ESTRO 38 Abstract book

S1097 ESTRO 38

W. Van den Wollenberg 1 , C. Carbaat 1 , P. De Ruiter 1 , P. Remeijer 1 , T. Janssen 1 , J. Sonke 1 1 Netherlands Cancer Institute, Radiation Oncology, Amsterdam, The Netherlands Purpose or Objective The MR-Linac (Elekta Unity, Elekta AB, Stockholm, Sweden) integrates a 7MV linac with a 1.5T MRI scanner. Currently, translational target misalignment can be corrected for using plan adaptations but rotational errors cannot. The purpose of this work was to develop and evaluate a method to correct for inter-fractional target rotation in prostate cancer patients. Material and Methods In this retrospective study, 4 prostate cancer patients with seminal vesicles (SV) invasion were included. Per patient, a planning CT scan, delineations of the prostate+SV (CTV), PTV (5mm margin), rectum, and anal sphincter, and a 9- beam step-and-shoot IMRT (20x3Gy) MR-linac treatment plan were available. The plans were created using the clinical version of Monaco 5.4. For each patient, delineations were rigidly transformed to simulate prostate rotations around the LR axis (-15⁰ to +15⁰ in 1⁰ steps). To avoid rotation into organs-at-risk (OARs) the rectum and anal sphincter were deformed by the same rotation that dampens with PTV distance (maximal within 5mm then drops off linearly over 1cm). Two adapt-to-shape (ATS) workflows available on the MR- linac were tested: 1. ATSseg starts with the segments of the original plan but uses the rotated structure set. It then tries to satisfy the planning constraints as best as possible. Per structure, electron density is set to the median density on the planning CT. . In addition we expanded the CTV into a dose evaluation volume (DEV) with increasing margins accounting for residual uncertainties and calculated the DEV D 99% . We compared these values to those obtained by the adapt-to-position (ATP) workflow, which does not correct for rotations. Results Distributions of DVH parameters are shown in figure 1. Figure 2 shows the CTV (plus margins) coverage as a function of rotation angle. Unlike the ATP method, the ATS methods maintained CTV coverage (D 99% > 95%) up to a margin of 3mm in all tested cases. CTV+4mm coverage was maintained in the large majority of cases. PTV coverage is lost incidentally for negative rotations, but often for positive ones (>5⁰); positive rotations wrap the rectum around the PTV hindering plan optimization. Median adaptation times were 4.0 min (ATP), 9.9 min (ATS seg ) and 10.9 min (ATS flu ). These were measured on a research computer and clinical times are expected to be faster. 2. ATSflu works as ATS seg , but starts from scratch with a fluence optimization. We evaluated the anal sphincter and rectum D mean , and the PTV D 98% and D 2%

Conclusion The ATP method fails to adequately correct for larger prostate rotations, while ATS does account for such rotations. ATS seg is faster than ATS flu , but ATS flu appears to be slightly better in terms of OAR sparing and target coverage. ATS is the preferred correction method for prostate rotations larger than a few degrees. The potential for PTV margin reduction following daily ATS is the subject of future studies. EP-2005 A novel method for rectal wall dose accumulation for prostate cancer patients F. Kamp 1 , H. Liu 1 , L. Ermoschkin 1,2 , M. Di Biase 1,3 , C. Kurz 1 , G. Landry 4 , C. Belka 1,5 , M. Li 1 1 University Hospital LMU Munich, Department of Radiation Oncology, München, Germany ; 2 HELIOS Hospital Krefeld, Department of Radiation Oncology, Krefeld, Germany ; 3 "G. D'Annunzio" University of Chieti- SS. Annunziata Hospital, Department of Radiation Oncology, Chieti, Italy ; 4 Faculty of Physics- Ludwig- Maximilians-Universität München, Department of Medical Physics, Munich, Germany ; 5 German Cancer Consortium DKTK, Munich, Munich, Germany Purpose or Objective Due to inter-fractional changes in rectum and bladder filling, precise dose estimation for the rectal wall over all fractions of a prostate treatment remains a major challenge. We developed a novel dose accumulation workflow for rectal wall dose, explicitly accounting for the daily changing geometry as well as physiological properties of this elastic organ, without the need for deformable image registration (DIR). Material and Methods In order to facilitate dose recalculations on the patient geometry of the day, the daily cone beam CT (CBCT) was intensity corrected using a previously published method. This method exploits the planning CT (pCT) as a prior to correct the projections yielding intensity corrected CBCT cor . All necessary structures are contoured on CBCT cor