461 / 1096
S446
ESTRO 36
_______________________________________________________________________________________________
result in more conformal and homogenous irradiation,
though robustness for anatomical and posture variations is
possibly an issue. Compared to conventional plans, the
beams are not fully opposing and fields cannot be opened
manually outside the outer contour of the breast and the
body. Therefore, in this study we evaluated the robustness
of both an IMRT and a VMAT technique for daily variations
in patient positioning in comparison to our conventional
technique.
Material and Methods
20 Patients treated with a dose of 16x2.66 Gy using a
conventional technique to the breast and axillary lymph
nodes levels I to IV (Figure 1a) were replanned with both
an IMRT and a VMAT technique using Pinnacle
autoplanning. The IMRT technique consisted of 6 beams
with 20
o
spacing, while the VMAT technique consisted of
opposing pairs of 24
o
arcs (Figure 1). The delivered dose
was calculated using the cone beam CT (CBCT) (Elekta XVI)
images for each fraction to quantify the influence of
patient positioning, both for an online and offline
correction protocol. Contours were transferred from
planning CT to CBCT by deformable image registration
using Mirada RTx. Density overrides were applied to
account for imperfections in Hounsfield unit values on the
CBCT. IMRT and VMAT techniques were compared to the
conventional technique for the V95%, conformity index
(CI), mean lung dose and mean heart dose. The CTV-PTV
margin used is 7mm. Since the setup error is accounted for
when evaluating dose on the CBCT, we used the CTV for
the evaluation.
Results
Evaluation of the treatment plans for 20 patients showed
that V95% coverage of IMRT and VMAT plans was
comparable to conventional plans (Table 1). Conformity
was significantly higher for IMRT and VMAT. Mean lung
dose was approximately 0.7 Gy lower on average, while
mean heart dose increased by approximately 0.7 Gy using
IMRT or VMAT. Robustness evaluation of the dose on daily
CBCT’s using an online positioning protocol showed that
V95% coverage remained stable for conventional, IMRT an
VMAT. Significant conformity improvement was obtained
using both IMRT and VMAT, and small differences in mean
heart dose (+0.7 G) and mean lung dose (-0.8 Gy) were
found. Evaluation of an offline positioning protocol
showed similar results.
Conclusion
Presented IMRT and VMAT techniques show a similar
robustness for interfraction motion in locoregional breast
irradiation compared to the conventional technique, while
conformity of the target volume is increased significantly.
An offline positioning protocol would be sufficient for
clinically acceptable set-up accuracy.
PO-0830 Quantification of density and tissue changes
in pencil beam scanning proton treatment.
F. Van den Heuvel
1
, F. Fiorini
1
, B. George
1
1
University of Oxford, CRUK/MRC Oxford Institute for
Radiation Oncology, Oxford, United Kingdom
Purpose or Objective
Proton pencil beam scanning (PBS) is becoming the
methodology of choice to deliver proton therapy in many
cases. Several authors have reported discrepancies
between the dose distributions generated by commercial
planning systems, using analytical models, compared to
those using stochastic methods. The differences are
greatest in areas with extensive tissue inhomogeneities.
In analytically based commercial planning systems,
inhomogeneities are taken into account using a water
equivalent path length (WEPL) scaling. In this work we
quantify and investigate the impact of different densities
and tissue on the dose deposition characteristics of a
proton pencil beam.
Material and Methods
A single pencil beam with nominal energy 226 MeV from
an IBA-facility is modeled in homogenous cubic 40x40x40
cm3 phantom using FLUKA. The pencil beam’s dose
deposition is uniquely characterised using a stable