ESTRO 2020 Abstract book

S948 ESTRO 2020

the CBCT. Original structures are deformed and transferred to the aCT. The dose is recalculated on the cCTs (cCT-plans) and on the aCTs (aCT-plans) based on the CBCTs that timely matches the cCTs. Calculations are performed with +/-3.5% range uncertainty. Clinical plans are robustly optimized with 0cm and +/-0.4cm isocenter shifts in addition to +/-3.5% range uncertainty. A coverage of V95%(CTV1) = 99% in the worst-case scenario is required for the clinical plan. The aCT-plans and the cCT-plans are compared with respect to changes in target coverage and dose to organs at risk (OARs). Results Calculations have been performed at each of the 5 weekly controls (C1-C5) for the first two patients with HNC who have completed their course of treatment, see fig. 1. Differences in CTV1 (66-68Gy) coverage with respect to clinical plans are shown in the top while dose to spinal cord and brainstem are shown in the middle and the bottom, respectively. Patient 1 (left) was re-scanned and -planned after C4 (top left). The aCT-plan did not immediately indicate a need for re-scan in this case with V95%(CTV1) = 99.5%. However, the cCT-plan showed that V95%(CTV1) = 98.5%. The 3D dose distributions (fig. 2), revealed the risk of target compromise in both the aCT-plan (top) and the cCT plan (bottom). The sub-optimal distribution was caused by a part of the immobilization device, marked with arrow in fig. 2, which led to re-planning. Critical OAR doses were similar for Patient 1 and the re-scan after C4 is clearly seen at C5. Patient 2 (right) was re-planned before C2 and re-scanned after C2. In this case both aCT- and cCT-plans indicated a need for re-planning at C2 (top right). The cCT-plan showed an increase in dose to spinal cord at C3 and brainstem at C4. However, the doses remained far below organ tolerances.

No differences are recorded for heart and contralateral lung, while some decrease was shown to occur in ipsilateral lung D mean and D 10Gy and in contralateral breast D mean (barely statistically significant). Conclusion Our analysis shows that breast treatment with HT is robust enough toward residual setup errors and daily small breast deformations. The adopted margins of 5 mm are adequate to guarantee the desired targets coverage, including nodal regions and tumour bed boost. PO-1637 Strategy for adaptive proton therapy for patients with head and neck cancer E. Samsøe 1,2 , K. Jensen 1 , C.R. Hansen 1,3* , P.S. Skyt 1 , J. Friborg 1,4 , B. Smulders 1,4 , I. Bahij 1 , A. Schouboe 1 , P. Randers 1 , A. Vestergaard 1 1 Danish Center for Particle Therapy, Aarhus University Hospital, Aarhus, Denmark ; 2 Copenhagen University Hospital Herlev, Radiotherapy Research Unit- Department of Oncology, Herlev, Denmark ; 3 Odense University Hospital, Laboratory of Radiation Physics, Odense, Denmark ; 4 Copenhagen University Hospital- Rigshospitalet, Department of Oncology, Copenhagen, Denmark Purpose or Objective Proton therapy of patients with head and neck cancer (HNC) was initiated at our institution in 2019. Since protons are sensitive to anatomical changes such as weight loss, fluid retention and tumor shrinkage, the proton treatment plan is recalculated on weekly control CTs (cCTs) to evaluate dosimetric consequences. Patients receive daily CBCTs to verify positioning. The hypothesis is that an artificial CT (aCT) generated from deformable registration of the planning CT (pCT) to the daily CBCT can replace some of the cCTs for recalculation of dose and evaluation of the need for adaptation of the treatment plan any time during the treatment course. Material and Methods An artificial CT is generated using deformable registration and resampling based on a daily CBCT and the pCT thereby creating an aCT that mimics the anatomical situation of