467 / 1082
S452
ESTRO 36 2017
_______________________________________________________________________________________________
PO-0844 Dosimetric Evaluation of MLC and Fixed Cone
for Patients in the Prone Position with CyberKnife
S.K. Ahn
1
, J.H. Cho
1
, K.C. Keum
1
1
Yonsei Cancer Center- Yonsei University, Department of
Radiation Oncology- Yonsei Cancer Center- Yonsei
University Health System- Seoul- Korea, Seoul, Korea
Republic of
Purpose or Objective
The constraints of systems using fixed cones have been
improved with the recent introduction of the multileaf
collimator (MLC) to the CyberKnife® system. This study
evaluated the dosimetric impact of the MLC in stereotactic
body radiation therapy for spine lesions, with the patient
in the prone position.
Material and Methods
Sixteen patients with spinal tumors, who were treated
with CyberKnife® M6
TM
, were placed in a body fixer and
scanned with four-dimensional computed tomography
(4DCT).
A total of 32 treatment plans were set up with two fixed
cones (ray tracing algorithm) and MLC (finite-sized pencil-
beam algorithm), using the MultiPlan® System. A total of
24 Gy in four fractions was prescribed to the 78%–83%
isodose line, encompassing at least 95% of the planning
target volume (PTV).
The XSight® prone tracking method was used for target
tracking, and the Synchrony® Respiratory Tracking System
was used for motion tracking. For the PTV, the maximum
dose, homogeneity index (HI), and conformity index (CI)
were analyzed. For the spinal cord and bowel, the
maximum dose (D 0.03 cc) was analyzed. The other
analyzed parameters included monitor unit, treatment
time, beam number, and node number.
Results
Regardless of the type of collimator, the difference among
the maximum dose, HI, and CI values of the PTV was
3.1±2%, while the maximum dose of the spinal cord and
bowel was 9.7±4.5%, indicating clinically insignificant
differences. For the other parameters, the values of the
treatment plan using MLC were lower by 53.8±8.4% for MU,
by 39.5±7.5% for treatment time, by 49.3±7.3% for beam
number, and by 49.7±7.1% for node number, compared to
the use of fixed cones. The differences were larger when
the tumors were greater in size.
Conclusion
There are dosimetric advantages to evaluating patients in
the prone position for lesions that are anatomically
located in the back, such as spinal tumors. However, MLC,
which has fewer treatment nodes and a shorter treatment
time, is also useful in the prone position because the
maintenance of positional reproducibility is critical.
PO-0845 Automatic treatment planning of FFF VMAT
for breast cancer: fast planning and fast treatment
E.L. Lorenzen
1
, K.L. Gottlieb
1
, C.R. Hansen
1
, H.R.
Jensen
1
, J.D. Jensen
2
, M.H. Nielsen
2
, M. Ewertz
2
1
Odense University Hospital, Laboratory of Radiation
Physics, Odense, Denmark
2
Odense University Hospital, Department of Oncology,
Odense, Denmark
Purpose or Objective
Forward planned tangential radiotherapy with wedges or
few segments is the standard technique in many centres
for radiotherapy after breast conserving surgery. Helical
techniques such as Thomotherapy and VMAT can be used
to increase conformity but may increase the volume
receiving low doses and the treatment planning can be
time-consuming. In the present study we evaluate FFF
VMAT using automated planning by comparison with
manually planned tangential radiotherapy on its plan
quality as well as its efficiency in both treatment planning
and delivery.
Material and Methods
Twenty patients, ten right-sided and ten left-sided, were
selected by including all patients receiving partial breast
radiotherapy between the 1/6-2016 and the 19/9-2016 at
our institution. All patients were treated with forward
planned tangential step-and-shoot 6MV fields and with
18MV fields used partly for larger breasts. The ten left-
sided patients were treated in breath hold using ABC from
Elekta. For each patient an additional plan was generated,
using two small (30-40 degrees) 6 MV FFF VMAT fields with
tangential like beam angels. Dose planning was done in
Pinnacle 9.10 and the Auto-Planning module was used for
generation of the VMAT plans. Mean doses to target
regions and organs and risk were compared using paired t-
tests.
Results
VMAT plans were generated fast with a median time for
complete plan generation by Auto-Plan of 10,5 min (range:
9 min – 12 min) with further adjustments needed for 7/20
patients (5 min -15 min additional time). Mean doses to
target regions and organs at risk are shown in the table.
The doses were similar from both plans except for the dose
to the ipsilateral lung being statistically significant lower
from the VMAT plans. Dose volume histograms for the
ipsilateral lung and the PTV are shown in figure a) and b)
respectively. As shown, the dose to ipsilateral lung was
lower for all dose levels in the VMAT plans even though
the coverage of the PTV was better. The measured
delivery time of all VMAT fields were 14,5 s (range: 10 s -
22 s). As a result all VMAT plans could potentially be
delivered within two breath holds (our threshold for
maximum breath hold duration is 25 s). In comparison the
median number of breath holds required for the ten left-
sided patients treated in breath hold in the forward
planned treatment was 4 (range: 2 – 7).