ESTRO 36 Abstract Book

252 / 1096

S239 ESTRO 36
_______________________________________________________________________________________________
relevant organs at risk (OAR) were pre-contoured. The
planning target volume (PTV) was derived at each centre
(5-10 mm). All plans were checked, PTV conformality (PTV
V95%/PTV Vtotal ) and PTV compromise (OAR V 95%/OAR&PTV
VTotal overlap) indexes were also calculated. The relevant OAR
for this case were the femoral head and neck (FHN) and
the femur in treatment field (FTF). The IMRT dose fall-off
gradient for FHN (FHN V95%/ FHN V80%) was also assessed.
Normal tissues and the joint were not analysed, as their
tolerances were easily met for this specific case.
Results
19 centres completed 20 IMRT plans. The plan quality of
9/20(45%) submissions was suboptimal and had to be
repeated. The results (see table) include the resubmitted
cases (total 29 plans). The case was particularly
challenging near FHN and FTF, due to an overlap of OARs
with the PTV. Depending on PTV margins, overlapping FHN
volumes varied from 13.8% (for PTV margins of 5mm) to
33.0% (for PTV margins of 10mm). FTF overlapping volume
with PTV ranged from 24.7% to 51.1%.Plans were very
conformal to PTV; however, the PTV conformality index
was not useful for areas where PTV overlapped with OAR.
We therefore calculated a compromise index for the PTV
areas overlapping with FHN and FTF, which support the
visual assessment of plans. The graph below highlights
plans in which V80% was suboptimal in relation to the V95%
(in total 5 plans had a suboptimal IMRT fall-off dose
gradient).
Conclusion
Limb STS tumours are a heterogeneous group of tumours
with significant variation in PTV shape and size. Evaluating
the plans for a newly implemented technique can be
challenging, particularly when determining if a plan is
optimal. We developed an objective assessment method
that is applicable to all limb STS. The first planning results
show that 45% of plans had either compromised PTV
coverage in favour of meeting OAR dose constraints, or
had not created a steep enough dose gradient near the
OAR. We attribute this to a change in the planning
technique paradigm, as many of the centres were using
IMRT for limb STS for the first time.
OC-0450 Geometric variation of the axillary lymph
node region in locoregional breast/chest wall
irradiation.
K.N. Goudschaal1 , N. Bijker 1, A. Bel1, N. Van Wieringen1,
M. Kamphuis 1
1Academic Medical Center, Radiotherapie, A msterdam,
The Netherlands
Purpose or Objective
Image-guided radiation therapy (IGRT) for p atients who
will be treated to both the breast/chest wall and the
axillary lymph node region (ALNR) is performed with
Conebeam-CT (CBCT). The position verification is based
on the breast/chest wall registration. The current
planning technique is a combination of a quarter field fast-
forward intensity modulated radiation therapy (IMRT) for
the breast/chest wall with an AP-PA beam technique for
the ALNR. This technique is robust for the daily position
variation of the ALNR but is not very conformal. We are
planning to introduce a volumetric arc therapy (VMAT)
technique which is highly conformal but the margins to
account for daily position variation are not known.
The aim of this study is to determine the daily positional
variation of the ALNR relative to the breast.
Material and Methods
The study population consisted of 20 female patients
treated with locoregional radiotherapy for stage II to IV
breast cancer. For all 20 patients the target volume was
the breast/chest wall and the ALNR level 1 and 2 or level
1 to 4, depending on the TNM classification. A standard
clinical target volume (CTV) to planning target volume
(PTV) margin of 5 mm is used for the entire axilla area and
the delineated breast is the PTV. The patient positioning
was supine with both arms up on a CQual breastboard
including a wedge position combined with a knee support
(CIVCO, USA). The clinical IGRT protocol with CBCT is
based on the position variation of the breast/chest wall.
A bony registration with a region of interest (ROI) on bony
anatomy (ribs and sternum) is used as a surrogate for the
breast position (XVI 4.5, Elekta).
For this study 138 CBCTs were retrospectively registered
on level 1 to 3 and 66 CBCTs on level 4. The CBCT analysis
was based on:
1. registration of the breast/chest wall using a ROI
(figure 1) on bony anatomy of the ribs and sternum;
2. registration of the ALNR using shaped ROI (SROI, figure
1) around level 1, 2, 3 and 4. The SROI is a ROI that can
be designed in the shape of each level separately. No
bony elements were included in these SROIs.
The geometric variation was expressed as the
displacement of the ALNR relative to the breast. The
mean, systematic and random setup errors of the
displacement of level 1 to 4 were calculated.
Results
The mean displacement for each level of the ALNR is small
(table 1). Considerable geometric variation was found for
Level 1 in ventral-dorsal (VD) direction. This may be due