quality M.F. Belosi^1, S. van de Water^2, F. Albertini^1, D.C. Weber^3, A.J. Lomax^4 ^1Paul Scherrer Institute, Centre for Proton Therapy, Villigen PSI, Switzerland, ^2Unaffiliated, Netherlands, ^3Paul Scherrer Institute- University Hospital of Bern and Zurich, Centre for Proton Therapy, Villigen PSI, Switzerland, ^4Paul Scherrer Institute- Swiss Institute of Technology ETH, Centre for Proton Therapy, Villigen PSI, Switzerland Purpose/objective: To investigate, through planning and experimental validation, the quality, deliverability, accuracy, robustness and delivery time reduction of spot-reduced PBS treatment planning. Material/methods: For a head-and-neck patient, conventional and ‘spot-reduced' SFUD plans were generated, with spot-reduced plans being calculated using the ‘pencil beam resampling' technique of Erasmus-iCycle (Erasmus MC Cancer Institute). This involves repeated inverse optimization while iteratively excluding low-weighted proton spots until the plan quality deteriorates. Beam setup was identical for both plans and the resulting dosimetric plan quality was comparable. Both plans were delivered on PBS Gantry 2 at PSI, measuring the delivery time per field and dose profiles in water, and subsequently recalculating dose distributions using the machine log-files. In addition, robustness analysis was performed to assess sensitivity to delivery inaccuracies and errors in patient setup and proton range. Results: The total number of spots for the plan could be reduced by 94% (26069 to 1540), resulting in an average delivery time reduction of 65% per field (Table 1). Measured dose profiles differed from the planned dose by 2.9%-4.3% for the spot-reduced plan and by <2% for the conventional plan. For both plans, the log-file dose reconstruction was within ±1% of the planned dose for all voxels (Figure 1). Spot-reduced plans were slightly more sensitive to delivery inaccuracies, requiring a spot position accuracy within ≤0.5mm, but were surprisingly less sensitive to setup and range errors. Conclusion: Delivery times per field could be reduced by 65% using spot reduction without substantially compromising plan quality, delivery accuracy or robustness. O 003: Limitations of Worst Case Robust Optimization: A Comparison with an All-scenario Approach J. Ma^1, H.S. Wan Chan Tseung^1, M. Herman^1, C. Beltran^1 ^1Mayo Clinic, Department of Radiation Oncology, Rochester, USA Purpose: Worst case optimization has been widely employed in intensity modulated proton therapy (IMPT) to achieve robust optimization. This work evaluates some of the limitations of worst case robust optimization in IMPT, and proposes an all-scenario robust optimization as an alternative to mitigate some of the limitations. Method: Worst case robust optimizations focus on a single scenario at each optimization iteration. The all-scenario approach strives to satisfy constraints under all uncertainty scenarios at each optimization iteration by including dose from all scenarios in the objective function rather than the worst case scenario only. Two different worst case robust optimizations were studied: composite worst case optimization and voxel-wise worst case optimization. The worst case approaches were compared with the all-scenario approach in two different clinical cases: a head and neck case and a brain case. The different optimization approaches were implemented with the same gradient based optimization engine. Plan quality and optimization convergence were compared between approaches. The optimization engine was GPU-accelerated. The dose calculation was based on an in-house GPU-accelerated Monte Carlo. Results: Compared with composite and voxel-wise worst case optimization, the all-scenario robust optimization converged faster, and arrived at solutions with tighter DVH robustness spread, better target coverage and lower OAR dose. Conclusion: The limitations of worst case robust optimization methods should be carefully evaluated for clinical use. The all-scenario robust optimization method has been shown to be a better choice. O 004: L1-Spot: Iteratively Constrained Spot Optimization with L1-Sparsity Regularization for Generating Efficient Proton PBS Plans Y. Lin^1, H. Gao^1, B. Clasie^2, F.F. Yin^1 ^1Duke University Medical Center, Radiation Oncology, Durham, USA, ^2Massachusetts General Hospital and Harvard Medical School, Radiation Oncology, Boston, USA The initial proton pencil beam scanning (PBS) treatment plans often consist of spots with weights that are too small to be deliverable and thus require post-processing to eliminate spots that fall below the deliverable threshold set by the treatment machine (indicated as Gmin). The purpose of this work is to develop an efficient spot optimization algorithm to generate PBS plans that eliminate low-weight spots while preserving the original intended dose distribution and also with potentially enhanced delivery efficiency by minimizing the number of spots to be delivered. The proposed method, namely L1-Spot, is a type of inverse planning algorithm that formulates the problem as a constrained optimization problem with L1 sparsity regularization. The minimum spot constraints are iteratively updated and enforced during the spot optimization, and the solution algorithm is based on alternating direction method of multipliers. The L1-Spot was compared against two state-of-art spot post-processing methods, round and redistribution (Figure1). The round method rounds spots between Gmin/2 and Gmin up to Gmin, and deletes spots below Gmin/2. The redistribution method interactively redistributes the lowest weight spot to its nearest neighbors until all spots comply with the machine limit (>Gmin). Its performance was evaluated using plans that use three different spot sizes (8mm, 4.6mm, and 2.3mm), see Table1. The evaluation criteria were the γ-index pass rate that compares the original and processed dose distributions. We developed a L1-Spot method to generate an efficient deliverable plan for PBS treatment and demonstrated that it was more effective than redistribution at large Gmin and/or small spot size. O 005: FROG: A New Calculation Engine for Physical and Biological Dose Investigations at CNAO K. Choi^1,2, S. Mein^3,4, B. Kopp^4,5, G. Magro^1, S. Molinelli^1, M. Ciocca^1, A. Mairani^1,6 ^1CNAO, Medical Physics, Pavia, Italy, ^2Pavia University, Physics, Pavia, Italy, ^3Heidelberg University, Physics, Heidelberg, Germany, ^4DKFZ, Imaging and Radiation Oncology, Heidelberg, Germany, ^5Heidelberg University Clinic, Radiation Oncology, Heidelberg, Germany, ^6HIT, Medical Physics, Heidelberg, Germany Fast and accurate dose calculation engine for hadrontherapy is important for both research and clinical application. FROG is a GPU-based forward calculation tool developed at CNAO (Centro Nazionale Adroterapia Oncologica) and HIT (Heidelberg Ion Beam Therapy Center) for fast and accurate calculation of both physical and biological dose for proton and carbon ion scanning beams. FROG calculation engine adopts a triple Gaussian parameterization for the description of the lateral dose distributions for a better agreement against FLUKA predictions. FROG provides dose/biological dose 2D maps, profiles and dose-volume-histograms. FROG-based RTDOSE DICOM files are created which can be read in by commercial treatment planning systems. For the benchmark of the FROG calculation engine, 6 Spread Out Bragg Peaks and few patients CNAO carbon treatment cases have been chosen and re-calculated with FROG. As a result, biological dose agreement against FLUKA is approximately within about 2%, while physical dose agreement varies with the beam energy from about 1% up to 3%. Both MKM (Microdosimetric Kinetic Model) and LEM (Local Effect Model) biological dose are implemented and tested in FROG to support the NIRS (National Institute of Radiological Sciences)-based to LEM-based biological dose conversion. FROG calculation engine is accurate and fast enough to re-calculate physical and biological doses applying various biological models. O 008: Uncertainties in Proton Stopping Power Ratio (SPR) from a Novel CT/MRI Tissue Classification Model A. Witztum^1, A. Sudhyadhom^1, T. Solberg^1 ^1University of California- San Francisco, Department of Radiation Oncology, San Francisco, USA Purpose: To analyze the uncertainty in SPR calculations based on errors in mean ionization potential (Im) and tissue classification in a four-component tissue classification (4CC) model using CT/MRI imaging. Methods: We proposed a 4CC system to classify molecules in the human body as proportions of water, fat, and protein (by MRI) and hydroxyapatite (HA) by CT so that: [22]graphic file with name i2331-5180-5-2-58-e01.jpg where 'w' is mass content percentage and 'I' is mean ionization potential for each type of molecule. Reference values were obtained for soft tissues (tissue) and cortical bone (bone) from ICRU report 44 and others in the literature. Errors in mass content percentages were normally distributed between 0-2% (water, HA), and 0-10% (fat). Errors in I were uniformly distributed between +/- 10 eV (water, fat, protein) and +/- 16 eV (HA). Proton (E = 250MeV) SPR was calculated using Bragg Additivity Rule (BAR) Im as the reference standard and compared to SPR using Im from: the 4CC system without errors SPR (4CC), and mean SPR from the distribution with errors. Results: Table 1 shows that for tissue and bone, the 4CC model SPR is within 0.2% of BAR. Adding the maximum systematic (0.2%) and random (2.4%) errors in quadrature yield a total maximum error of 2.4%. Conclusion: Current SPR calculation methods do not account for uncertainty in I. Even with a conservative error distribution the 4CC method performs well compared to current techniques. Content quantification by CT/MRI using the 4CC model has been demonstrated to be an accurate method for SPR determination. O 009: Comprehensive Log-file Based Monte Carlo Validations of a Fast, Analytical Dose Calculation for Pencil Beam Scanned Proton Therapy C. Winterhalter^1, Y. Tian^1, G. Meier^1, A. Bolsi1, M. Dieterle^1, J. Hrbacek^1, D. Oxley^1, D.C. Weber^1,^2,^3, S. Safai^1, A. Lomax^1 ^1Paul Scherrer Institute, Center for Proton Therapy, Villigen PSI, Switzerland, ^2University Hospital of Zurich, Radiation Oncology Department, Zurich, Switzerland, ^3University Hospital of Bern, Radiation Oncology Department, Bern, Switzerland Purpose: To validate a fast ray casting dose calculation (RCDC) for pencil beam scanned (PBS) proton therapy using log-file based Monte Carlo (LF-MC) calculations. Materials: PBS treatment planning at PSI uses a fast (0.5-5s/field using standard computer hardware) RCDC (Schaffner et al 1999). To validate the accuracy of this, clinical treatment plans have been reconstructed using log-file based Monte Carlo (TOPAS 3.0.p1) simulations. Results: MC parameters have been defined such that beam widths in air and depth doses in water match measurements to within 0.1mm. For all simulations, proton numbers per pencil-beam are directly derived from Faraday cup based monitor unit calibration measurements. As such, MC calculated energy specific absolute doses in water were found to agree to within -2%+/-0.5% to measurements for a range of energies (figure 1). When applied to clinical skull base and lung cases, and after the LF-MC distribution was normalized to the mean PTV dose of the RCDC (LF-MC normalization factors), 99%/92% of dose voxels with dose >10% of the prescription dose agreed to within +/-5% between the LF-MC and RCDC distributions respectively (figure 2). Correcting for the -2% absolute dose offset, the LF-MC normalization factors agreed with those from field specific verifications to between 0.3-1.9% per field for the skull base and 1.8-3.0% for the lung case. Conclusion: The fast RCDC algorithm provides excellent relative agreement to LF-MC reconstructed dose, even when including delivery uncertainties derived from log-files. Absolute dose prediction of the MC is good and matches well with patient specific absolute dose measurements. O 010: The Impact of LET/RBE Modelling and Robust Analysis on Skull-Base and Paediatric Craniopharyngioma Proton Plans, Benchmarked against VMAT A. Gutierrez^1, V. Rompokos^2, K. Li1, C. Gillies^2, D. D'Souza^2, F. Solda^3, N. Fersht^3, C. Yen Ching^3, G. Royle^1, R. Amos^1, T. Underwood^1 ^1University College London, Department of Medical Physics and Biomedical Engineering, London, United Kingdom, ^2University College London Hospitals NHS Foundation Trust, Department of Radiotherapy Physics, London, United Kingdom, ^3University College London Hospitals NHS Foundation Trust, Department of Clinical Oncology, London, United Kingdom Purpose: To investigate how physical and biological uncertainties shift the balance between VMAT and IMPT, considering whether IMPT plans remain dosimetrically superior when such uncertainties are considered. Methodology: We retrospectively studied ten cases: four chondrosarcoma, two chordoma and four paediatric craniopharyngioma. VMAT and IMPT plans were created according to modality-specific protocols using a PTV for VMAT and robust optimisation for IMPT. For IMPT we considered (i) variable RBE modelling using the McNamara model for different values of α/β, and (ii) robust analysis with +/-3mm set-up and 3.5% range uncertainties. Results: Comparing the VMAT and nominal IMPT plans, the dosimetric advantages of IMPT were clear: IMPT led to reduced integral dose (especially to the normal brain) and typically, improved CTV coverage given our OAR constraints. When robustness analysis was performed, IMPT doses exceeding the constraints were predicted for small volumes of OARs such as the brainstem and chiasm. However, variable RBE-weighted dose analyses predicted even more substantial hotspots. Within the nominal VMAT and IMPT plans, generally less than 0.1cm3 of the brainstem received the dose constraint, as shown in Figure 1. However, when the IMPT robustness analyses (worst case scenarios assuming RBE=1.1) and variable RBE modelling were performed, substantial increases to the volume receiving the constraint were observed. Conclusion: Both physical and biological robustness analyses should be considered for IMPT: these analyses can substantially affect the sparing of OARs and shift the dosimetric balance relative to VMAT. O 011: Impact of Radiotherapy Modality on Neuropsychological Outcomes of Pediatric Brain Tumor Patients J. Gross^1, S. Powell^2, F. Zelko^2, W. Hartsell^3, S. Goldman^4, J. Fangusaro^4, R. Lulla^4, N. Pillay Smiley^4, J. Chang^3, V. Gondi^3 ^1Northwestern University, Radiation Oncology, Chicago, USA, ^2Ann and Robert H. Lurie Children's Hospital of Chicago, Psychology, Chicago, USA, ^3Northwestern Medicine Chicago Proton Center, Radiation Oncology, Warrenville, USA, ^4Ann and Robert H. Lurie Children's Hospital of Chicago, Pediatric Oncology, Chicago, USA Purpose: To evaluate predictors of neuropsychological outcomes for pediatric brain tumor patients undergoing X-ray radiotherapy (XRT) or proton radiotherapy (PRT). Methods: 125 patients received treatment for brain tumors. All received age-appropriate neuropsychological assessment of intelligence quotient (IQ), processing speed (PS), visual motor integration (VMI), executive function, memory and parent-reported function at our institution. Results: Median age at diagnosis was 7.0 years; median time from treatment to last assessment was 4.0 years. Patients receiving PRT had higher socioeconomic status (SES), differing distributions of race and tumor locations (Table 1), as well shorter median follow-up time compared to XRT (9.5 vs. 5.0 years, p<0.001). Univariate and multivariate analyses with tests for interaction of treatment group and follow-up identified higher verbal IQ (β=0.84 points/year, p=0.02) and full-scale IQ (β=1.03, p=0.01) for older patients; lower PS (β=-15.9 points, p=0.04) and VMI (β=-14.0, p=0.006) following CSI; and higher full-scale IQ (β=10.6 points, p=0.048), PS (β=12.6, p=0.02), and parent-reported practical function (β=13.8, p=0.049) following PRT relative to XRT. VMI was higher in those with higher SES (β=1.2 points/$10,000 household income, p=0.04), but lower following receipt of vincristine chemotherapy (β=-16.6, p=0.01). Parent-reported practical function was lower in those with posterior fossa tumors (β=-10.8, p=0.048). Conclusions: Neuropsychological outcomes of pediatric brain tumor patients are impacted by age, SES, receipt of CSI or vincristine chemotherapy, and tumor location. Relative to XRT, PRT is associated with favorable neuropsychological outcomes in terms of full-scale IQ, PS, and parent-reported practical function. O 012: Comparison of Symptomatic and Asymptomatic Imaging Changes following Pencil-Beam Scanning and Uniform Scanning Proton Therapy for Posterior Fossa Tumors J. Gross^1, B. Kreydick^2, E. Chang^2, M. Pankuch^2, C. Bregman^3, J. Kalapurakal^1, J. Chang^2, W. Hartsell^2, V. Gondi^2 ^1Northwestern University, Radiation Oncology, Chicago, USA, ^2Northwestern Medicine Chicago Proton Center, Radiation Oncology, Warrenville, USA, ^3Ann and Robert H. Lurie Children's Hospital of Chicago, Radiology, Chicago, USA Purpose: Strategies to mitigate the RBE effects at the Bragg peak end-of-range have unknown impact on post-treatment imaging changes. We compared treatment-related imaging changes following proton therapy (PT) delivered with uniform scanning (US) without mitigation, US with mitigation (blocking, range feathering, range pushing), and pencil beam scanning (PBS). Methods: MRI and follow-up for all patients with medulloblastoma and infratentorial ependymoma receiving PT between 2011 and 2016 at our institution were reviewed with a Neuroradiologist. New lesions on contrast-enhanced T1- or T2-weighted MRI following PT were recorded and graded as per the Common Terminology Criteria for Adverse Events (CTCAE) v4.0. Comparisons were made with the Kaplan-Meier method and multivariable Cox Proportional Hazards models. Results: 127 patients met inclusion criteria (Table 1). Crude rates of asymptomatic imaging changes were 38.7% for US without mitigation, 48.2% for US with mitigation, and 51.3% for PBS. On time-to-event analysis, PBS was associated with higher rate of asymptomatic imaging changes in the posterior fossa (p<0.001, Figure 1). On multivariable analysis accounting for age, total dose and receipt of chemotherapy, PBS retained significant association with asymptomatic imaging changes in the posterior fossa (HR 10.1, 95% CI 2.7 - 38.4, p<0.001). Symptomatic changes associated with CTCAE grade 2+ events occurred in 5 patients (crude rate 3.9%). No difference in symptomatic imaging changes was observed between cohorts. Conclusions: Following PT for medulloblastoma or infratentorial ependymoma, symptomatic imaging changes are rare and not impacted by PT modality, but PBS is associated with increased asymptomatic imaging changes in the posterior fossa. O 013: Factors Associated with Insurance-Related Referral Delays in Children Irradiated with Proton Therapy in the Pediatric Proton Consortium Registry C. Hess^1, D. Indelicato^2, A. Paulino^3, W. Hartsell^4, C. Hill-Kayser^5, S. Perkins^6, A. Mahajan^7, N. Laack^7, R. Ermoian^8, T. Yock^1 ^1Massachusetts General Hospital, Radiation Oncology, Boston, USA, ^2University of Florida, Radiation Oncology, Jacksonville, USA, ^3MD Anderson, Radiation Oncology, Houston, USA, ^4Northwestern Medicine, Chicago Proton Center, Chicago, USA, ^5University of Pennsylvania, Radiation Oncology, Philadelphia, USA, ^6Washington University, Radiation Oncology, St. Louis, USA, ^7Mayo Clinic, Radiation Oncology, Rochester, USA, ^8University of Washington, Radiation Oncology, Seattle, USA Background/Objectives: We report factors associated with insurance-associated treatment delays in children irradiated with proton therapy in the Pediatric Proton Consortium Registry. Design/Methods: A multi-institutional registry of childhood cancer patients undergoing proton radiation therapy was opened to enrollment in summer 2012. Data was frozen for analysis on October 28, 2017. Results: Of 1,578 children enrolled, 1,304 (83%) had available data regarding insurance-associated delays. A total of 83 of these children (6%) were delayed by insurance: 51/1029 (5%) for private/foreign insurance compared to 32/275 (12%) among Medicaid patients (p<0.001). On average, 28.3% of children with rare tumors were delayed compared to 8.8% with non-rare tumors (p<0.001). Among non-rare tumors, non-CNS tumors were more likely to be delayed than CNS (p=0.001). Neither age nor gender predicted delay. Black, Hispanic, and Asian children made up 6%, 8%, and 5% of the total cohort, 13%, 21%, and 2% of the Medicaid population, and 12%, 16%, and 1% of children with delay, respectively. “Black or Hispanic” children (8%) had higher rates of insurance-related delays compared to “white and non-Hispanic” children (5%) (p=0.001), yet Medicaid insurance status increased delays to 12% and 14% in both groups, respectively. Delays were rare among Asian children (2%). Conclusion: Insurance-related RT delay for proton therapy referral is more common in Black or Hispanic children than white but Medicaid coverage doubles delays across racial/ethnic groups. Rare and non-CNS tumor types are risk factors for delay. Awareness of risk factors for referral delay can guide efforts to address disparities in access to advanced treatments. O 014: Craniospinal Irradiation: Evaluation of Proton Range Deviation Using Daily CBCT W. Yao^1, T. Merchant^1 ^1St. Jude Children's Research Hospital, Radiation Oncology, Memphis, USA Purpose: Proton range deviation on the surface of the clinical target volume (CTV) provides important information about actual delivered dose. Cone-beam computed tomography (CBCT) accurately reflects patient anatomy relative to the treatment delivery system. CBCT may be used to calculate range deviation to set the appropriate set-up margin. This work focused on range deviation in the clinical setting of craniospinal irradiation (CSI). Method: 396 CBCT data sets from 10 pediatric patients treated with intensity-modulated proton CSI were used to investigate range deviation. The water equivalent thickness along the CTV surface was calculated for each beam using the first day CBCT which was then used as the reference for the remaining treatments. Additional pCTs performed in 2 patients for boost planning were used to check the consistency of the calculated range deviations in CBCT and pCT. Results: The mean range deviation (mean ± RMS) for the 10 study patients was 2.7 ± 1.5 mm in the cranium, 2.3 ± 1.3 mm in the cervical, 1.5 ± 0.7 mm in the upper (thoracic) spine and 1.9 ± 0.9 mm in the lower (lumbar) spine. In the pCTs the mean range deviation for the cranium was 2.4 ± 0.7 mm that underwent consistency evaluation. Conclusion: Range deviation calculated using daily CBCT can be reliably used to monitor range difference in patients undergoing daily CSI. We estimate that an inter-fraction margin of 5 mm for the cranio-cervical, and 3 mm for the upper and lower spine volumes would be sufficient in the absence of personalized set-up margins. O 015: Practice Trends for Vertebral Body Coverage of Pediatric Patients Undergoing Proton Craniospinal Irradiation S. Medek^1, B. De^2, J. Breneman^1,3, L. Pater^1,3, N. Laack^4, A. Mahajan^4, S. Wolden^2, R. Vatner^1,3 ^1University of Cincinnati Medical Center, Radiation Oncology, Cincinnati, USA, ^2Memorial Sloan Kettering Cancer Center, Radiation Oncology, New York, USA, ^3Cincinnati Children's Hospital Medical Center, Radiation Oncology, Cincinnati, USA, ^4Mayo Clinic, Radiation Oncology, Rochester, USA Introduction: Craniospinal irradiation (CSI) is an important therapeutic component for pediatric brain tumors with a propensity for dissemination. While the thecal sac is the spinal clinical target volume (CTV), many physicians treat the entire vertebral body (VB) in growing children due to risk of lordosis from partial VB treatment. Currently there is no standard of care for determining VB coverage in proton based CSI, leading to variable practice patterns. Materials/Methods: Pediatric radiation oncologists were identified from membership in the PT-COG pediatric subcommittee or affiliation with U.S. proton centers. Potential participants were contacted by email with link to an IRB approved, anonymized web-based survey distributed on June 21st 2017 with follow-up on October 10th 2017. The survey utilized skip logic and included up to 11 questions regarding practice patterns. Results: Thirty-three radiation oncologists representing all regions of the U.S. responded. Five were excluded for lack of recent pediatric proton CSI experience. Of 28 responses, 23 physicians sometimes cover, five always cover and zero never cover the entire VB. Factors influencing VB coverage included patient age (n=17), bone age (n=12), growth curves (n=7) and parental height (n=6). Most physicians modify anterior CTV margins to protect ventral structures. Conclusion: Vertebral body coverage varies amongst radiation oncologists treating pediatric patients with proton CSI, with multiple factors influencing physician technique. These data suggest the need for a standardized approach to VB coverage in pediatric proton based CSI. O 016: Proton versus Photon Radiation Therapy for Pediatric Head and Neck Rhabdomyosarcoma: Disease Control, Overall Survival, and Toxicity D. Casey^1, L. Wexler^2, S. Wolden^1 ^1Memorial Sloan Kettering Cancer Center, Department of Radiation Oncology, New York, USA, ^2Memorial Sloan Kettering Cancer Center, Department of Pediatrics, New York, USA Introduction: Proton therapy (PT) provides promise in sparing late toxicity without compromising clinical outcomes for children with head and neck rhabdomyosarcoma (HNRMS). We compared outcomes after photon versus proton irradiation. Methods: Sixty-six HNRMS pediatric patients (age <25 years) were treated with definitive chemoradiation at Memorial Sloan Kettering. Fifty-one patients were treated with intensity modulated radiation therapy (IMRT) and 15 patients with PT. Locoregional control (LRC), event-free survival (EFS), overall survival (OS), and acute toxicity were compared. Results: Median follow-up was 9.8 versus 1.6 years in the IMRT and PT cohorts, respectively. The cohorts were similar with respect to age, gender, race, tumor site, size, group, and stage. Patients treated with IMRT were more likely to be PAX/FOXO1 fusion-positive (p=0.06) and node-positive (p=0.03). Five-year LRC was 91% in the IMRT cohort versus 85% in the PT cohort (p=0.28). All local failures were in-field; there were no marginal failures. Five-year EFS (71% vs 74%, p=0.97) and OS (75% vs 71%, p=0.54) were similar for both cohorts. There were no grade 4 acute toxicities. Grade 3 mucositis rates were slightly higher in the IMRT cohort (20% versus 13%, p=0.58), while grade 3 dermatitis rates were higher in the PT cohort (33% versus 2%, p<0.0001). Conclusions: IMRT and PT appear to result in similar LRC and survival for pediatric HNRMS. Longer follow-up is needed in the proton cohort prior to fully establishing their equivalence. Given the encouraging short-term outcomes and the potential late morbidity benefits, we recommend PT for HNRMS whenever feasible. O 017: Proton Beam Therapy for orbital Rhabdomyosarcomas at West German Proton Therapy Center Essen (WPE) D. Geismar ^1,2,3, T. Steinmeier^1,3, S. Nagaraja^1,2,3, S. Peters^1,2,3, S. Plaude^1,3, S. Tippelt^4, B. Timmermann^1,2,3 ^1West German Proton Therapy Center Essen WPE- Hufelandstraße 55- 45147 Essen- Germany, University Hospital Essen, Essen, Germany, ^2Clinic for Particle Therapy- University Hospital Essen- Hufelandstraße 55- 45147 Essen- Germany, University Hospital Essen, Essen, Germany, ^3West German Cancer Center WTZ- Hufelandstraße 55- 45147 Essen- Germany, University Hospital Essen, Essen, Germany, ^4University Hospital of Essen- Pediatrics III- Pediatric Hematology and Oncology- Essen- Germany, University Hospital Essen, Essen, Germany Purpose: Proton beam therapy (PT) is of increasing interest especially in tumors in close proximity of critical structures or in particular sensitive tissues like orbital rhabdomyosarcomas. Initial results of PT at WPE are presented. Methods: Between April 2014 and September 2017, 16 patients (8 male, 8 female, median age 7.4 y (3.1-16.8 y)) with orbital rhabdomyosarcomas were treated at WPE and were prospectively enrolled in the in-house registry KiProReg. Histology types were embryonal rhabdomyosarcoma (87.5%) or alveolar rhabdomyosarcoma (12.5%). 37.5% had a parameningeal involvement. All patients were treated with chemotherapy (CTx) before radiotherapy and in 81.2% concomitant CTx was applied. The median PT dose was 50.4 Gy (41.4-54 Gy) applied in mean 28 fractions (23–30) by using uniform scanning (68.8%), pencil beam scanning (12.5%) or both techniques (18.7%), respectively. Results: The median follow-up after last fraction is 1 year (0.0-2.4 y). 11 patients (68.8%) showed disease control. Local recurrence occurred in 4 patients and 1 patient developed a metastatic disease after treatment. No patients died so far. PT was well-tolerated. New high-grade (CTCAE ≥°3) acute toxicities occurred only in the field of hematological toxicity (n=5). Long-term data for 12 months after PT show a low number of new high-grade (CTCAE °3) toxicities regarding hematological toxicities. No new grade 4 or grade 5 effects occurred. Conclusion: Current data support safety, tolerance and effectivity of PT in orbital rhabdomyosarcomas. However, long-term follow-up data is still needed to assess long-term outcomes. O 018: Carbon Ion Radiotherapy for Inoperable Pediatric Osteosarcoma O. Mohamad^1, R. Imai^2, T. Kamada^2, Y. Nitta^2, N. Araki^3 ^1National Institute of Radiological Sciences Hospital of Charged Particles and University of Texas Southwestern Medical Center, Department of Radiation Oncology, Chiba, Japan, ^2National Institute of Radiological Sciences, Hospital of Charged Particles, Chiba, Japan, ^3Ashiya Municipal Hospital, Department of Surgery, Ashiya City- Hyogo, Japan Background: Osteosarcoma, the most common primary bone malignancy in children, has poor outcomes with conventional treatments especially if inoperable or in truncal sites. Methods: We retrospectively reviewed the records of pediatric and adolescent patients who received carbon ion radiotherapy (CIRT) for inoperable osteosarcoma between 1996 and 2014. Results: The cohort consisted of 26 patients aged 11-20 years (median 16) with inoperable high-grade osteosarcoma of the trunk (24 pelvic, 1 mediastinal and 1 paravertebral) without any other lesion at initial examination. There were 22 primary, 1 locally recurrent and 3 metastatic cases. All patients received prior chemotherapy and only 4 patients received prior surgery. Median CIRT dose was 70.4 Gy RBE (relative biological effectiveness) delivered in 16 fractions. Median follow-up was 32.7 months. Overall survival was 50.0% and 41.7% at 3 and 5 years, respectively. Ten patients survived for more than 5 years (range 5-20.7 years). Local control was 69.9% and 62.9% at 3 and 5 years, respectively. Fourteen cases developed distant metastasis and, thus, progression-free survival was 34.6% at 3 and 5 years (Figure 1). Only largest tumor diameter correlated with 5-year overall survival and local control. All patients tolerated the treatments well and there were 4 grade 3-4 CIRT-related late toxicities and 1 case of fracture in an irradiated bone. No treatment-related mortalities were observed. All patients (except 1) were able to ambulate after CIRT. Conclusions: CIRT was safe and efficacious in the treatment of inoperable osteosarcoma with improved local control, overall survival and adverse events compared to conventional treatments. O 019: Gamma Electron Vertex Imaging (GEVI) for Proton Beam Range Measurement: Experimental Results S.H. Kim^1, J.H. Park^1, Y. Ku^1, H.R. Lee^2, C.H. Kim^1, S.K. Cho^3 ^1Hanyang University, Nuclear Engineering, Seoul, Korea Republic of, ^2Korea Atomic Energy Research Institute, Neutron Utilization Technology Division, Daejeon, Korea Republic of, ^3Samsung Medical Center, Radiation oncology, Seoul, Korea Republic of A prototype gamma electron vertex imaging (GEVI) system was developed to monitor the proton beam range in the patient by measuring prompt gammas. In the GEVI system, the prompt gammas from the patient are converted to electrons in the electron converter, and the electrons are then measured by the subsequent detectors (i.e., two hodoscopes and a calorimeter). Measuring the trajectories and the energies of the electrons, the vertices of the prompt gammas can be located, thereby producing a 2D prompt gamma image. In the present study, prompt gamma distributions were obtained using the prototype GEVI system for the therapeutic proton beams at Samsung Medical Center (SMC) in Korea. The pencil beams, from 90 MeV to 180 MeV in 15 MeV steps, were delivered to a high-density polyethylene phantom. The GEVI system was fixed at a location for all proton energies considered in the study, and 6.24×109 protons were delivered for each measurement case. Figure 1 shows the 2D prompt gamma images measured using the GEVI system. For all of the proton beams, the positions of the maximum counts on the 2D images were found to be directly in front of the beam range. Using the falloff positions of the GEVI images, a linear correlation between the falloff positions and beam ranges was derived with the coefficient of determination (r2) of 0.999. With this correlation, we believe that the proton beam ranges can be measured accurately, i.e., within ∼3 mm of error, for a wide range of proton beam energies. O 020 - How Risky Is 4D Planning? An FMEA-Based Human Error Comparison of 3D/4D PT QA Planning Risks Performed at CPT/PSI F. Emert^1, D. Pandya^2, J. Hrbacek^1, P. Morach^1, J. Rottmann^1, S. Zepter^1, L. Podofillini^2, A. Lomax^1, D.C. Weber^1, V.N. Dang^2 ^1Paul Scherrer Institute, Center for Proton Therapy, PSI-Villigen, Switzerland, ^2Paul Scherrer Institute, Risk and Human Reliability- LEA, Villigen PSI, Switzerland Purpose: Identification and quantification of human error risks providing failure minimization is of critical importance for patient safety in radiotherapy. This especially applies to new, complex workflows such as 4D-treatment planning. Therefore, a risk analysis comparing traditional 3D- with modern 4D-workflow implementation in proton therapy at Paul-Scherrer-Institute was performed. Overall goal is to compare failure risks of important tasks in a 4D-planning workflow with their counterparts in the 3D-planning process. Methods: 5 tasks (3 specific for each workflow type, 2 common to both; 4D-example, Table1), their failure modes and causes (FC) were identified based on a Human Reliability Analysis (HRA) method developed at PSI and screening analysis by 5 experts who are familiar with 3D/4D-planning techniques. The risk estimation was based on the Failure Mode and Effect Analysis (FMEA) approach outlined in AAPM-TG100 report. Results: Risk Priority Number (RPN) analyses of the tasks revealed a significantly higher risk of 4D- compared to 3D-planning processes (Fig1). RPN ratios [RPN(4D)/RPN(3D)] for both common tasks ranged from 1.2-1.9, for 3 related but workflow specific tasks between 1.3-8.8. The top FCs that received high risks were “less experienced with process” and “high workload” (Table1). High occurrence and low detectability were related to high RPNs. Conclusions: Due to complicated 4D-planning workflow and its relatively newness, significantly higher risks are observed in 4D tasks than their 3D counterparts. By studying complete failure scenarios using the in-house developed HRA method, we will investigate tasks with highest risks and minimize failure rates by optimized distribution of QM/QA resources. O 021: Proton Range Verification through Acoustic Measurements W. Nie^1, K. Jones^2, S. Petro^3, A. Kassaee^3, C. Sehgal^4, S. Avery^3 ^1University of Pennsylvania, Department of Radiation Oncology- West Pavilion 2nd Floor. Room CN-495A, Philadelphia, USA, ^2Rush University, Radiation Oncology, Chicago- IL, USA, ^3University of Pennsylvania, Radiation Oncology, Philadelphia, USA, ^4University of Pennsylvania, Radiology, Philadelphia, USA Proton radiation therapy requires a simple and accurate method to measure the proton beam Bragg peak (BP) depth for monitoring patient treatment. Protoacoustics - measurement of the pressure waves emitted by thermal expansion resulting from proton dose deposition - may be used to obtain the position of the BP in a phantom by measuring the time-of-flight (TOF) of the maximum amplitude pressure wave. We performed acoustic measurements and developed four different methods for analyzing the protoacoustic signals produced in homogeneous phantoms made from metallic and plastic materials. Analysis shows that among these four methods, the best exhibits minimal error (0.2cm). After comparing multiple materials and shapes, we conclude that the rectangular high density polyethylene phantom is most appropriate for proton range verification due to its durability and convenience. To characterize the ideal detection device, the protoacoustic signals generated in the polyethylene phantom were measured with three different detectors: hydrophone, accelerometer, and laser vibrometer. The accelerometer and vibrometer demonstrate similar sensitivity as the hydrophone: the accelerometer has light weight and small size, therefore it is easy to be attached on the patient exterior; the vibrometer has advantages in multi-spot scanning remotely and simultaneously. Psuedospectral wave-equation simulations (k-Wave MATLAB toolbox) were performed using phantom CT images to study the effect of CT number error on protoacoustic range verification. To reduce the uncertainty in TOF measurements and CT error, we can apply a triangulation algorithm. Simulation study shows we can reduce the uncertainty due to the uncertainty to less than 0.5mm. O 023: An Activity to Improve Quality through Dose Inter-comparison by End-To-End Testing in Carbon-Ion Radiation Therapy Centers in Japan H. Mizuno^1, A. Fukumura^2, N. Kanematsu^2, S. Yonai^2, K. Yusa^3, T. Yanou^4, M. Suga^5, M. Mizota^6, S. Minohara^7, T. Kanai^3 ^1National institute of Radiological Sciences- QST, Dept. of Radiation Measurement and Dose Assessment, Chiba, Japan, ^2National institute of Radiological Sciences- QST, Clinical Research Cluster, Chiba, Japan, ^3Gunma University, Heavy Ion Medical Center, Maebashi, Japan, ^4Hyogo Ion Beam Medical Center, Department of radiation technology, Tatsuno, Japan, ^5Hyogo Ion Beam Medical Center, Department of radiation phyiscs, Tatsuno, Japan, ^6SAGA HIMAT Foundation, Ion Beam Therapy Center, Tosu, Japan, ^7Kanagawa Cancer Center, Department of radiation therapy, Yokohama, Japan For multicenter clinical studies of carbon-ion radiation therapy (C-ion RT), dose uniformity among centers is crucial. In the clinical studies of The Japan Carbon-ion Radiation Oncology Study group (J-CROS), dosimetric inter-comparison for all centers participating in clinical studies was performed using the end-to-end test method. All five centers that had performed C-ion RT in Japan in 2017 participated. A dedicated water phantom was produced for the test. CT imaging and treatment plan was performed using the same method as the patient, the physical dose of the isocenter was measured by the ionization chamber dosimeter, and compared with the calculated dose from the treatment planning system. As a result, the difference between the measured and calculated doses in all five facilities were within ± 3%, tolerance level. Uniformity of dose among institutions is guaranteed. In addition, in response to the inter-comparison dosimetry, some centers have begun actions to improve the dosimetry precision. For instance, thermometers and barometers used for absolute dosimetry in one center that had not been calibrated as those mounted in the irradiation system were finally calibrated. In another center, an old type of ionization chamber used as the reference dosimeter was updated to a new type. The team of J-CROS medical physicists continues with this kind of QA activity among C-ion RT centers. O 024: Differences in Human Endothelial Cell Activation after Irradiation with Photons vs. Protons in the Proximal and Distal Spread-Out Bragg Peak R. Vatner^1, H. Shen^2, J. Goines^2, D. Ionascu^3, E. Wolf^3, W. Kassing^3, M. Lamba^3, E. Janssen^4, A. Mascia^3, E. Boscolo^5 ^1University of Cincinnati / Cincinnati Children's Hospital Medical Center, Radiation Oncology, Cincinnati, USA, ^2Cincinnati Children's Hospital Medical Center, Cancer and Blood Diseases Institute, Cincinnati, USA, ^3University of Cincinnati, Radiation Oncology, Cincinnati, USA, ^4Cincinnati Children's Hospital Medical Center, Immunobiology, Cincinnati, USA, ^5Cincinnati Children's Hospital Medical Center, Experimental Hematology and Cancer Biology, Cincinnati, USA Purpose/Objectives: Radiation necrosis is an infrequent but morbid late effect of radiotherapy for brain tumors, with anecdotes of a propensity at the distal end of the spread-out Bragg peak (SOBP). The mechanism is unknown, but a prevailing hypothesis is that radiation necrosis results from inflammation and microvascular dysfunction due to endothelial cell injury. We test the hypothesis that proton irradiation at the distal SOBP induces more endothelial activation than irradiation with photons or proximal SOBP protons. Materials/Methods: Human umbilical vein endothelial cells (HUVECs) were treated to 0Gy, 2Gy, or 8Gy with 0.662MeV photons, or 70-240MeV protons in the proximal or distal 2cm of 10cm SOBP. Cells were incubated at 37C for 24h and inflammatory cytokine and chemokine expression was tested using the Human Angiogenesis Array C1000 (RayBiotech). At 48h flow cytometry was performed using antibodies for MHC Class I, ICAM-1 and HCAM. Results: Proximal and distal SOBP proton treated cells had diminished viability compared with photon treated cells. There was a dose dependent increase in expression of MHC-I and CD54/ICAM-1, and decreased CD44/HCAM after proton vs. photon treatment. Differences in inflammatory cytokines and chemokines were also dose dependent with a trend toward greater effects in proton vs. photon treated groups, and no significant difference between cells treated with proximal vs. distal SOBP. Conclusion: Proton and photon radiation activate a dose dependent inflammatory profile in HUVECs. Proton radiation induces a stronger inflammatory response compared with photons. No significant differences were detected in effects between proximal and distal SOBP. O 025: Drosophila Melanogaster may Be a Useful Model Organism to Study the Biological Effects of Proton Radiation K. Nakajima^1,2, H. Iwata^1,2, M. Naito^3, S. Hirai^3, K. Kume^4, G. TianXiang^4, C. Omachi^5, J.E. Mizoe^1,6, H. Ogino^1, Y. Shibamoto^2 ^1Nagoya Proton Therapy Center- Nagoya City West Medical Center, Department of Radiation Oncology, Nagoya, Japan, ^2Nagoya City University Graduate School of Medical Sciences, Department of Radiology, Nagoya, Japan, ^3Aichi Medical University, Department of Anatomy, Nagakute, Japan, ^4Nagoya City University, Department of Neuropharmacology, Nagoya, Japan, ^5Nagoya Proton Therapy Center, Department of Proton Therapy Physics, Nagoya, Japan, ^6Osaka Heavy Ion Therapy Center, Department of Radiation Oncology, Osaka, Japan Background: Biological effects of proton beams have been well investigated in vitro. In contrast, however, their effects in in vivo systems are not sufficiently understood. Drosophila melanogaster is an organism well-suited for genetic analysis and commonly used in medical research. This is a study to demonstrate the usefulness of the fly as a model organism to investigate the biological effects of proton beams. Materials/Methods: Third-instar larvae of wild-type Drosophila melanogaster (Oregon-R) were irradiated with 6-MV X-rays or passive proton beams at the center of the spread-out Bragg peak. They were exposed to single doses of 0-50 Gy and observed for up to 48 hours after eclosion. RT-qPCR for mRNA was performed using 40-Gy irradiated samples. Results: The surviving fraction within 48 hours was: control (0 Gy), 0.87; 10 Gy, 0.75/0.92 (photon/proton); 20 Gy, 0.81/0.73; 30 Gy, 0.73/0.37; 40 Gy: 0.52/0.30; and 50 Gy, 0.12/0.05. Two-way ANOVA revealed significant differences between the radiation types (f1) and doses (f2); p-values were 0.023 for f1, <0.001 for f2, and 0.015 for f1f2. In the analysis of mRNA expression of 26 genes, remarkable differences were observed between photons and protons in genes associated with DNA repair, such as Ku70, Ku80 and Rad51. Conclusions: To date, this is the first study using Drosophila to investigate the biological effects of proton beams. We demonstrated the differences between photons and protons in both phenotypes and genotypes in the in vivo system. This experimental model may be useful in elucidating biological aspects of proton beams. O 026: Radiobiological Impact of Mixed Field from Target Fragments in Proton Treatment Plans V.E. Bellinzona^1, E. Scifoni^1, M. Krämer^2, F. Tommasino^1,3, G. Petringa^4, P. Cirrone^4, F. Romano^4,5, L. Grzanka^6, T. Friedrich^2, M. Durante^1 ^1INFN - Istituto Nazionale di Fisica Nucleare, TIFPA-Trento Institute for Fundamental Physics and Applications, Trento, Italy, ^2GSI Helmholtzzentrum für Schwerionenforschung GmbH, Biophysics, Darmstadt, Germany, ^3Università degli Studi di Trento, physics, Trento, Italy, ^4INFN - Istituto Nazionale di Fisica Nucleare, LNS - Laboratori Nazionali del Sud, Catania, Italy, ^5National Physical Laboratory, CMES - Medical Radiation Science, Teddington, United Kingdom, ^6Institute of Nuclear Physics Polish Academy of Sciences, Physics and Astronomy, Cracow, Poland Among the various radiobiological effects of proton beams, the impact of target fragments on Relative Biological Effectiveness (RBE) is an open issue and TPS would possibly benefit from an implementation accounting for this effect. Recently this topic has become timelier since new experiments like FOOT (FragmentatiOn Of Target) promise to get relevant physics data with unprecedented details. In this scenario, we ask: Is this correction actually needed in clinical practice? Within the MoVe¬IT project, we aim to answer, analyzing he overall mixed field arising in the irradiated medium and the RBE composed by all the produced secondary fragments compared to the pure primaries effect. By using Monte Carlo methods (Geant4), we have produced the spectra of those fragments for different initial proton energies in water to implement them in TRiP98 and to evaluate their impact with a combination of LEM IV model and mixed field algorithm, in order to account the biological dose contribution of each single fragment and reconstruct the overall RBE. First steps of the analysis of the spectra indicate a sensible contribution of physical dose of the target fragments, including those with Z>2 and thus considering their low energy, a possible correction on the biological dose in treatment plans is expected. The fragmentation spectra data will be validated against experimental data from the partner experiment FOOT on a second step. Several scenarios of TRiP98 computed plans to investigate the dose variation by accounting for the RBE variation due this effect will be presented. O 027: Differential Inflammatory and Metabolic Gene Expression Profiles of Murine and Human Cancer Cell Lines after Proton vs. Photon Irradiation S. Singhaviranon^1, H. Shen^2, M. Sertorio^2, S. Langevin^3, D. Ionescu^4, W. Kassing^4, M. Lamba^4, A. Mascia^4, E. Janssen^5, R. Vatner^6 ^1University of Connecticut School of Medicine, Immunotherapy of Cancer and Infectious Diseases, Farmington- CT, USA, ^2Cincinnati Children's Hospital Medical Center, Cancer and Blood Diseases Institute, Cincinnati- OH, USA, ^3University of Cincinnati, Department of Environmental Health, Cincinnati- OH, USA, ^4University of Cincinnati, Department of Radiation Oncology, Cincinnati- OH, USA, ^5Cincinnati Children's Hospital Medical Center, Division of Immunobiology, Cincinnati- OH, USA, ^6University of Cincinnati / Cincinnati Children's Hospital Medical Center, Department of Radiation Oncology, Cincinnati, USA Purpose: The biological effects of proton and photon radiation are assumed to be qualitatively equivalent with only the quantitative difference in relative biological effectiveness of 1.1 CGE/Gy of proton physical dose. However, the physical interactions between these modalities and cells are fundamentally different. Using whole transcriptomic analysis, we test the hypothesis that proton and photon radiation elicit qualitatively different biological responses in cancer cells. Methods: Murine embryonal rhabdomyosarcoma (M3-9-M) and human hypopharyngeal squamous cell carcinoma (FaDu) cell lines (50-80% confluent) were treated to a dose of 0, 2, or 8 Gy with photons (6 MV or 0.662 MeV) or protons (70-240 MeV SOBP), and incubated 37C for 2 or 24 hours. RNA-seq (75bp paired-end, 40M reads) was performed from cDNA libraries constructed from isolated mRNA using a HiSeq2500 (Illumina). Differential expression and pathway enrichment analysis was performed using AltAnalyze and WikiPathways, and induced genes were cross referenced against expression data from irradiated human and murine leukemia cell lines. Results: Differential gene expression patterns induced by proton vs. photon irradiation involve multiple pathways, including cell cycle control, DNA repair, inflammation, metabolism and cholesterol biosynthesis. Photons induce a pronounced relative increase in oxidative stress genes with concordant down-regulation of the cholesterol biosynthesis, consistent with shunting of NADPH towards the glutathione pathway for management of oxidative stress. Conclusion: Proton and photon radiation induce differential patterns of gene expression in both human and mouse cancer cell lines, with a pattern suggestive of a relative increase in oxidative stress induced by photons vs. protons. O 028: Comparison of the Effect of X-ray, Carbon Ion Beam and Proton Beam on Metastatic Potential R. Kondo^1, T. Adachi^1, K. Minami^2, M. Koizumi^1 ^1Osaka University Graduate School of Medicine, Department of Medical Physics and Engineering, Suita- Osaka, Japan, ^2Osaka University Graduate School of Medicine, Department of Radiation Oncology, Suita- Osaka, Japan Metastasis, the biggest threat to survival for patients with solid tumors, is the spread of tumor cells from the original growth to the other sites in the body. In the clinic, ionizing radiation has been established as a highly effective modality used in the local control of tumor growth. Recently, particle beam therapy using carbon ion or proton beam has good local control to radio-resistant carcinoma too. However, there are little known that effects of these radiations for cancer metastasis. Therefore, in this study, we investigated about effects of three types of radiations (X-ray, carbon ion beam and proton beam) on metastatic potential via in vitro and in vivo. Cell migration and invasion capability were increased by sublethal dose X-ray. On the other hand, carbon ion beam was decreased them dose-dependently. In the case of proton beam, metastatic potentials were not changed on sublethal dose and were decreased on higher dose. in vivo, X-ray irradiation resulted in a 1.2 fold increase in the number of metastatic lung nodules in mice as compared to mice injected with untreated cells. However, a significant suppression of lung metastases was observed in cells irradiated with carbon ion or proton beam. As a cause of these differences, we focused on the effect of radiations on cytoskeletal related protein. O 030: Is Deposited Energy a Useful Metric in the Quality Assurance of Pencil-Beam-Scanning Plans? R. Slopsema^1, S. Flampouri^1 ^1University of Florida Health, University of Florida Health Proton Therapy Institute, Jacksonville, USA Summation of the MU-weighted energy of all spots, over all beams in a PBS plan yields the total deposited energy. The locally deposited energy is calculated by limiting the integration to energies above the lowest spot-energy in the beam (Fig1). Because of margins and beam penumbrae, only a fraction of this energy is deposited inside the target volume. The fraction of the energy deposited outside the target volume is defined as the spill factor SF. For any PBS plan the spill factor can be calculated from the energies and MUs in the plan combined with the target volume in cc and prescribed dose. SF depends on the target volume and site-specific planning margins. Per treatment site we parameterized the SF as function of target volume by fitting to multiple treatment plans. The line in Fig2 shows this parameterization for lumbar-chordoma plans. Next we evaluated the quality of plans by determining the deviation of the SF from the parameterization. Plans that undercover the target had a SF lower than expected; plans that over-cover larger than expected. The effect of deliberate planning errors on the SF was evaluated (green/red points). This SF-analysis was performed for 125 PBS QA plans for 11 treatment sites. Overall 89% of the plans had a SF within ±10% of expected. For sites with regular-shaped targets and standardized protocols (spinal-axis, brain, prostate) the agreement with the fit was best and deliberate planning errors could be detected, indicating that for these sites deposited energy can be a useful QA metric. O 031: Proton Beam Ruler: A Fast Proton Range Measurement Tool Using a Scintillator Block and Camera D. Robertson^1, X. Ding^1, M. Bues^1 ^1Mayo Clinic Arizona, Radiation Oncology, Phoenix, USA Introduction: Proton beam range measurement is important, but it is also resource-intensive. Scintillator-based detectors can quickly and accurately measure proton range with high spatial resolution. We describe the “proton beam ruler”, a scintillator-based proton range measurement device that is optimized for clinical use. We report on its accuracy and measurement efficiency. Methods: The proton beam ruler comprises a 10x10x40 cm3 block of plastic scintillator imaged by a digital camera. Self-calibration marks are inscribed into the scintillator, which is placed in a light-blocking housing and indexed on the treatment couch (Figure 1a). Camera acquisition is triggered using signals from the synchrotron. Range measurement for 97 clinical beam energies was performed during a single irradiation using 0.1 monitor units per energy (Figure 1b). The water-equivalent thicknesses (WET) of tissue-equivalent inserts from a commercially available CT Hounsfield Unit phantom were also measured. Results: The difference in measured ranges between the proton beam ruler and a water tank with Bragg peak chamber was 0.01 mm (σ=0.05 mm) with 0.18 mm maximum deviation for scintillator measurements repeated over 5 days (Figure 2a). Setup time was 5 minutes, and measurement time for 97 beam ranges was 5 minutes. WET measurements of the Hounsfield Unit phantom inserts differed by 0.12 mm (σ=0.05 mm) from water tank measurements (Figure 2b). Each WET measurement required 10 seconds of beam time. Conclusions: The proton beam ruler can quickly and accurately measure proton range and the WET of clinical devices, decreasing quality assurance time requirements and facilitating more comprehensive beam testing. O 032: Highlights from the Second International Proton Treatment Efficiency Workshop Sponsored by the PTCOG Treatment Efficiency Subcommittee C. Beltran^1, D. Mundy^1 ^1Mayo Clinic, Radiation Oncology, Rochester, USA Purpose: The purpose of the 2nd International Proton Treatment Efficiency Workshop with a focus on Quality Assurance and Treatment Planning was to inform each other of our current clinical practices, identify current limitations, and discuss new approaches and techniques with the goal of improving the safety and efficiency in these areas. Methods: A 2.5 day meeting was held in Knoxville Tn. Short presentations on Daily, monthly, and patient specific quality assurance (PSQA) followed by in depth group discussion/debate were done on the first day. The second day covered treatment planning including robust optimization, robust analysis, verification CT imaging/planning, and image guidance for proton therapy. The last half day was used to summarize and document the discussions. Results: Nearly30 participants attended the workshop, spanning the US and Europe, some with facilities under construction and others that have well established facilities. Discussion on QA and limitations of current vended solutions were expressed. There was consensus on the merit of log based PSQA however, many proton system vendors do not allow access to the log files. Recommendation that access to meaningful log files should be negotiated upfront before purchase was made. Use of robust optimization was encouraged; limitations on speed and usability were expressed. Surface imaging was stated as a large efficiency gain for breast patients. Conclusion: Vended solutions for proton QA are still lacking and access to log files should be a priority. Daily imaging, particularly surface imaging for breast, can lead to efficiency gains. O 033: In-Room Characterization, Using an Anthropomorphic Phantom, of a Novel Monitor Exploiting Secondary Charged Particles Emission in Light Ion PT Treatments A. Sarti^1,2,3, M. De Simoni^4,5, Y. Dong^6,7, C. Mancini Terracciano^4,5, M. Marafini^2,5, R. Mirabelli^4,5, S. Muraro^8, G. Traini^4,5, S.M. Valle^6,7, I. Mattei^7 ^1Università di Roma “La Sapienza”, Scienze di Base e Applicate per l'Ingegneria, Rome, Italy, ^2Museo Storico della Fisica e Centro Studi e Ricerche “E. Fermi”, Fisica, Rome, Italy, ^3Istituto Nazionale di Fisica Nucleare, Laboratori Nazionali di Frascati, Frascati, Italy, ^4Università di Roma “La Sapienza”, Fisica, Rome, Italy, ^5Istituto Nazionale di Fisica Nucleare, Sezione di Roma, Rome, Italy, ^6Università di Milano, Fisica, Milan, Italy, ^7Istituto Nazionale di Fisica Nucleare, Sezione di Milano, Milan, Italy, ^8Istituto Nazionale di Fisica Nucleare, Sezione di Pisa, Pisa, Italy The use of C, He and O as beam particles when administering Particle Therapy (PT) treatments is getting more and more widespread as a consequence of the enhanced Relative Biological Effectiveness and Oxygen Enhancement Ratio of such projectiles. The advantages in the tumour control probability, related to the improved efficacy of the incoming radiation, require an accurate online monitor of the dose release spatial distribution. The monitor main purpose is to prevent unwanted damage to the tissues surrounding the tumour that can arise, for example, due to morphological changes occurred in the patient during the treatment with respect to the initial CT scan. PT treatments with C, He and O ions can be monitored by detecting the secondary radiation produced by the primary beam interactions with the patient body along the path towards the target volume. Secondary charged fragments (produced mainly by the projectile fragmentation) can be emitted at very large angles with respect to the incoming beam direction and can be detected with high efficiency in a nearly background free environment. The Dose Profiler (DP) detector, developed within the INSIDE project, is a scintillating fibre tracker that allows an online charged fragments reconstruction and backtracking. The construction and preliminary tests performed on the DP, carried out using the 12C ions beam of the CNAO treatment centre using a RANDO® anthropomorphic phantom as a target, will be reviewed in this contribution. A discussion of the results implications for a pre-clinical trial on CNAO patients, foreseen in 2018, will be made. O 034: Scintillator-Based System for Transversal Dose Pro Files Reconstruction R. Catalano^1, G. Petringa^1,2, G. Cuttone^1, M. Durante^3, G. Pitta^4, S. Puglia^1, L. Raffaele^5, E. Scifoni^3, F. Tommasino^3, P. Cirrone^1,6 ^1Istituto Nazionale di Fisica Nucleare, Laboratori Nazionali del Sud, Catania, Italy, ^2Università degli Studi di Catania, Dipartimento di Fisica e Astronomia, Catania, Italy, ^3INFN, Trento Institute for Fundamentals Physics and Applications, Trento, Italy, ^4DE.TEC.TOR., Devices and Technologies Torino s.r.l., Torino, Italy, ^5Università degli Studi di Catania, Azienda Ospedaliero-Universitaria Policlinico, Catania, Italy, ^6Institute of Physics ASCR, ELI-Beamlines Project, Prague, Czech Republic A fast and reliable system to measure transversal charged particles relative dose profiles is desirable in any hadron therapy facility, being the basis for an accurate treatment quality assessment procedure. For this purpose a system was developed at the “Laboratori Nazionali del Sud” of Italian Institute for Nuclear Physics (INFN-LNS, Catania, I) consisting of a plastic scintillator screen (50 × 50 mm2, variable in thickness from 0.5 mm to 1.0 mm), mounted perpendicularly to the beam axis and coupled with a highly sensitive cooled CCD detector (resolution 1928 × 1452 pixels) in a light-tight box. In-house custom software for real-time data acquisition and processing was also developed in the LabView 2016 environment. In this work the characterization of the system is reported: transversal dose profile in terms of FWHM, field lateral penumbra, flatness, symmetry, reproducibility and linearity with beam current have been investigated. A comparison with other common quality control devices able to perform transversal beam profiles reconstruction (Gafchromic films, silicon diodes, Lynx detector and Timepix detector) has been also carried out both in small- and large-field proton beams at the INFN-LNS proton therapy facility (Catania, Italy). No significant differences have been observed in respect to conventional systems at both the INFN-LNS proton therapy facility (Catania, Italy) and the Trento Institute for Fundamental Physics and Applications (TIFPA, Trento, I). Results will be reported and discussed. O 035: Implementation of a Non-Invasive Online Beam Monitor at a 60 Mev Proton Therapy Beamline R. Schnuerer^1, J.S.L. Yap^1, H. Zhang^1, G.J. Haefeli^2, O. Girard^2, M. Hentz^3, T. Szumlak^4, C.P. Welsch^1 ^1Cockcroft Institute/ University of Liverpool, Physics, Warrington, United Kingdom, ^2École Polytechnique Fédérale de Lausanne, Physics, Lausanne, Switzerland, ^3University College London, Physics, London, United Kingdom, ^4Akademia Górniczo-Hutnicza, Physics, Krakow, Poland Online beam monitoring in medical accelerators is essential in assuring patient safety as well as the high quality and efficacy of cancer treatment. In clinical practice for proton therapy, currently used ionization chambers are interceptive devices, degrading both the beam profile and its energy spread. Therefore, a new non-interceptive approach of online beam monitoring is highly desirable. The Vertex Locator (VELO) detector is a multi-strip silicon detector used in the LHCb experiment at CERN. The semi-circular design and position of its sensitive silicon detector offers a non-invasive way to measure the beam intensity through a precise measurement of the beam halo without interfering with the beam core. To be integrated to the specific conditions of the clinical environment in a proton therapy center, VELO was adapted in a standalone system. In this contribution, Geant4 Monte Carlo simulations of the beam parameters including the beam profile, energy spread and halo to dose correlations were performed to investigate the integration of the detector in the treatment line and behavior of the beam during delivery. Furthermore, initial results with the VELO detector integrated at the 60 MeV proton therapy beamline at the Clatterbridge Cancer Centre (CCC), UK are presented. Synchronized with a locally constructed Faraday Cup, the quality of the beam monitor is assessed by measuring the beam current at different dose rates and by monitoring the beam halo profile at different positions along the beamline. Future beam and halo propagation studies will look into further application for the VELO standalone online monitor. O 036: Prospective Study of 3 Fraction Pencil-Beam Scanning Partial Breast Irradiation: Early Provider and Patient-Reported Outcomes of a Novel Regimen K. Jethwa^1, K. Gonuguntla^1, T. Whitaker^1, S. Park^1, T. Hieken^2, L. McGee^3, K. Ruddy^4, K. Corbin^1, N. Remmes^1, R. Mutter^1 ^1Mayo Clinic, Radiation Oncology, Rochester- MN, USA, ^2Mayo Clinic, Surgery, Rochester- MN, USA, ^3Mayo Clinic, Radiation Oncology, Scottsdale- AZ, USA, ^4Mayo Clinic, Medical Oncology, Rochester- MN, USA Purpose/Background: The optimal dose and fractionation for proton accelerated partial breast irradiation (APBI) is not known. Herein, we report dosimetry, early adverse effects (AEs), cosmetic, and patient-reported outcomes (PROs) of a prospective study of 3-fraction pencil-beam scanning (PBS) APBI. Methods: Eligibility criteria included women age ≥ 50 years with estrogen receptor positive (ER+), sentinel lymph node negative invasive or in-situ breast cancer measuring ≤ 2.5 cm. The prescription was 21.9 Gy (RBE 1.1) in 3 daily fractions to the post-operative tumor bed with a 1 cm expansion. Common toxicity criteria for adverse effects (CTCAE v 4.0), 10-point linear analog scales (LASA), patient-reported outcomes version of the CTCAE (PRO-CTCAE), and Harvard Breast Cosmesis Scale (HBCS) were utilized for provider and patient-reported assessments. Results: 76 women were treated between 2015 and 2017. The mean age was 66 (SD 8.6) years. Most patients had grade 1-2 (92%) invasive breast cancer (80%), of ductal histology (92%), measuring ≤2 cm (95%). A median of 2 (range 1-3) treatment fields were used to achieve favorable target coverage and normal tissue dosimetry (table 1). Mean follow-up was 6 months (SD 5.8). Physician-assessed AEs and PROs are shown in table 2. At last follow-up, no treatment-related grade ≥ 2 AEs have been observed, and all patients are alive without relapse. Conclusions: 3-fraction PBS-APBI is well tolerated with low rates of physician and patient-reported adverse effects. Follow-up is ongoing to assess the late toxicity and disease control outcomes of this novel strategy for APBI. O 037: Improved Long-Term Patient-Reported Health and Well-Being Outcomes of Early-Stage Breast Cancer Treated with Partial Breast Proton Therapy S. Teichman^1, S. Lum^2, J. Slater^1, D. Bush^1 ^1Loma Linda University Medical Center, Radiation Medicine, Loma Linda, USA, ^2Loma Linda University Medical Center, Surgical Oncology, Loma Linda, USA Background: Long-term quality of life (QoL) is important in assessing outcomes of breast conservation therapy for patients with early-stage breast cancer. This study compares patient-reported QoL outcomes among women with stage 0-II disease treated via lumpectomy followed by whole-breast photon (WBI) or partial-breast proton therapy (PBPT). Methods: Subjects receiving WBI or PBPT were recruited from institutional research database and prior phase II clinical trials. Participants evaluated QoL several years post-treatment by responding to subjective instruments, including established scalar questionnaires and self-report measures. Responses were averaged between the two groups for statistical analysis. Results: 129 subjects completed QoL evaluations, 57 following WBI and 72 following PBPT. At 6.5 years (median) post-diagnosis, participants' demographic and clinical characteristics were similar. Patient-reported outcomes were reported as mean scale scores for the two groups, all displaying significant differences favoring proton irradiation: Cosmesis (p<0.001); pain/sensitivity(p<0.05); breast texture or shape (p<0.001); clothing fit (p<0.001); fatigue (p<0.001); daily-life fatigue impact (p<0.05 to p<0.001); self-consciousness (appearance dissatisfaction) (p<0.05); attractiveness self-opinion (p<0.05); contentedness (p<0.01); fear of recurrence (p<0.001); and happiness with treatment choice (p<0.001). Conclusion: Patients' responses suggest that partial-breast proton radiotherapy is associated with higher overall QoL compared with WBI. These self-perceptions prevail up to >10 years post-treatment. Proton therapy may enhance QoL in a cascade of ways. O 038: Initial Experience with Barrier Film Dressing for Dermatitis Prophylaxis in Breast Proton Therapy K. Corbin^1, K. Roberts^1, S. James^1, S. Park^1, E. Yan^1, K. Klein^1, J. Lubahn^1, C. Loprinzi^1, I. Petersen^1, R. Mutter^1 ^1Mayo Clinic, Radiation Oncology, Rochester, USA Purpose/Background: Radiation dermatitis is a common acute toxicity of breast cancer radiotherapy. Mepitel film (MF) appears to reduce acute radiation dermatitis (RD) in photon patients. We report the feasibility and initial outcomes of PMRT patients treated with proton beam (PBT) and MF. Methods: Patients were treated with multi-field optimized pencil-beam scanning IMPT to the chest wall and regional lymph nodes, most commonly with a 2-field oblique beam arrangement, per institutional standards. The median prescription dose was 50 Gy (RBE 1.1) in 25 fractions. Skin dose was individualized, with typical planning goals of D90>90% and D1cc < 105%. MF was applied by nurses prior to treatment and replaced approximately weekly. Common toxicity criteria for adverse effects (CTCAE v 4.0), patient-reported outcomes (PRO) version of the CTCAE (PRO-CTCAE), and 10-point linear analog scales (LASA) were collected prospectively to evaluate skin toxicity. Results: 33 patients underwent proton PMRT with MF. 20 were reconstructed with tissue expanders and 13 were non-reconstructed. Four of 33 patients discontinued for pruritis (2), film non-adherence (1), and rash unrelated to MF (1). Median skin D1cc was 103.5% (Range: 95.8-120%). Physician assessed and patient reported outcomes are shown in Table 1. Maximal physician reported RD was grade 1 -2 in 97% at completion. Just 9 patients reported severe or very severe radiation burns. Conclusions: Post mastectomy IMPT with MF was feasible, resulting in promising patient and physician reported acute toxicity. Based on this preliminary experience, we are planning a randomized trial of MF in PMRT. O 039: Proton Beam Therapy for Hepatocellular Carcinoma Associated with Inferior Vena Cava Tumor Thrombus Y. Sekino^1, N. Fukumitsu^1, T. Okumura^1, H. Numajiri^1, M. Mizumoto^1, K. Ohnisihi^1, T. Aihara^1, H. Ishikawa^1, K. Tsuboi^1, H. Sakurai^1 ^1University of Tsukuba, Radiation Oncology, Tsukuba, Japan Background: Patients with hepatocellular carcinoma (HCC) associated with inferior vena cava tumor thrombus (IVCTT) have poor prognosis. The purpose of this study is to evaluate the safety and effectiveness of proton beam therapy (PBT) for patients with HCC associated with IVCTT. Methods: Fifteen patients who had HCC with IVCTT were treated with PBT at University of Tsukuba between 2005 and 2014. Median tumor size was 9 (range 4-20) cm. Tumor thrombus extended to right atrium in 10 patients. Portal vein tumor thrombosis was found in 4 cases. Total delivered dose ranged from 50 to 74 Gray equivalent (GyE) and the most frequently used dose-fractionation was 72.6 GyE in 22 fractions. Seven patients were treated with curative intent. The local tumor control rates (LC), overall survival rates (OS), and toxicities were evaluated. Results: Median follow-up was 20.1 months. The OS and LC rates at 1 year were 66.7% and 85.1%, respectively. For patients with curative intent, median OS was 25.1 months and LC was achieved in all cases. Recurrence was observed in 12 patients, intrahepatic recurrence in 5 patients and lung metastases in 5 patients. Lung metastases occurred within 3 months in 4 cases. No patients were suffered from radiation-induced liver disease or other severe toxicity of grade 3 or greater. One patient had grade 2 gastrointestinal hemorrhage. Conclusions: PBT showed favorable effectiveness and safety, and can play a role in treatment for HCC associated with IVCTT. O 040: Proton Radiotherapy for Locally Advanced Pancreatic Cancer (LAPC) – Limiting Duodenal Toxicity P. Vitek^1, J. Kubes^1, V. Stepan^1, V. Vondracek^2, J. Kvech^1, B. Ondrova^1, K. Dedeckova^1, A. Pasztorova^1, G. Kasacova^1 ^1Proton Therapy Center Czech, Proton radiotherapy, Praha 8, Czech Republic, ^2Proton Therapy Center Czech, Dept. of Physics, Praha 8, Czech Republic Introduction: Dose escalation and concomitant chemotherapy for LAPC may improve local control and prolong survival. The duodenum, adjacent to pancreas head, potentially is the most critical organ. If the regional lymph-nodes (LNN) are encompassed to CTV, the substantial volume of D1-D4, receives the full prescribed dose. The dosimetry of pencil beam scanning (PBS) proton radiotherapy was analysed and correlated to the observed toxicity. Methods: Proton PBS radiotherapy was administered in 30 patients with inoperable LAPC with concomitant chemotherapy (capecitabine or gemcitabine), hypofractionated regimen – 54 GyE/18 fractions. The regional lymph nodes - peripancreatic, coeliac, portal, and upper paraaortic group were encompassed CTV. It extended caudally beyond lower-most margin of duodenum - to lower edge of L2 or L3. Results: The median volume of delineated duodenum was 53 ccm (24 – 118) median of Dmean in duodenum 37 GyE, median of V45GyE 29 ccm (2,8 -50 ccm). The medians of V45GyE for stomach and small intestine were 21,5 and 8,0 ccm respectively. (Table 1). Signs of acute duodenitis developed in 6 patients, duodenal ulceration in 3 patients, perforation in 1, with a fatal outcome. The time of ulceration onset was 3-6 months after radiotherapy. Other side effects did not exceed grade 2. Conclusions: Duodenal toxicity limits the dose and extent of PTV for LAPC. The anatomy of pancreas and LNN allows only limited saving of duodenum. Low duodenal dose may be maintained if LLN are included in CTV. A separate dose constraint of duodenum and small intestine for normo and hypofractionated radiation is warranted. O 041: Is Proton SBRT an Attractive Option for Liver Metastases? Results of a 5-Year Observational Study for 80 Hepatic Lesions D. Sufficool^1, P. McGee^1, J. Slater^1, G. Yang^1 ^1Loma Linda University, Radiation Medicine, Loma Linda, USA Purpose/Objectives: To evaluate local control outcomes of liver metastases treated with proton stereotactic body radiotherapy (SBRT). Secondary objectives include evaluation of toxicity. Materials/Methods: Patients diagnosed with metastatic disease to the liver treated with proton SBRT between 2012 and 2017 were included. Tumor motion was accounted for using either respiratory management or 4D CT scan. In-field local control was evaluated according to RECIST 1.1 criterion. Radiation induced liver disease was evaluated using CTCAE 5.0 criterion. Results: A total of 41 patients with 80 lesions were included in this analysis with a median age of 65.5 years (range 33-87). Dose/fractionation included 50GyE / 5fx (19%), 30GyE / 5fx (16%), and 30GyE / 3fx (16%). The primary cancer site was 51% colorectal, 17% breast, and 37% other. The number of metastases treated ranged from 1-5 with mean tumor diameter of 2.7 cm ± 1.6 cm (range 0.7 – 7.3). With a mean follow up of 14.5 months (range 1 – 44), local control at 6 months and 1 year was 92% and 75%, respectively. Grade 1 and grade 2 toxicities were 38% and 6% respectively. No patient developed grade 3 or higher toxicity. No radiation induced liver disease was detected. Conclusions: Proton SBRT for liver metastases demonstrated a high rate of local tumor control with minimal toxicity. An ongoing institutional phase II trial to further validate current findings will also be updated at time of presentation. O 042: Innovative Strip Silicon Detectors for Proton Beam Monitoring: Preliminary Results A. Vignati^1, O. Hammad Ali^2, A. Attili^1, M. Donetti^3, F. Fausti^1, S. Giordanengo^1, L. Manganaro^2, V. Monaco^2, R. Sacchi^2, R. Cirio^2 ^1National Institute for Nuclear Physics INFN, Turin division, Turin, Italy, ^2National Institute for Nuclear Physics INFN- Università degli Studi di Torino, Physics Department, Turin, Italy, ^3Fondazione CNAO, Medical Physics, Pavia, Italy Unlike the legacy gas ionization chambers, solid state detectors offer large granularity and sensitivity to single protons and would ideally be suited for beam monitoring in therapy applications. However, signal pileup, radiation damage and the readout complexity prevented their use so far on high flux therapeutic beams. Innovative silicon low-gain avalanche detectors optimized for time resolution (Ultra Fast Silicon Detectors - UFSDs) were recently proposed, where sensors as thin as 50 μm provide signals of ∼ 1ns time duration with time resolutions of tenths of ps, and large enough signal-to-noise ratio to efficiently discriminate the proton signal. The MoVeIT project of the INFN is investigating the use of this technology in Particle Therapy by developing two monitoring devices, one to directly count individual protons at high rates, the second to measure the beam energy with time-of-flight techniques. This requires the design of custom UFSD sensors as well VLSI readout electronics. From the simulations' results and the first beam tests with UFSD pads, strip detectors were produced by FBK in Trento, with two geometries (30 mm and 15 mm length) and different doping modalities to improve radiation hardness. In parallel, prototypes of a new TERA10 readout chip have been submitted to the foundry. The aim of this contribution is to review the advancement of the project and to report on the results of the tests of UFSD strip sensors with the therapeutic proton beam of CNAO. O 043: Dynamic Beam Current Control for Improved Dose Accuracy in PBS Proton Therapy C. Bula^1, M. Eichin^1, J. Hrbacek^1, M.F. Belosi^1, D. Meer^1 ^1Paul Scherrer Institut, Center for Proton Therapy, Villigen, Switzerland The step-and-shoot method of pencil beam scanning applies the dose on a three-dimensional grid in the target volume, with one dimension defined by the proton energy. While the spot dose may vary substantially within an iso-energy layer, the beam current typically remains constant. In this static operation mode, the inherent latency of the beam switch-off mechanism results in a lower limit for the deliverable spot dose, which may conflict with part of the low-weighted spots prescribed by the treatment planning system. To overcome this limitation, we enhanced the control system of the PSI Gantry 2 with a direct link to the vertical deflector located at the center of the cyclotron. This connection allows much faster beam current changes (∼ 0.1 ms) and hence opens the possibility of temporarily reducing the current for individual low-dose spots. This new dynamic operation mode improved the accuracy of the delivered dose compared to the planned distribution without compromising treatment time. Figure 1 shows an example field where 4% of the spots (0.4% of dose) were skipped in the static mode, while the dynamic mode allowed to deliver all spots and reduced the maximum missing dose per voxel from 2.3% to 1.3%. The method was successfully commissioned and is in clinical operation since fall 2017. We consider dynamic beam current control to be a valuable contribution to cyclotron-based spot scanning technology, especially in the context of new modalities such as rescanning and high-intensity deliveries, where the number of low-weighted spots is even more pronounced. O 044: Commissioning of a Unique Penumbra Sharpening Adaptive Aperture of HYPERSCAN M. Kang^1, H. Chen^1, R. Cessac^2, D. Pang^1 ^1Georgetown University Hospital, Radiation Medicine, Washington DC, USA, ^2Mevion Medical System, Engineering, Littleton- MA, USA Purpose: To present the commissioning of a unique adaptive aperture(AA) on the penumbra sharpening for proton pencil beam scanning treatment. Materials and Methods: The AA is essentially a mini-MLC consisting of 7 pairs of leafs made of nickel. The inner 5 and the outer 2 pairs have width of 0.5 and 2 cm, and the thickness of the leaves in the beam direction is 10 cm(Fig1(a)). The leaves follow the scanning spots and trim the spots on the periphery of a field to sharpen the penumbra. Both single spot and uniform spot maps of square fields were created with and without AA for typical therapeutic energies. A PTW Octavius ion chamber array was used to measure the absolute dose and field profiles at the Isocenter plane. Results and Discussion: Fig1 (b) shows the ratio of the absolute dose between with and without AA can be 5% different, which is energy, field size and nozzle position dependent. Fig 2(a)-(c) show that the penumbra and dose reduction by using AA for single spots. AA can reduce the penumbra by 2 mm to 12 mm from highest energy 227 to 64MeV for square fields. The significant penumbra reduction for low energy beams make it feasible to use very low energy scanning proton to achieve conformal dose distribution without using any external range shifter. However, the Raystation TPS currently cannot model the actual MLC motion patterns in planning so the final delivery MUs should be verified from measurement and adjusted by using appropriate correction factors. O 045: External Beam Radiotherapy for Ocular Melanoma: A Comparative Study Low Energy Protons vs. Pencil Scanning High Energy Protons vs. Cyberknife A. Gerard^1, M.L. Peyrichon^1, C. Peucelle^1, M. Vidal^1, A. Claren^1, S. Wolfgang^2, A. Carnicer^1, G. Angellier^1, J. Thariat^3, J. Herault^1 ^1Centre Antoine Lacassagne, Department of Radiation Oncology, Nice, France, ^2Strahlenklinik- University Hospital, Department of Radiation Oncology, Essen, Germany, ^3Centre François-Baclesse, Department of Radiation Oncology, Caen, France The Centre Antoine Lacassagne (CAL) hosts all three facilities potentially competitive for radiotherapy of ocular tumors. Therefore, CAL can identify the advantages and disadvantages of each type of facility. Based mainly on radiochromic films irradiation, the study analyzes dosimetric data for both proton facilities and Cyberknife using respectively one and multiple beam incidences. Two clinical cases were considered: a conjunctival melanoma and a posterior uveal melanoma. The contributions of beam modifiers as well as dosimetric planning data were analyzed. The doses from primary and secondary particles received by distant organs were studied. Beyond static image views, the influences of eye positioning systems were also analyzed. Always related to positioning, the irradiation duration, the tracking system as well as eyelids sparing were compared. The study is carried out disregarding clinical results as well as economic imperatives and is based solely on dosimetric data and positioning quality. Based on dosimetric and positioning criteria, stereotactic photon irradiation delivered for example by Cyberknife is not suitable for the treatment of ocular malignancies. Irradiation with proton beams leads to a more favorable dose distribution. The use of a high-energy proton pencil beam scanning however requires additional collimator and positioning accessory equipment and fails to compete with both distal (6.7 vs. 0.6 mm) and lateral (8.7 vs. 1.6 mm) 20-80% penumbras of a low energy proton beam. In addition, its prohibitive treatment duration for 15 GyEBR (5 min vs. 10 s) jeopardizes an effective movement limitation resulting larger irradiated volumes and higher doses to OARs. O 046: Utilization of an Isocentric Rotating Chair for Treating Head/Neck Cancer at Seating in Fixed Particle Beamline W.C. Hsi^1, R. Zhou^2, Z. Wang^2, F. Yang^2, X. Zhang^2, J. Sun^1, S. Yinxian^1, H. Xue^3, M. Wang^3, F. Liu^3 ^1Shanghai Proton and Heavy Ion Center, Medical Physics, Shanghhai, China, ^2Sichunan University, College of physical science and technology, Chengdu, China, ^3Jiangsu Supersense Technology Co. LTD, Technology, Suzhou, China Purpose: Present systematic evaluation of chair with six-degree-freedom (6DoF) isocentric movement for head/neck cancers at seating positioning in fixed carbon-ion and proton beamline. Methods and Materials: When patients are well re-positioned at chair, a movement around isocenter required only rotation along vertical axis with a translation in IEC fixed coordinate; referred beam-angle vector. Any re-positioning error required additional 6 DoF displacement on IEC patient-support coordinate; referred as a correction vector. A patient-support coordinate is rotated accordingly to required beam angle. In Fig. 1 for built chair, mechanisms of a 360rotating platform for beam-angle movement, and a six-pod 6DoF movement for correction vector was utilized. The mathematical presentation for combined beam-angle movement and correction vector was evaluated to be coincided with mathematical formula used in X-ray based image-guiding positioning system during acceptance tests. Results: In Fig. 2, a 4x4 matrix to execute a defined sequence rotation (3x3 matrix) of 3 axes followed by a translation vector (3x1 vector) was found to a unique solution for any 6 DoF movement in fixed coordinate. This 4x4 matrix was also proved to equivalently present a translation following by a sequence rotation around 3-rotated axes in patient-support coordinate. Any movement combined with beam-angle movement and correction vector could be presented by six-parameter set in fixed coordinate. However, accuracy of immobilization was evaluated by extracted six-parameter set for correction vector. Conclusions: With understanding of mathematical characteristics of beam-angle movement and correction vector in separated mechanisms used in chair, treated patients can be correctly positioned. O 047: Clinical Implementation of the Patient-Specific Apertures Used in Commercial Proton Pencil Beam Scanning Systems C. Chen^1, H. Liu^1, S. Luckman^1, E. Readdy^1, K. Hall^1, Z. Han^1, N. Ju^1, N. Schreuder^2, D. Mah^1 ^1ProCure Proton Therapy Center, Medical Physics, Somerset, USA, ^2ProVision Center for Proton Therapy, Medical Physics, Knoxville, USA Purpose: Patient-specific apertures for proton pencil beam scanning (PBS) were commissioned using commercially available beam delivery and treatment planning systems (TPS). We performed: 1. collinearity tests amongst the X-ray imaging, spot steering and the aperture mount; 2. dose verification of the Monte Carlo algorithm including aperture transmission and scatter; 3. refinement of the patient-specific QA procedure to verify overall implementation. Materials and Methods: The gantry-dependent spot steering was tested using star-shot films. A uniform PBS field with a single centralized spot was generated and over-scanned on a square aperture (Fig. 1) to measure the isocenter offsets in between the PBS spot positioning and aperture center. The measured depth-doses and lateral profiles of aperture-shaped fields were compared with TPS calculations. Finally, clinical PBS aperture-collimated plans were generated and patient-specific QA procedure was refined to account for the larger ion chamber spacing on the 2D array. Results: The star-shot diameters were 0.14 mm and 0.34 mm for gantry rotated clockwise and counterclockwise respectively (Fig. 2). The offsets in between the spot positioning and aperture center were <1 mm. Acceptable dose agreements were found in the dose comparisons for the aperture-shaped fields. It was noticed that the TPS created heavily weighted spots abutting the aperture edge to ensure coverage. Conclusions: The use of PBS aperture-shaped fields reduces the penumbra significantly at shallow depths permitting sparing of OARs for selected cases. The alignment in between the spot positioning and aperture shaping should be carefully verified to ensure accurate delivery. O 048: A Novel Deliver Sequence and Efficiency Optimization Algorithm for Spot-Scanning Proton Arc Therapy X. Ding^1, X. Li^1, J. Zhou^1, C. Stevens^2, Y. Di^2, P. Kabolizadeh^1 ^1Beaumont Health, Radiation Oncology Proton Therapy Center, Royal Oak, USA, ^2Beaumont Health, Radiation Oncology, Royal Oak, USA Purpose: Spot-Scanning Proton Arc therapy (SPArc) has been a great interest of the society because of the improved dosimetric outcome. However, there are a lot of technique challenges to implement it in the clinical settings especially the delivery efficiency. Due to the hysteresis and technical difficulties in changing the beamline magnetic field, it costs significant time in the energy layer switching especially switching from low to high energy. Thus, we presented a new energy sequence optimized SPArc algorithm(SPArc_seq) to shorten the delivery time of the SPArc. Material and Methods: SPArc_seq includes an energy layer sorting and control point re-sampling mechanism taking into account of delivery sequence through the gantry rotation. It is optimized for high to low energy delivery sequence instead of random layer switching. Both SPArc and new SPArc_seq were tested on 5 prostate patients. Both plans were delivered at a fixed 0 degree gantry angle. Total actual delivery time was recorded and dose measurements were performed using MatriXXONE at 3cm depth. Results: The result showed that with similar plan quality, SPArc-seq (mean beam-on-time: 330 seconds) was able to successfully reduce about 56% of delivery time compared to SPArc (mean beam-on-time: 756 seconds) (p<0.01). Absolute dose measurements showed within 2% difference compared to the plans. 2D Gamma Index(3%/3mm) showed more than 97% passing rate in both SPArc and SPArc_seq. Conclusion: This is the first study to test the SPArc delivery feasibility at a fixed gantry angle. SPArc_seq algorithm could effectively reduce the total treatment delivery time while keeping the similar plan quality. O 049: Superconducting Gantry for Proton Therapy with a Large Momentum Acceptance K.P. Nesteruk^1, C. Calzolaio^1, A. Gerbershagen^2, D. Meer^1, V. Rizzoglio^1, M. Seidel^1, J.M. Schippers1 1Paul Scherrer Institut, Large Research Facilities GFA, Villigen, Switzerland, 2CERN, Engineering Department, Geneva, Switzerland Utilization of superconductivity for proton therapy gantries allows them to be much lighter and somewhat smaller. These features bring obvious advantages, especially for new hospital-based facilities. In our design we have included the degrader into the gantry. Since such a gantry will not need an external degrader and energy selection system, a substantial footprint reduction can be achieved in new cyclotron-based facilities. In addition to this, superconducting magnets deployed for a gantry can also significantly improve the treatment process. For our development of a new SC gantry, we have designed specific beam optics with a very large momentum acceptance, enabled by using superconductivity. This achievement allows a wide energy spectrum to be transported so that a typical depth range in the order of +/- 30% can be covered without changing the magnetic field of these magnets. Such a property enables a decrease of the treatment time. In particular, the large momentum acceptance combined with a ridge-filter instead of a degrader, would give a possibility of using multiple energies in the gantry at the same time, and consequently, would allow an SOBP to be delivered at once. Moreover, with the potentially achievable treatment time in the order of a few seconds, various new techniques can be employed to deal with organ motion. O 051: 4D Robust Optimization Including Time Structures Can Reduce the Interplay Effect in Proton Pencil Beam Scanning Radiotherapy E. Engwall^1, A. Fredriksson^2, L. Glimelius^1 ^1RaySearch Laboratories AB, Physics, Stockholm, Sweden, ^2RaySearch Laboratories AB, Research, Stockholm, Sweden Interference between the delivery time structure and organ motion can create large distortions in the dose distributions of proton pencil beam scanning radiation therapy, and is known as the interplay effect. Standard methods for mitigating this effect include abdominal compression, rescanning and gating. We propose a new method (implemented in a research version of the RayStation treatment planning system), where the time structures of the delivery and the organ motion are included in a 4D robust optimization. The time structures are used to distribute the pencil beams over the different breathing phases and partial beam doses delivered to each phase are calculated. The partial beam doses are accumulated on a reference phase using deformable image registration. Uncertainties in the delivery and organ motion are handled by the inclusion of multiple scenarios in the robust optimization implemented by means of minimax optimization. In this study, we limit the uncertainties to variations in the breathing pattern. The method is evaluated for three different non-small cell lung cancer patients with different motion amplitudes. The ability of the method to reduce dose distortions from interplay is compared to results from standard 4D robustly optimized plans with and without rescanning (see Figure 1). Our study shows that the new method is beneficial, especially for large tumor motion, where rescanning alone cannot mitigate the interplay effect. The most efficient mitigation is achieved when our method is combined with rescanning. O 052: Robust Treatment Planning with 4D Intensity Modulated Particle Therapy for Multiple Targets in Stage IV Non-Small Cell Lung Cancer M. Wolf^1, C. Graeff^1, K. Anderle^1 ^1GSI Helmholtz Centre for Heavy Ion Research, Biophysics, Darmstadt, Germany Purpose: Recently, we introduced a 4D IMPT optimization approach handling multiple targets and including all motion states of a 4DCT. This method was expanded by a robust non-linear biological optimizer for carbon ions – accounting for setup and range uncertainties – to explore its potential in improving plan robustness and sparing of critical organs. Methods: The implemented worst case scenario method considers 9 different scenarios: nominal scenario, under- and overestimation of particle ranges (±3.5%) and isotropic shifts of patient's isocenter (±3 mm in the 6 major anatomical directions). Within the 4D IMPT optimization approach, robust IMPT on CTV is compared to conventional IMPT with 3 mm margins using 4D dose calculation and robustness analysis with 21 different uncertainty cases. Rescanning is used for motion mitigation. Results: Both optimization methods are compared in two lung cancer cases with multiple lesions in proximity to critical structures, like smaller airways (SA). By using robust IMPT the given constraints for SA could be fulfilled in 98.4% of all regarded cases compared to just 55.5% with conventional IMPT. This was enabled by reducing average target coverage (see D99 values in Table 1) which decreased the D0.5cc for SA from 56.8±4.1% to 46.6±2.4% averaged over all cases. Additionally, the uncertainty bands in the DVHs could be reduced as well (Fig.1). Conclusion: The study above showed that plan robustness could be improved by using robust optimization in 4D IMPT. Focusing on improved OAR sparing implicates a decrease in target coverage in the majority of the regarded uncertainty scenarios. O 053: Evaluation of 4D Plan Robustness Using Lung 4DCT(MRI) M. Krieger^1,2, J. Käser^2, R.L. Perrin^1, M. Peroni^1, O. Bieri^3,4, Z. Celicanin^3,4, D.C. Weber^1,5, A.J. Lomax^1,2, Y. Zhang^1 ^1Paul Scherrer Institute, Center for Proton Therapy, Villigen PSI, Switzerland, ^2ETH Zürich, Department of Physics, Zürich, Switzerland, ^3University of Basel Hospital, Department of Radiology- Division of Radiological Physics, Basel, Switzerland, ^4University of Basel, Department of Biomedical Engineering, Basel, Switzerland, ^5University Hospital Zürich, Department of Radiology, Zürich, Switzerland Purpose: 4D planning for lung cancer is typically performed using a single 4DCT. Here we demonstrate the potential variability in 4D dose-distributions for PBS lung treatments under variable breathing conditions by using the 4DCT(MRI) approach applied to lung cancer cases. Materials: Lung motion deformation fields from 10 breathing cycles have been extracted from 4DMRI datasets of 2 volunteers. From these, 10 synthetic 4DCT(MRI) datasets, one for each extracted breathing cycle, were generated for each motion scenario by deforming the end-exhale phase of a patient 4DCT with the extracted 4DMRI motion fields. A 2Gy(RBE), 3-field PBS plan was optimised to the geometric ITV on the static CT, and 4D dose calculations (4DDCs) were performed on the different 4DCT(MRI) datasets. For comparison, 4DDCs were also performed assuming perfectly periodic motion by repeating the first breathing cycle of each 4DCT(MRI) in the 4DDC. Results: Figure 1 shows the median SI motion of the lung for the two volunteers, comparing the repeated motion to the actual (variable) motion, with the differences in the resulting 4D dose-distributions shown in Figure 2. Dose differences in the ITV of close to +/-20% (mean absolute difference (SD): 3(9)%) are observed between periodic and variable motions from the same volunteer (Figure 2b), and of up to 50% (mean absolute difference (SD): 20(40)%) between the two different motion scenarios (Figure 2c and d). Conclusion: Calculating a lung plan under different motion scenarios is important to capture the range of possible effects of breathing. 4DCT(MRI) could be a valuable tool for this purpose. O 054: Alternating Intra-Field Scan Direction in Rescanning for Improved Motion Mitigation G. Fattori^1, G. Klimpki^1, Y. Zhang^1, M. Krieger^1, J. Hrbacek^1, D.C. Weber^1, A. Lomax^1, S. Safai^1 ^1Paul Scherrer Institute, Center for Proton Therapy, Villigen, Switzerland For moderate organ motion, rescanning is effective in averaging dose distortions due to interplay. However, unsought positional and temporal correlations between patient breathing and the dynamics of rescanning may arise, undermining its efficacy. Here we investigate the effectiveness of systematically changing meander direction within the field to increase interplay mitigation due to rescanning. Alternation of the meander path for rescanning can be performed by either switching the primary direction of scanning between each energy layer or between each volumetric rescan, both of which have been experimentally investigated. Using a platform-mounted ionisation chamber array, measured doses were compared to those of a stationary delivery. The detector was moved to replicate a cranio-caudal target displacement (ca. 6 mm) of a liver carcinoma patient (PTV 76.59 cm3), and the conventionally generated control files modified to scan either parallel or orthogonally to the motion, or to alternate between energy layers (EE) or between each rescan (ER). Results from central plane measurements demonstrate that, to achieve a high gamma pass rate (∼90% at 1%/1mm), a substantially smaller number of rescans was required when using ER (4x) compared to best-case conventional (non-alternating) rescanning (8x), and that ER was marginally more effective than EE. When introducing additional random amplitude and breathing fluctuations however, agreement was compromised for all scenarios, but was still consistently higher for the EE and ER scenarios (87.2%/95.7% pass rates for 8x) than conventional rescanning (best case 71.7% for parallel re-scanning). In conclusion, alternating scanning directions during re-scanning can further help mitigate interplay effect. O 055: Feasibility of Noninvasive Cardiac Ablation Utilizing Intensity Modulated Proton Therapy to Treat Ventricular Tachycardia S.M. Goddu^1, J. Hilliard^1, N. Knutson^1, T. Zhao^1, P. Samson^1, G. Hugo^1, S. Mutic^1, J. Bradley^1, P. Cuculich^2, C. Robinson^1 ^1Washington University, Radiation Oncology, Saint Louis, USA, ^2Washington University, Cardio-vascular Division, Saint Louis, USA Introduction: We have recently demonstrated safety and feasibility (Cuculich et al, NEJM-2017) of using noninvasive single-fraction photon-based-SBRT to treat patients with refractory, life threatening, ventricular-tachycardia (VT). Proton Therapy has the potential to reduce dose to non-target-heart-tissue (NT-HT). We evaluated the feasibility and potential dosimetric improvements using proton-beam-therapy for noninvasive treatment of VT. Method: Sixteen patients, who underwent single-fraction VMAT-SBRT (3-5 arcs) to arrhythmogenic focus on a prospective phase I/II trial, were retrospectively re-planned using Varian's ProBeam-IMPT and compared against their respective clinical plans. Four non-coplanar-beams were selected by minimizing the irradiated volumes of the surrounding NT-HT, stomach, esophagus and bowel while maximizing plan robustness. A single-fraction dose of 35Gy to ITV and 25Gy to PTV were prescribed. Varian pencil-beam model in Eclipse Treatment-Planning-System was used to optimize the plans. Results: Typical isodose distributions and dose-volume-histograms are shown in Figure-1. Esophagus and/or stomach were adjacent to PTVs in 25% of patients, which limited the PTV coverage. However, all IMPT plans showed adequate target coverage (V95%Rx=>99%) while meeting the OAR objectives/constraints. Although conformity-index was similar, excellent NT-HT sparing and reduction in V50%Rx was observed in IMPT plans. NT-HT volume reduction at different dose-levels is shown in Table-1. Conclusion: The advantage of distal dose-fall-off by protons is highly promising for this group of patients where cardiac toxicity may play a vital role in long-term survivors. However, delivery accuracy in the presence of tissue-motion should be tested before clinical use. Furthermore, this may open doors to a new set of patients otherwise treated using catheter ablation. O 056: Plan Analysis Tools to Quantify Proton IMPT Complexity A. Gosling^1, A. Warry^1, C. Gillies^1, V. Rompokos^1, A. Poynter1, D. D'Souza^1 ^1University College London Hospitals, Radiotherapy Physics, London, United Kingdom QA tasks take up a significant amount of time for any proton therapy centre. The United Kingdom's National Health Service (NHS) model for a standard treatment day at the new proton therapy centres in London and Manchester allocate 12.5% of the total available treatment time to patient specific QA. However, there are currently no agreed upon metrics that measure the complexity of a proton plan to offer an indication of its likelihood of passing such QA. In preparation for the opening of the NHS proton therapy centre at University College London Hospitals (UCLH), we are developing a suite of tools to extract and analyse data from DICOM ION plan files. These tools will generate metrics to quantify the complexity of plans, and display these in easily interpreted plots; thus allowing planners to determine plan modulation levels and quantitatively compare plans designed using different techniques. We generate metrics and plots based on the MUs in beams and individual spots, their change over the course of beam delivery, as well as the derivatives of these values. Once we begin treatment operation, we plan to correlate these metrics with the outcomes of patient plan QA performed using the beamline. From this we hope to determine threshold levels at which we will be confident of a plans passing QA. In the long term this may allow us to reduce the amount of patient specific QA performed and therefore improve patient throughput and working time efficiency. O 058: The Development of a Unique Co-Linearity Set-Up and Test Phantom for a Patient Positioner Mounted Image Guidance System N. Schreuder^1, D. Hu^2, J. Shamblin^2, J. Treffert^2, D. Slater^2, P. Bagwell^2, L. Derenchuk^2 ^1Provision Cares Proton Therapy Center - Knoxville, Medical Physics, Knoxville, USA, ^2ProNova Solutions, ProNova, Knoxville, USA The ProNova SC360 system facilitates image guidance Radiotherapy (IGRT) with a cone-beam CT (CBCT) system mounted on the patient positioner. Classic IGRT systems acquire images independently from the patient positioner relative to the beam isocenter allowing a simple relationship between the beam and the IGRT system. A new method was required to test co-linearity between beam, PPS and IGRT systems for the SC360. A precision phantom containing four 2 mm diameter steel BBs rigidly mounted, with one at isocenter, and several precision laser tracker targets was developed. The isocenter BB is approximately aligned to the beam isocenter. The dose delivery system generates beam spots centered on the BBs. The BB shadows in the beam spots are clearly visible on EBT Gafchromic film or a scintillation detector placed behind the phantom (figure 1). The BB shadows relative to the center of the beam spots are digitally analyzed and the PPS adjusted accordingly to bring the isocenter BB precisely to the beam isocenter. The laser tracking system is used to determine the location of the isocenter BB in the room coordinate system. A CT scan of the phantom is used to create a treatment plan with beam spots centered on the BB's. The IGRT system generates a 3D CBCT image and uses automatic registration to generate all required corrections. The spot pattern is delivered and any deviations between the beam spot and the BBs indicates a calibration error between the PPS mounted IGRT system and beam and room coordinate systems. O 060: Utilizing Imaging Thresholds to Improve Setup Efficiency for Proton Therapy Patients S. Petro^1, B. Robison^1, J. Renegar^1, M. Blakey^1, M. Artz^1, N. Schreuder^1 ^1Provision CARES Proton Therapy Center, Medical Physics, Knoxville, TN, USA Proton therapy relies on image guidance to position the patient, due to protons'sensitivity to changes in the treatment pathway. Typically, therapists align to “zero”, i.e. exact alignment between the DRR and acquired X-ray. This is rarely achievable, leading to increased imaging time and dose and reduced throughput. To improve imaging efficiency, we developed thresholds to provide therapists an objective measure of sufficient patient alignment based on the patient's robust plan evaluation. We perform image guidance using pairs of pre-treatment orthogonal X-rays. Prior to using thresholds, each patient typically received 2-4 pairs of X-rays. To reduce imaging time and dose, we implemented thresholds. Each patient undergoes a robust evaluation by a physicist, who perturbs the plan with isocenter shifts. The perturbed shift is based on the patient's immobilization and imaging, e.g. boney anatomy or fiducials. Imaging thresholds are 2/3 of the perturbation. For example, a breast patient, immobilized with a VacLoc bag and aligned to external BBs, is shifted 5mm and rotated 3deg. The thresholds would then be 3mm in X, Y, and Z and 2deg of rotation. During pre-treatment imaging, therapists acquire an X-ray pair. If shifts are less than the thresholds, therapists are to move on to treatment without shifting or re-imaging. Since the introduction of imaging thresholds, each patient typically receives only 1-2 pairs of X-rays and therapy reports improved efficiency. Utilizing imaging thresholds based on each patient's individual setup has improved pre-treatment imaging efficiency and reduced imaging time and dose. O 061: Which Proton Imaging Setup Should We Use in a Proton Therapy Facility? N. Krah^1, F. Khellaf^1, J.M. Létang^2, S. Rit^1, I. Rinaldi^3 ^1CNRS, Creatis, Villeurbanne, France, ^2INSA Lyon, Creatis, Villeurbanne, France, ^3CNRS, Ipnl, Villeurbanne, France Proton imaging has been proposed to reduce range uncertainties in the treatment planning by circumventing or improving the conversion from Hounsfield Units to relative stopping power (RSP). In the last decades, different types of proton imaging setups have been developed, but not yet integrated into the clinical workflow. In this contribution, we present a comprehensive comparison of four types of proton imaging setups. For this purpose, we develop a mathematical framework to quantify the spatial resolution achievable with each of them. We find that the spatial resolution in setups combining pencil beam scanning with X-ray flat panel detectors is 10% better than in setups with pixel-less detectors such as multilayer ionisation chambers. In both setup types, performance can be significantly improved by reducing the pencil beam size down to 2-3 mm FWHM. In this case, the achievable spatial resolution is only 50% lower than in considerably more complex single proton tracking setups. Our results show that imaging setups combining double scattering with a pixel detector provide sufficient spatial resolution only under very stringent conditions. We further investigate the setups' integrability in a proton center and analyse their performance in the low dose regime based on characteristic detector sensitivities and image acquisition scenarios. Finally, we develop methods to extract physical properties of the imaged object complementary to the RSP and validate the results against experimental data. Our results indicate that detector hardware already available in proton centers for quality assurance provide well performing proton imaging setups. O 062: Estimation of the Risk for Sequelae Following Radiosurgery of Liver Metastases G. Mondlane^1, A. Ureba1, M. Gubanski^2, P. Lind^2,3, A. Siegbahn^4 ^1Stockholm University, Department of Physics - Medical Radiation Physics, Stockholm, Sweden, ^2Karolinska Institutet, Department of Oncology and Pathology, Stockholm, Sweden, ^3Södersjukhuset, Department of Oncology, Stockholm, Sweden, ^4Stockholm University, Department of Physics – Medical Radiation Physics, Stockholm, Sweden Purpose: The aim of this study was to estimate the risk of side-effects following photon- and proton-beam based radiosurgery of liver metastases. Materials and Methods: Ten liver metastases patients previously treated with photon-beam radiosurgery were selected. Intensity-modulated proton therapy (IMPT) plans were thereafter created by performing a CTV-based selective robust optimisation. Setup and range uncertainties were considered and a simultaneous PTV-based conventional optimisation was also performed. A robustness criterion was defined for the CTV (V95%>98% for at least 10 of the 12 simulated scenarios). The dosimetric values and the NTCPs obtained with the clinically used photon plans were compared with those obtained with IMPT. A generic proton RBE of 1.1 was assumed. Results: For all patients, the robustness criterion was respected. Similar dose coverage of the PTV was obtained with the photon and proton plans. An improved dosimetric sparing of the healthy part of the liver, right kidney, lungs, spinal cord and the skin was achieved with IMPT. However, similar NTCP values (median = 0 %) were obtained for all the OARs, except for the healthy part of the liver, for which lower NTCPs were obtained with IMPT (Figure 1: NTCP values obtained for the healthy part of the liver with the photon plans and the IMPT plans). Conclusions: Use of the selective robust optimisation approach in the proton radiosurgery planning carried out for liver metastases patients could improve the plan quality in terms of target-dose coverage, overall OAR sparing and reductions of the normal liver toxicity. O 063: Normal Tissue Complication Probability Models in Plan Evaluation of Children with Brain Tumours Referred to Proton Therapy C. Stokkevag^1, D.J. Indelicato^2, H. Magelssen^3, M.E. Evensen^3, M. Ugland^1, M. Brydoy^1, T. Nordberg^1, G.M. Engeseth^1, P. Brandal^3, L.P. Muren^4 ^1Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway, ^2University of Florida, Department of Radiation Oncology, Jacksonville, USA, ^3Oslo University Hospital, Department of Oncology, Oslo, Norway, ^4Aarhus University Hospital, Department of Medical Physics, Aarhus, Denmark Purpose: Children with brain tumours are at particular risk of radiation-induced morbidity and are therefore routinely considered for proton therapy (PT) aiming at reducing the dose to healthy tissues. The aim of this study was to apply normal tissue complication probability (NTCP) models derived specifically for children when evaluating the gain of PT compared to contemporary photon-based radiotherapy (VMAT). Methods: The study included sixteen patients (2-16 years) referred from two Norwegian institutions to PT abroad (2014-2016). All patients received cranial passive modulation PT, with CTV-dose prescriptions of 50.4-59.4 Gy(RBE). All cases were re-planned with VMAT to match target coverage of the delivered PT plan using original CT-images and structure sets. PT and VMAT plans were compared by relevant dose/volume metrics and NTCP for organs associated with growth hormone levels, auditory toxicity, visual impairment, and neurocognitive outcome. Results: Low-dose volumes were considerably reduced with the delivered PT plans compared to VMAT re-plans, while the high-dose volumes were comparable (Fig.1). For low-to-intermediate risk levels for VMAT, PT risk was close to baseline, while for individual organs at intermediate-to-high dose levels, some cases favoured VMAT (Fig.2). Conclusions: Most of the investigated parameters resulted in lower median values for the delivered PT plans compared to VMAT re-plans, with statistical significant differences for predicted audiometric toxicity. For normal tissue within and immediately adjacent to the target volume, NTCP estimates resulted in comparable elevated risk, independent of modality. This suggests that further toxicity reduction may depend on systematic de-escalation of the prescription dose in conjunction with technological advancement. O 064: Helical Tomotherapy Alone versus Helical Tomotherapy + Proton Beam Therapy Combination in Nasopharynx and Oropharynx Cancer: Comparison of Acute Toxicity S.G. Park^1,2, Y.C. Ahn^1, D. Oh^1, J.M. Noh^1, C.S. Hong^1, S.G. Ju^1, D. Kwon^1, K. Jo^1, E. Chung^1, W. Lee^1 ^1Samsung Medical Center- Sungkyunkwan University School of Medicine, Department of Radiation Oncology, Seoul, Korea Republic of, ^2Keimyung University Dongsan Medical Center- Keimyung University School of Medicine, Department of Radiation Oncology, Daegu, Korea Republic of In this contribution, we evaluate feasibility of combining helical Tomotherapy (HT) and proton beam therapy (PBT) in nasopharynx (NPC) and oropharynx (OPC) cancer. From January 2016 till October 2017, 158 patients (90 NPC, 68 OPC) received definitive radiation therapy (92.4% with concurrent chemotherapy). Using simultaneous integrated boost and adaptive re-plan, 68.4 Gy to GTV and 36.0∼60.0 Gy to CTV were delivered in 30 fractions: initial 18 fractions by HT in all, and, after rival plan evaluation, later 12 fractions by HT in 111 (70.3%) or by PBT in 47 (29.7%). Acute toxicities and analgesic usage were evaluated. Patients receiving HT/PBT presented more frequently with ipsilateral neck involvement and lower stage (p<0.001 and <0.001). With median 11 months' follow-up, locoregional and distant failures occurred in 7 (5.8%) and 12 (10.0%) without difference by technique or primary disease. Among all, grade ≥2 weight loss was less frequent in HT/PBT group (52.3% vs 27.7%, p=0.004). Among NPC patients, grade ≥2 mucositis was less frequent in HT/PBT group (69.4% vs 42.9%, p=0.017). Among OPC patients, grade ≥2 dermatitis was more frequent in HT/PBT group (18.4% vs 52.6%, p=0.005). Non-regular analgesic usage was more frequent in HT/PBT group (34.2% vs 48.9%, p=0.083), which was significant among NPC patients (45.2% vs 67.9%, p=0.046). With very high disease control rates, HT/PBT combination, when compared with HT alone, was advantageous with respects to grade ≥2 weight loss in all, grade ≥2 mucositis and analgesic usage in NPC patients, but was disadvantageous in grade ≥2 dermatitis in OPC patients. O 065: Cost Comparison of Hepatocellular Carcinoma Treatment with Either Proton Radiotherapy or Transarterial Chemoembolization as Assessed from a Randomized Clinical Trial D. Bush^1, J. Smith^2, M. Volk^3, M. Reeves^4, M. de Vera^5 ^1Loma Linda University Medical Center, Radiation Medicine, Loma Linda, USA, ^2Loma Linda University Medical Center, Radiology, Loma Linda, USA, ^3Loma Linda University Medical Center, Hepatology, Loma Linda, USA, ^4Loma Linda University Medical Center, Surgical Oncology, Loma Linda, USA, ^5Loma Linda University Medical Center, Transplant Institute, Loma Linda, USA Objective: To determine the costs associated with treatment and post-treatment care for hepatocellular carcinoma (HCC) patients treated with either proton beam therapy (PT) or transarterial chemoembolization (TACE) Methods: Eligible subjects had either clinical or pathologic diagnosis of HCC and met either Milan or San Francisco transplant criteria. Patients were randomly assigned to TACE or PT. Proton beam radiotherapy was delivered to all areas of gross disease to a total dose of 70.2 Gy in 15 daily fractions over 3 weeks. Treatment delivery costs of each treatment was determined by collecting billable CPT codes to calcutate medicare allowable charges. Post treatment costs were assessed by tabulating days of hospitalization within 30 days of treatment completion using medicare allowable charges associated with the admitting diagnostic code. Results: At the time of this analysis 69 subjects were available for analysis. 36 were randomized to TACE and 33 to proton. The 36 TACE patients received 63 courses of TACE as primary HCC treatment, treatment of persistent disease, or for new HCC in other parts of the liver. The 33 PT patients received 38 courses of PT. (treatment costs to be presented at meeting). Total days of hospitalization within 30 days of TACE/proton was 166 and 24 days respectively (p<0.001). Costs of hospitalization following treatment was significantly less with PT which when combined with treatment delivery costs, made PT a less costly method of treatment. Conclusion: This analysis indicates the overall cost PT is less than TACE potentially making it a more cost effective treatment for primary HCC. O 066: Stereotactic Proton Radiotherapy in the Treatment of Low and Intermediate Risk Prostate Cancer – 2-Year Results J. Kubes^1, S. Sláviková^1, P. Vítek^1, K. Dědečková^1, V. Vondráček^1, S. Vinakurau^1, B. Ondrová^1, A. Pasztorova^1, G. Kasáčová^1, J. Kvěch^1 ^1PTC Prague, Proton therapy dept., Prague, Czech Republic Purpose: Stereotactic radiotherapy of prostate cancer is a common modality in photon therapy. Pencil beam scanning (PBS) in similar fractionation allows better dose distribution and makes proton therapy more available for such patients. Material and methods: 200 patients with early stage prostate cancer were treated with IMPT (intensity modulated proton therapy), stereotactic schedule (36.25 GyE in 5 fractions) between February 2013 and December 2015. Mean age was 64,3 years, mean value of PSA before treatment was 6,83 μg/l (0,6-17,3 μg/l). 93 patients (46,5%) were in risk group 1, 107 patients (53,5%) were in risk group 2, 29 patients (14,5%) had neoadjuvant hormonal therapy, no patients had adjuvant hormonal therapy. Acute toxicity, early late toxicity and short term results were evaluated. Results: All patients finished radiotherapy without interruptions. Median of follow up time is 25,7 months. Mean treatment time was 9.5 days (median 9 days). Acute toxicity (CTCAE-v.4) was: GI G1-17%, G2-3,5%, GU G1-40%,G2-19%, no G3 toxicity was observed. Late toxicity was: GI G1-19%, G2-5,5%; GU G1-17%,G2-4%, no G3 toxicity was observed. PSA relapse was observed in one patient (1,03%) from risk group 1 (pelvic lymph node involvement was detected) and in 7 patients (6,5%) from risk group 2 (3 lymph node metastasis, 2 lymph node and bone metastasis, 2 PSA relapses). No patient died from prostate cancer, 3 patients died from other reasons. No local recurrence in prostate was observed. Conclusion: Proton beam radiotherapy of prostate cancer is feasible with low rate of acute toxicity and promising late toxicity and effectivity. O 067 - Prospective Patient-Reported Quality-of-Life Outcome of the Initial 100 Prostate Cancer Treated With the First Clinical Image-Guided Compact Pencil-Beam Proton Unit J. Wang^1, R. Guarisco^2, P.D. Dang^1, J.B. Wilkinson^3, S. Katz^1, M. Durci^1, H.T. Wu^1, L. Rosen^1 ^1Willis-Knighton Proton Center, Radiation Oncology, Shreveport, USA, ^2LSU-Shreveport, School of Medicine, Shreveport, USA, ^3Provision Proton Therapy Center, Radiation Oncology, Knoxville- TN, USA Introduction: Although proton therapy for prostate cancer has evolved, quality-of-life data for patients treated with modern techniques remains limited. Materials/Methods: Men with localized prostate cancer treated consecutively on a single-gantry compact pencil-beam scanning (PBS) proton unit (ProteusOne, IBA) between 2014 and 2017 were enrolled on an IRB-approved registry protocol. Patients' quality-of-life (QoL) questionnaires were prospectively collected at baseline and subsequent follow-ups. Patients without any follow-up or missing baseline assessment were excluded. Urinary symptoms and erectile dysfunction (ED) were assessed by the American Urological Association (AUA) Symptom Index and the Sexual Health Inventory For Men (SHIM) Score, respectively. Data was analyzed using two-tail Fisher's exact test comparing each time point to their baseline. Results: Of the total of 113 patients, median age was 68.4 (47.6-89.8), PSA 5.9 (0.012-382), and Gleason score 7 (6–9). Low, intermediate, and high risk patients were 15.0%, 57.4%, and 32.7%, respectively. 36% received androgen deprivation therapy. Median follow-up was 9 months. At baseline, 58.0%, 34.5%, and 7.6% had mild, moderate, and severe urinary symptoms. Most patients with initial moderate or severe urinary symptoms reported improvement at their last follow up (88% and 100%, respectively). Approximately half (52%) of patients reported being sexually active at baseline, with no changes in percentage of patients experiencing moderate-to-severe ED symptoms (15-36%, p=0.2872-1.000) at any given time point. Conclusion: Modern PBS proton radiotherapy possibly improves urinary symptoms and preserves sexual QoL in patients with localized prostate cancer. Long-term follow up is needed to confirm these favorable outcomes. O 068: Peripheral Lymphocyte Subpopulation Variation and Hematologic Changes after Carbon Ion Radiotherapy in Patients with Prostate Cancer Z. Yang^1, Z. Ning^1, Z. Qing^1, F. Shen^1 ^1Fudan University Shanghai Cancer Center FUSCC/Shanghai Proton and Heavy Ion Center SPHIC, Radiation Oncology Dept, Shanghai, China Purpose: We aimed to assess peripheral lymphocyte subpopulations variations and hematologic changes after carbon ion radiotherapy (CIR) in patients with prostate cancer for 3 years. Experimental Design: Eligible patients who had not undergone previous radiotherapy, were pathologically confirmed localized prostate adenocarcinoma. CIR was administered in daily fractions of 2.74GyE with a total dose of 63-66 GyE. Variations in lymphocyte subset counts were investigated pre-radiotherapy, during radiotherapy, post-radiotherapy, 1 month after CIR, and 3-year follow-up. Erythrocyte and platelet levels were evaluated pre-radiotherapy, during radiotherapy, post-radiotherapy, and in 6-month, 12-month, 24-month, and 36-month follow-up. CIR-related parameters were calculated using the Syngo system. Results: An increase in CD8+ was observed in 3-year follow-up(P<0.01) while CD8+ remained stable during radiotherapy; (P<0.05 for both). An decrease in CD4+ (P<0.01) and the CD4/CD8 ratio(P<0.01) was observed in 3-year follow-up while increase in the CD4+ and CD4/CD8 ratio during treatment during treatment and 1-month follow-up (P<0.05 for both). NK remained stable during radiotherapy and follow-up; CD19+ gradually decreased during radiotherapy (P<0.01) but increased in 1-month follow-up, then remained stable. Erythrocyte, leukocyte and platelet levels decreased during radiotherapy and in 3rd year (P<0.05), while they recovered to the pre-treatment level after completing CIR, then gradually increased during 2nd year of follow-up. Conclusion: Prostate CIR has a small but significant effect on the blood count and peripheral lymphocyte subpopulations in different ways. Endocrine therapy might be related to peripheral lymphocyte subsets variation after 3 years of CIR. O 069: Minimal Acute Toxicity in Major Salivary Gland Patients after Proton Beam Therapy: Outcomes from the Proton Collaborative Group REG001-09 Trial M. Chuong^1, W. Hartsell^2, H. Tsai^3, G. Larson^4, S. Badiyan^5, G. Laramore^6, L. Rosen^7, C. Vargas^8 ^1Miami Cancer Institute, Radiation Oncology, Miami, USA, ^2Northwestern Medicine Chicago Proton Center, Radiation Oncology, Warrenville, USA, ^3ProCure Proton Therapy Center - New Jersey, Radiation Oncology, Somerset, USA, 4ProCure Proton Therapy Center - Oklahoma City, Radiation Oncology, Oklahoma City, USA, ^5Maryland Proton Treatment Center, Radiation Oncology, Baltimore, USA, ^6University of Washington, Radiation Oncology, Seattle, USA, ^7Willis-Knighton Cancer Center, Radiation Oncology, Shreveport, USA, ^8Mayo Clinic, Radiation Oncology, Scottsdale, USA Background: Proton therapy (PBT) reduces normal organ dose compared to intensity modulated radiation therapy (IMRT) for major salivary gland patients. It is not well described whether this dosimetric advantage is clinically meaningful. Methods: We evaluated treatment parameters and acute toxicity outcomes of patients with major salivary gland cancers enrolled on the Proton Collaborative Group REG001-09 trial ([23]NCT01255748). Results: Of 717 head and neck trial patients, 105 were treated for parotid (N=90) or submandibular gland (N=15) tumors across 7 institutions from 2010-2017. Median age was 61 years. The most common histologies were mucoepidermoid and squamous cell carcinomas. Most were T1/T2 (54%), N0 (57%), and all were M0. Treatment prior to PBT included surgery alone (55%) or surgery followed by X-ray therapy (11%), most commonly 66 Gy in 33 fractions. Median PBT dose was 66.5 GyE (range 14.8-70.9) in 33 fractions (range 4-66); only 1 patient was prescribed less than 50 GyE. Uniform scanning (45%) was used more than pencil beam scanning (33%). Target volume details were not available. Chemotherapy was given concurrently with PBT to 20%. Median followup was 14.3 months (range 0.8-60.3). Acute grade 2 or higher nausea (1.5%), dysgeusia (4.8%), xerostomia (7.6%), mucositis (10.5%), and dysphagia (10.5%), were uncommon (Table 1). Conclusions: These are the first prospective data to demonstrate that major salivary gland patients who receive PBT experience less acute grade 2 or higher toxicity than is expected from IMRT. Longer follow up is needed to more accurately characterize late toxicity outcomes. O 070: Boron Neutron Capture Therapy for Locally Recurrent Head and Neck Squamous Cell Carcinoma: A Retrospective Analysis of Dose Response H. Koivunoro^1,2, L. Kankaanranta^1, A. Haapaniemi^3, A. Mäkitie^3, H. Joensuu^1 ^1University of Helsinki and Helsinki University Hospital, Department of Oncology, Helsinki, Finland, ^2Neutron Therapeutics Finland Oy, Medical Physics, Helsinki, Finland, ^3University of Helsinki and Helsinki University Hospital, Department of Otorhinolaryngology - Head and Neck Surgery, Helsinki, Finland Boron neutron capture therapy (BNCT) is based on accumulation of a boron carrier in cancerous tissues followed by external neutron irradiation. Boron fission due to neutron capture releases high-LET α-particles and lithium-nuclei, which have a range <10 μm, depositing most of the absorbed dose in cancer. Tissue boron uptake can only be estimated, and therefore the delivered radiation dose remains uncertain. In this study we examined the relationship between the estimated tumor dose and treatment response. Seventy-nine patients (49 men, 30 women; median age, 62 yrs), with inoperable, recurrent head and neck squamous cell carcinoma (rHNSCC) were treated with BNCT once or twice. Seventy-five patients (95%) received BNCT as salvage treatment after radiotherapy failure. Median minimum tumor dose from BNCT was 22 Gy(W) and tumor volume 105 cm3. Twenty-five (36%) patients had a complete response (CR), 22 (32%) partial response, 17 (25%) stable disease for a median of 4.2 months (range, 2.2-19.2), and 5 (7%) progressed. The median survival was 10 months (range, 0.1-124) and time to local progression 10.8 months (range, 0.1-124). A small median tumor volume (53 cm3) was associated with achieving CR. A minimum tumor dose >30 Gy (W) was associated with achieving CR, more favorable survival and a longer time to local progression. The tumor volume and the minimum tumor dose are prognostic factors for achieving CR and for survival in patients with rHNSCC. Accelerator-based neutron sources have recently became available for hospital installation, and will enable conducting of new prospective trials to confirm our findings. O 071: Comparison of Dose-Response of Acute Mucositis after C-ion RT for Patients Treated at NIRS and CNAO J.E. Dale^1,2, P. Fossati^3,4, A. Hasegawa^4, S. Molinelli^5, A. Mairani^5,6, O. Dahl^1,2, N. Matsufuji^7, T. Ohno^8, T. Kamada^9 ^1Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway, ^2University of Bergen, Faculty of Medicine, Bergen, Norway, ^3MedAustron, Carbon Ion Program, Wiener Neustadt, Austria, ^4National Center for Oncological Hadrontherapy, Clinical radiotherapy unit, Pavia, Italy, ^5National Center for Oncological Hadrontherapy, Medical physics unit, Pavia, Italy, ^6Heidelberg Ion-Beam Center, Medical physics, Heidelberg, Germany, ^7National Institute of Radiological Sciences, Department of Accelerator and Medical Physics, Chiba, Japan, ^8Gunma University Graduate School of Medicine, Department of Radiation Oncology, Gunma, Japan, ^9National Institute of Radiological Sciences, Research Center for Charged Particle Therapy, Chiba, Japan Objective: To investigate to which extent institution-specific treatment factors other than the applied RBE-model (delivery technique, no. of fields per day, etc.) affect the dose-response of acute mucositis in patients treated with C-ion RT at NIRS and CNAO. Methods: Palate was contoured as a surface structure of 5 mm thickness in 24 patients treated at CNAO and 27 patients treated at NIRS. For patients treated at CNAO, both the Local effect model (LEM) and NIRS clinical dose (NIRS clin) were applied to the physical dose by using the open-source TPS matRad. The Lyman-Kutcher-Burman (LKB) method was used to derive dose-response curves for CTCAE grade ≥ 3 palate mucositis for the groups: 1) NIRS patients with NIRS clin dose (NIRSNIRS clin), 2) CNAO patients with LEM dose (CNAOLEM) and 3) CNAO patients with NIRS clin dose (CNAONIRS clin). Results: Parameters for the LKB-model and the corresponding dose-response curves are presented in table 1 and figure 1. Changing only the RBE-model (CNAOLEM vs. CNAONIRS clin) resulted in a steepening of the dose-response curve and a 10% decrease in TD50. In contrast here was a more pronounced steepening of the curve and a 36% decrease in TD50 when comparing the toxicity taking all institution-specific factors into account (CNAOLEM vs. NIRSNIRS clin). Conclusion: Although this study is based in a small sample size, the results suggest that the different RBE-models only partially explain the observed difference in dose-response. These findings highlight the difficulties involved in comparing treatment outcome between different C-ion RT centers. O 072: Patient and Public Involvement in Design of a Phase III Trial Comparing IMPT versus IMRT in Low Risk Oropharyngeal Cancer C. Hague^1, B. Foran^2, E. Hall^3, R. Moule^4, C. Nutting^5, S. Parsons^6, R. Prestwich^7, N. Slevin^1, C. West^8, D. Thomson^1 ^1The Christie NHS Foundation Trust, Department of Head and Neck- Clinical Oncology, Manchester, United Kingdom, ^2Weston Park Hospital, Clinical Oncology, Sheffield, United Kingdom, ^3The Institute of Cancer Research, Clinical Trials and Statistics Unit, London, United Kingdom, ^4The University College London Hospital, Clinical Oncology, London, United Kingdom, ^5The Royal Marsden NHS Foundation Trust, Clinical Oncology, London, United Kingdom, ^6Manchester University NHS Foundation Trust and The University of Manchester, Public Programmes Team, Manchester, United Kingdom, ^7St James' University Hospital, Clinical Oncology, Leeds, United Kingdom, ^8The University of Manchester- The Christie NHS Foundation Trust-, Division of Cancer Science- Manchester Academic Health Science Centre, Manchester, United Kingdom Objective: The Christie Hospital, Manchester and University College London Hospital are developing the UK's high energy proton beam service due to open in August 2018. We report the results of public and patient involvement in the design of the first UK trial of Intensity Modulated Proton Therapy (IMPT) versus Intensity Modulated Radiotherapy (IMRT) for low risk oropharyngeal cancer. Material and Methods: Focus groups were held in Manchester, Sheffield and Leeds to understand patients' views of proton beam therapy, randomisation within the trial, willingness to travel and stay in Manchester for treatment, trial design and endpoints. Fifteen consecutive patients previously treated by IMRT for low risk oropharyngeal cancer from each centre were invited. Focus groups consisted of discussions around the questions shown in Table 1. Information was recorded on laminates, questionnaires, audio recordings and by telephone/email contact with a sample of patients. Data were analysed using thematic analysis. Results: Thirty three patients and eight relatives attended. Existing public knowledge of protons and the differences with IMRT was good. Beliefs about randomisation included disappointment if not receiving IMPT but willingness to participate to help future patients. Feedback on travelling and staying near the proton beam centre for treatment were generally positive, accepting the need for family and clinical support. The patient pathway and proposed trial outcomes were viewed favourably. Conclusion: This was an important piece of work to understand people's perceptions on the proton trial which has helped input into the trial design. The trial grant application is in process and due to be submitted in the near future. O 073: Commissioning of the ProNova SC360 PBS for First Patient Treatment at the Provision CARES Proton Therapy Center L. Derenchuk^1, J. Shamblin^1, R. Moore^1, J. Volk^1, H. Li^1, A. Xia^1, M. Heminway^1, P. Osucha^1, J. Matteo^2 ^1ProNova Solutions- LLC, Research and Development, Knoxville, USA, ^2ProNova Solutions- LLC, ProNova, Knoxville, USA The ProNova SC360 is a multi-room proton therapy system utilizing permanent and superconducting magnets to reduce cost and size. The SC360 fixed beam treatment room at the Provision CARES Proton Therapy Center in Knoxville has been calibrated to meet or exceed clinical performance of existing proton therapy systems. A novel and highly automated tool set has been developed incorporating a calibration phantom that links imaging, positioning, and beam isocenters. Daily QA tests verify the calibration as well as beam performance data. The SC360 span of ranges from 4 cm through 32 cm WET has been commissioned with spot position accuracy of +/- 0.5 mm throughout the full 25 cm x 25 cm scanning extent (Figure 1). Dose linearity is within +/- 1% down to 2.5 × 106 protons per spot (Figure 2a, 2b). Day-to-day reproducibility and inter-day dose reproducibility are demonstrated to be better than 0.5%. Scanning performance, positioning precision and positioning reproducibility are measured in daily QA and are demonstrated to be robust with respect to changes of the isochronous cyclotron performance (Figure 1c). Treatment delivery times for clinical prostate treatment plans are less than one minute. Daily automated data collection and processing can be completed efficiently and generates a summary report of data graphically displayed. O 074 - Implementation of Clinical Helium Beams at the Heidelberg Ion Therapy Centre T. Haberer^1, A. Mairani^2, J. Debus^3 ^1Heidelberg Ion Therapy Center, Directorate, Heidelberg, Germany, ^2Heidelberg Ion Therapy Centre, Biophysics, Heidelberg, Germany, ^3Heidelberg University Hospital, Radio-Oncology, Heidelberg, Germany About 4.500 patients were treated with Protons or Carbon ions at the Heidelberg Ion Therapy Centre (HIT) using the fully active 3D-dimensional intensity-controlled rasterscan technique [^1, 2]. An optimized linac-synchrotron combination generates libraries of energy-, focus- and intensity-variable pencil-beams for the dose-delivering scanning systems at two horizontally-fixed beam lines and a scanning ion gantry. The availability of low-LET and high-LET beams at HIT ranging from protons to oxygen under identical conditions optimally supports clinical trials aiming to clarify the question of which particle species is best suited for what indication. In order to pave the way for new therapy protocols HIT comprises a laboratory infrastructure and a dedicated research beam line offering low-LET proton and Helium beams and Carbon and Oxygen beams at the high-LET end. Comprehensive pre-clinical studies [^3, 4], i.e physical beam characterization, radiobiological experiments, biophysical modelling and treatment plan comparisons in realistic clinical settings, the evident potential of Helium ions for the use in particle therapy triggered an implementation project for this new modality at HIT. This paper will report about the approach and status of HIT's Helium treatment project. References: