460 / 1096
S445
ESTRO 36
_______________________________________________________________________________________________
Material and Methods
10 HN oropharynx consecutive patients treated with PBS
PT underwent prospective evaluation computer
tomography scans (eCTs) during their course of treatment
(average 4 eCTs per patient). The robustness of the
treatment plans containing two-posterior oblique(PO) PBS
fields (2PBS), contingency VMAT plans (2 arcs) and
retrospectively generated 3-field PBS plans (3PBS) was
evaluated against anatomy changes and residual setup
uncertainties via evaluation plans generated on eCTs.
3PBS plan matched two PO fields (same to 2PBS) with an
anterior field at thyroid notch level in order to treat the
lower neck nodal target. Plan robustness was assessed
based on the accumulated dose through deformable image
and dose registration between treatment and evaluation
plans using VelocityAI. The D98% dosimetric indicator of
target coverage and OARs planning constraints were used
to evaluate the plan robustness. Changes over 5% in target
coverage, excessive cord dose and/or clinical decision
triggered proton replan or the use of the VMAT
contingency plan.
Results
The average change of D98% in the accumulated plans for
2PBS, 3PBS and VMAT were:4.1%±4.8%, -0.1%±0.8% and -
2%±3.2% for low risk CTV, -1.7±1.8%, -0.5±0.8% and -
0.73±1.2% for high risk CTV, -0.2%±0.2%, -0.1%±0.1% and
-0.4%±1.3% for gross CTV respectively. The main source of
coverage loss at low risk CTV level for 2PBS was found to
be due to variable soft tissue deformation of the posterior
neck for elderly or short neck patients leading to
replanning for 2 out10 patients. OARs robustness was
maintained within planning constraints.
Conclusion
2PBS plans were not consistently robust relative to target
coverage due to variable folding neck tissue, and
therefore it should be cautiously employed for elderly and
short neck patients. 3PBS was proved to be consistently
robust, similarly with VMAT.
PO-0828 Stereotactic body radiotherapy (SBRT) for
localised prostate cancer on the MR-Linac
A. Pathmanathan
1
, A. Mitchell
2
, K. Thomas
3
, D.
Henderson
2
, S. Nill
1
, U. Oelfke
1
, R. Huddart
1
, N. Van As
2
,
A. Tree
2
1
Institute of Cancer Research, Radiotherapy and Imaging,
London, United Kingdom
2
The Royal Marsden NHS Foundation Trust, Department
of Radiotherapy, London, United Kingdom
3
The Royal Marsden NHS Foundation Trust, Statistics
Unit, London, United Kingdom
Purpose or Objective
As the estimated alpha-beta ratio for prostate cancer is
low (1), moderate hypofractionation has been shown to be
isoeffective (2). The MR-Linac (MRL) combines an MR
scanner and linac allowing intrafractional tracking of the
target (3). However, dose distributions are affected by the
magnetic field (4). The first Elekta system MRL (1.5T/
7MV) will deliver step-and-shoot intensity modulated
radiotherapy (IMRT), a technique rarely used for
stereotactic body radiotherapy (SBRT). This planning study
assesses whether adequate dose distributions for MRL-
based prostate SBRT are possible with comparison to non-
MRL based planning techniques: IMRT, volumetric
modulated arc therapy (VMAT) and CyberKnife.
Material and Methods
Using planning CT scans acquired for ten patients with
localised prostate cancer, clinical target volume (CTV)
was defined as prostate plus proximal 1cm of seminal
vesicles. The planning target volume (PTV) was created by
addition of a 5mm isotropic margin, except 3mm
posteriorly. For the MRL, 5, 7 and 9-field step-and-shoot
IMRT plans were created to deliver 36.25Gy in 5 fractions
to the PTV with an integrated dose of 40Gy in 5 fractions
to the CTV using Monaco 5.19 (research version, Elekta AB,
Stockholm, Sweden). Non-MRL comparison plans included:
7-field 6MV IMRT for a conventional Elekta Agility linac
(Elekta AB, Stockholm, Sweden), 6MV FFF single 360° arc
VMAT using Pinnacle 9.10 (Philips Radiation Oncology
Systems, Fitchburg, WI) for a non-MRL and CyberKnife
treatment using Multiplan (Accuray inc, Sunnyvale, CA).
Plans were acceptable if the 16 dose constraints of the
PACE trial (NCT01584258) were achieved, without a major
variation to the protocol.
Results
Clinically acceptable 7-field IMRT MRL plans (see Figure 1)
were achieved in all ten patients. Clinically acceptable
plans were also achieved for all ten patients using 9-field
IMRT, non-MRL 7-field IMRT, non-MRL VMAT and
CyberKnife treatment. Clinically acceptable 5-field IMRT
MRL plans were only possible in seven patients. Table 1
summarises the number of exceeded constraints, mean
rectal doses and mean bladder V37Gy for each plan type.
Given the small patient group, exploratory ANOVA
analyses were undertaken for the number of co nstraints
missed, the rectum D1cc and the two most challenging
constraints to achieve- rectum V36Gy and bladder V37Gy.
For the MRL, 5-field IMRT MRL plans performed
significantly worse in all these analyses compared to 7-
field IMRT. 7-field IMRT MRL plans had significantly lower
rectal doses compared to CyberKnife plans. No significant
differences were seen between 9-field IMRT MRL plans and
non-MRL VMAT plans compared to 7-field IMRT.
Conclusion
MRL IMRT plans for prostate SBRT achieved the PACE trial
constraints in all patients with 9-field appearing similar to
7-field IMRT. 5-field IMRT in this set-up appears inferior
for the MRL. All platforms could produce clinically
acceptable plans. Further work is needed for dosimetric
validation and feasibility of MRL delivery.
PO-0829 Robustness of IMRT and VMAT for
interfraction motion in locoregional breast irradiation
R. Canters
1
, M. Kunze-Busch
1
, P. Van Kollenburg
1
, M.
Kusters
1
, P. Poortmans
1
, R. Monshouwer
1
1
Radboud University Medical Center, Radiation oncology,
Nijmegen, The Netherlands
Purpose or Objective
Conventional techniques for locoregional breast
irradiation using field abutment are challenging, even
more in combination with breath-hold irradiation and with
hypofractionation, since over- or underdosage may occur
more consistently in the abutment region in these
circumstances. IMRT and VMAT techniques are likely to