907 / 1082
S891
ESTRO 36 2017
_______________________________________________________________________________________________
Four pts had a second adaptation during RT. For these pts,
changes above tolerance were solely observed for SS. The
median decrease in CTV V95% was 0.2% [0-2.1%] and PTV
V95% was 4.5% [0.6-7.2%]. None of the twice adapted
patients showed changes in anatomy which justified
reverting to the original treatment plan – either the
changes were further in the same direction or in a
different region.
Conclusion
Target coverage during the chemoRT in EC patients was
compromised in some cases due to interfractional
anatomical changes. Changes observed during RT
persisted and in some cases they increased, making
adaptation of the RT plan necessary.
EP-1662 Interfractional trend analysis of sinograms: a
decision-making for adaptive radiotherapy
S. Bresciani
1
, A. Garello
1
, A. Miranti
1
, A. Maggio
1
, A. Di
Dia
1
, P. Gabriele
2
, M. Stasi
1
1
Candiolo Cancer Institute - FPO- IRCCS, Medical Physics,
Candiolo TO, Italy
2
Candiolo Cancer Institute - FPO- IRCCS, Radiotherapy,
Candiolo TO, Italy
Purpose or Objective
The aim of this study is to investigate how geometric and
anatomical changes can be detected in daily sinogram
informations and how this information can be used to
examine interfractional trends, building toward a
methodology to optimize treatment and support adaptive
replanning.
Material and Methods
Sensitivity of detectors and sinograms complex to detect
shift errors and anatomical variations was previously
tested on thoracic phantom. In particular systematic
variations in shifts (1-5 mm in lateral direction),
anatomical variations (adding 1.25-2.5 cm bolus over
phantom) were applied.
Subsequently, a total amount of 106 patients treated with
Tomotherapy and their related 1573 sinograms were
analyzed. The sinograms, measured using Xenon
detectors integrated in Tomotherapy unit, were compared
with a reference one (usually the first fraction) using both
organ (CTV) and global percentage gamma pass (%GP)
evaluation (criteria: 3%/3mm, 2%/2mm and 1%/1mm). For
each patient mean %GP, standard deviation (σ) and
angular coefficient of linear fit of %GP where
evaluated. In particular σ is used to monitor random set-
up and preparation errors while the angular coefficient is
used to monitor the target size variation and tumor
response during treatment course. The results were
correlated to treated pathologies.
Results
The phantom results showed a sensitivity equal to 100% in
detecting all simulated errors.
The obtained results are described in the table
ANOVA analysis pointed out that the significance of the
difference between %GP and pathologies exists only when
calculating mean %GP,
s,
and angular coefficient with the
1%/1mm CTV criterion, obtaining respectively p<0.006,
p<0.04, and p<0.04. Applying the other criteria, the
obtained results were p>0.05.
The results showed that gynecological patients, followed
by lung, head and neck, and rectum pathologies are the
most responsive patients.
Conclusion
Based on these results, we can state a general correlation
law between angular coefficient of %GP and treated
pathology to search a quantitative parameter to help
predicte adaptive radiotherapy. This methodology could
provide an important element toward informed decision-
making for adaptive radiotherapy.
EP-1663 Automated full-online replanning of SBRT
lymph node oligometastases for the MR-linac
D. Winkel
1
, P. Kroon
1
, J. Hes
1
, G. Bol
1
, B. Raaymakers
1
, I.
Jürgenliemk-Schulz
1
1
UMC Utrecht, Department of Radiotherapy, Utrecht, The
Netherlands
Purpose or Objective
Diagnostic imaging on the MR-linac most probably provides
better visibility of lymph nodes compared to CBCT on
conventional linacs. While commercially available plan
adaptation methods are feasible, full-online replanning is
the preferred method to reach good plan quality. The aim
of this study is to investigate the feasibility of fast online
replanning on the MR-linac to account for inter-fraction
motion for stereotactic body radiotherapy (SBRT) of lymph
node oligometastases.
Material and Methods
Patient imaging data and delineations from seven
advanced cervix cancer patients with a combined total of
33 lymph nodes in the abdominal and pelvic region were
included. A planning simulation study was performed on
these lymph nodes with a 7-field IMRT technique and a
prescription dose of 5x 7Gy to 95% of the PTV. Treatment
plans were automatically generated using the research
version of Monaco by Elekta AB (Stockholm, Sweden) with
the use of their research automation API and in-house
developed automated treatment planning software. A
CTV-PTV margin of 3mm in all directions was applied. To
decrease optimization time an additional margin of 5mm
around the PTV was applied and only the parts of the
OAR’s within this margin were considered as OAR during
the optimization (Figure 1). All plans were generated using
the MR-linac machine model and a 1.5T magnetic field in
superior-inferior patient direction. Dosimetric outcomes
were evaluated against clinical dose constraints and
optimization time was measured. When required, PTV
coverage (V
100%
>95%) was sacrificed to meet all OAR dose
constraints.