851 / 1082
S835
ESTRO 36 2017
_______________________________________________________________________________________________
In each case the CI is improved using the VMAT solution.
The average CI for SCF plans was 0.64, this increased to
0.81 when using VMAT. The GI appeared to increase in
several cases, with improvement only shown in 4 out of 10
cases. In each case, NTOF was reduced to close to zero.
Perhaps surprisingly for a VMAT solution, the 5Gy volume
was reduced for 9 out of 10 cases. The average 5Gy volume
for the SCF plans was 36.65cc; this reduced to 26.55cc
when using VMAT.
Conclusion
Paddick CI has proven to be a reliable metric for
conformity, while NTOF gives a good indication of how
baggy the prescription isodose is. However, Paddick GI was
not found to be a reliable indicator of low dose spread,
especially for small volumes. In many cases the 40% and
80% volumes had each been reduced and conformity
improved but a higher GI was recorded.
The Eclipse VMAT solution is adequate for a wide range of
tumour shapes, sizes and locations and can be used for all
single-target SRS patients.
EP-1570 The dynamic jaw mode of tomotherapy:
Better neural structure protection for advanced NPC
patients?
P.T. Fang
1
, T.Y. Lu
1
, M.N. Wu
2
, M.Y. Huang
1
, C.J. Huang
1
1
Kaohsiung Medical University Hospital, Radiation
Oncology, Kaohsiung, Taiwan
2
Kaohsiung Medical University Hospital, Neurology,
Kaohsiung, Taiwan
Purpose or Objective
This study investigated the neural structure protection
effects of the dynamic jaw mode of tomotherapy for
advanced nasopharyngeal cancer (NPC) by comparing use
of the dynamic jaw mode and the fixed jaw mode with
different field widths.
Material and Methods
Twenty patients with locally advanced NPC were selected.
All T classifications were T3-4. Plans were using the
simultaneous integrated boost technique in 3 dose levels.
Four plans were generated utilizing the 2.5 cm dynamic,
2.5 cm fixed, 5.0 cm dynamic, and 5.0 cm fixed jaw
modes. Plan efficiency was evaluated in terms of monitor
unit and beam-on-time measurements. Plan quality was
evaluated using homogeneity index (HI) and conformity
index (CI). Dose indices of neural structures such as the
optic pathways, temporal lobe, and hippocampus, as well
as of other organs at risk (OAR), were compared.
Volumetric parameters of the hippocampus and temporal
lobe were also assessed. Only plans of the same field width
were compared with one another.
Results
The comparison of the 2.5 cm dynamic jaw plan with the
2.5 cm fixed jaw plan indicated no differences between
plan efficiency, HI and CI results, or the dose indices of
OARs within the radiation field. However, use of the 2.5
cm dynamic jaw mode significantly improved the
maximum dose to the left optic nerve (p=0.005) and
chiasm (p=0.035). It also lowered the maximum dose and
mean dose to the hippocampus (both p values were 0.035).
In addition, the volume receiving 5 Gy (V5) values for the
temporal lobe and hippocampus were significantly smaller
when the dynamic jaw mode was used (p=0.035; p=0.013).
The comparison of the 5.0 cm dynamic jaw plan with the
5.0 cm fixed jaw plan also indicated no differences
between plan efficiency, HI and CI results, or the dose
indices of OARs within the radiation field. However, use of
the 5.0 cm dynamic jaw mode significantly improved the
maximum dose to the left and right optic nerve (p<0.0005)
and chiasm (p=0.005). The mean dose to the temporal lobe
was significantly improved using the dynamic jaw mode
(p=0.013). It also lowered the maximum dose and mean
dose to the hippocampus (both p=0.035). Finally, the
dynamic jaw mode also significantly reduced the V5, V10,
and V20 values for the temporal lobe and hippocampus.
Conclusion
When the same field width, the dynamic jaw mode
provided better neural structure protection than the fixed
jaw mode, in addition to reducing the low-dose volume to
the temporal lobe and hippocampus. These advantages
were more pronounced when using a larger field width.
There were no significant differences, however, in plan
quality and efficiency between the two modes for the
same field width. The above results indicate that for the
treatment of locally advanced NPC patients, use of the 2.5
cm dynamic jaw mode rather than the 2.5 cm fixed jaw
mode currently used in clinical practice would provide
better neural structure protection and lower low-dose
volumes to the temporal lobe and hippocampus with equal
plan
efficiency.
EP-1571 Radiotherapy treatments using a prototype
MLC design
P. Nitsch
1
, Y. Li
1
, T. Netherton
1
, P. Balter
1
, S. Gao
1
, M.
Muruganandham
1
, S. Shaitelman
2
, S. Frank
2
, S. Hahn
2
, A.
Klopp
2
, L. Court
1
1
The University of Texas MD Anderson Cancer Center,
Radiation Physics, Houston, USA
2
The University of Texas MD Anderson Cancer Center,
Radiation Oncology, Houston, USA
Purpose or Objective
We are evaluating the clinical efficacy of a prototype
multi-leaf collimator (MLC) design which obviates the
need for collimating jaws. The new MLCs are 1.0cm wide,
potentially giving increased reliability, and have a
maximum speed of 5.0cm/sec. The increased leaf width
may reduce the achievable intensity modulation, but the
impact of this may be mitigated by the increase in MLC
speed and by using IMRT/VMAT treatment planning. Here
we evaluate (1) whether clinically acceptable plans can be
created using such an MLC design, and (2) the agreement
between the planned and delivered dose distributions.
Material and Methods
IMRT, VMAT, field-in-field and electronic compensator
plans were created in the Eclipse treatment planning
system using the prototype MLC design and a flattening-
filter-free 6MV beam, for the following treatment sites:
head/neck, lung (standard fractionation, palliative and
SBRT plans), cervix (pelvis and extended fields), intact
breast (left and right), prostate (SBRT and involved
nodes), and whole brain treatments. The planned dose
distributions and DVHs reviewed for clinical acceptability
by radiation oncologists, and compared with our original
clinical plans (120leaf Millennium MLC, 0.5cm MLCs).
Delivered dose distributions for IMRT and VMAT plans were
evaluated using the ArcCHECK array.
Results
In most situations, the plan quality (particularly
homogeneity in the target) was highest for IMRT (with
multiple collimator angles), followed by VMAT, electronic
compensator, and field-in-field. The optimal IMRT and
VMAT plans were typically considered clinically
acceptable, while the electronic compensator and field-