ESTRO 36 Abstract Book

89 / 1082

S76 ESTRO 36 2017
_______________________________________________________________________________________________
Conclusion
The systematic sensitivity study revealed the capability of
the PGI slit camera to detect range shifts under clinical
conditions. In both treatment modalities, global range
shifts can be detected. Additionally, in PBS a spot-wise
comparison allows also the determination of
interfractional local range shifts. Moreover, a still ongoing
evaluation of PBS measured and simulated spot-wise
profiles for absolute range verification will be presented.
OC-0154 Proton therapy patient selection for
oropharyngeal cancer patients: the impact of treatment
accuracy
M. Hoogeman1, S. Breedveld1, M. De Jong2, E.
Astreinidou2, L. Tans1, F. Keskin-Cambay1, R. Bijman1, S.
Krol 2, S. Van de Water1, T. Arts1
1 Erasmus MC Cancer Institute, Radiation Oncology,
Rotterdam, The Netherlands
2 Leids University Medical Center, Radiation Oncology,
Leiden, The Netherlands
Purpose or Objective
Comparative treatment planning including Normal Tissue
Complication Probability (NTCP) evaluation has been
proposed to select patients for proton therapy. NTCP,
however, does not only depend on the type of radiation
used, but also on the size of the safety margins or degree
of robustness needed to account for treatment-related
uncertainties. In this study, for the first time to our
knowledge, the impact of margins and robustness settings
to the selection of oropharyngeal cancer patients is
investigated using fully automated comparative treatment
planning.
Material and Methods
CT and contour data of 78 consecutive oropharyngeal
patients were imported in our in-house developed system
for automated treatment planning for Intensity-Modulated
photon (IMRT) and proton radiotherapy (IMPT). Treatment
plans were generated fully automatically for a
simultaneously integrated boost scheme prescribing 70
Gy RBE to the primary tumor and pathological lymph nodes
and 54.25 GyRBE to the elective nodal areas in 35 fractions.
IMRT treatment plans were generated with a 0, 3, or 5mm
margin. IMPT 'minimax” robust optimized treatment plans
were generated for five different setup and range
robustness settings. Five validated NTCP models (see Fig.
1) proposed for IMPT patient selection were used in this
study. Following Dutch consensus guidelines, patients
were selected for IMPT if IMPT reduced NTCP by 10% or 5%
for a grade II or a grade III complication, respectively.
Results
In total 624 treatment plans were generated automatically
and approved by the authors. Figure 1 shows that the
percentage of patients selected for IMPT decreases with
increasing robustness setting for a given margin and also
decreases with decreasing margin for a given robustness
setting. In contrast to the size of the margin, the degree
of robustness has little impact on patient selection for
tube feeding dependence, which is the only grade III
complication. For the other complications the impact of
the degree of robustness setting is noticeably higher. For
patient-rated sticky saliva, nearly no patient is selected
for IMPT if robustness is included. If we consider high-
precision IMRT using a 3mm margin and high-precision
IMPT using a robustness setting of 3mm for setup and 3%
for range errors, most patients are selected for proton
therapy based on problems swallowing solid food (51.3%),
followed by tube feeding dependence (37.2%) and
decreased parotid flow (29.5%). Patient-rated sticky saliva
and patient-rated xerostomia contributed only with 1.3%
and 7.7% respectively.
Conclusion
This study shows that treatment accuracy cannot be
ignored in estimating the number of patients that will be
selected for proton therapy based on comparative
treatment planning and NTCP evaluation. We also
conclude that IMRT as well as IMPT should be optimized
for accuracy to ensure a sustainable use of proton therapy.
Proffered Papers: Imaging and image analysis
OC-0155 Automated lung tumour delineation in cine
MR images for image guided radiotherapy with an MR-
Linac
B. Eiben1, M.F. Fast2, M.J. Menten2, K. Bromma2, A.
Wetscherek2, D.J. Hawkes 1, J.R. McClelland1, U. Oelfke2
1University College London, Centre for Medical Image
Computing, London, United Kingdom
2The Institute of Cancer Research and The Royal Marsden
NHS Foundation Trust, Joint Department of Physics,
London, United Kingdom
Purpose or Objective
Respiratory-induced lung tumour movement is a
significant challenge for precise dose delivery during
radiotherapy. MR-Linac technology has the potential to
monitor tumour motion and deformation using
continuously acquired 2D cine MR images. In order to
target tumours in their current shape and position the
tumour outline must be established automatically. In this
study we compared four automatic contouring algorithms
that delineate the tumour in sequential cine MR images
based on manually contoured training images.
Material and Methods
Five 1 min 2D cine MR images (Fig. 1) were acquired for
two patients. Each sequence was split into a training set
of ten source images and a test set of about 100 images.
Method (1) is a multi-template matching, with a template
taken from each source image centred on the tumour. For
every test image the best position of each template is
evaluated and the most similar match is selected. Method
(2) uses a pulse-coupled neural network (PCNN) to improve
the grey-value contrast between tumour and healthy
tissue thus aiding the auto-contouring. The PCNN and
associated erosion and dilation parameters were trained
on the training sets using an accelerated particle swarm