75 / 1082
S62
ESTRO 36 2017
_______________________________________________________________________________________________
5. CTV for the vaginal vault and upper vagina (CTVv),
bladder and rectum were manually contoured on each
image set.
6. Dose was computed on each CBCT and then deformed
to the planning CT by a computed deformable registration.
7. For the purpose of summation of simulated dose,
because patients did not receive daily CBCTs, the
allocated dose for adjacent CBCTs where an interval of
more than one fraction occurred was calculated by
interpolation.
8. The planned dose to the CTVv, rectum and bladder was
compared to the simulated delivered dose using paired t-
test with Bonferroni correction.
9. The deformation vector field (DVF) for each voxel
within a 1mm internal annulus of the CTVv contour was
used to calculate mean and standard deviation (SD)
displacement in left-right (LR), anteroposterior (AP) and
superoinferior (SI) directions.
Results
A total of 169 CBCTs from 17 patients were analysed.
There were statistically significant differences in the
planned and delivered dose to the target CTVv (Table 1),
with clinically significant under dosing of CTVv D95% in 2
patients (4068cGy and 4135cGy, target 4275cGy). In these
patients there was substantial reduction in rectal volume
during treatment compared to the reference CT (mean
rectal volume relative to baseline 51.1% and 58.1%). This
resulted in posterior displacement of the CTVv (Fig 1). A
further 4 patients had clinically significant under dosing of
CTVv D50%. As a group there was no statistically significant
difference in delivered dose to the organs at risk (OARs),
but individually some patients showed marked
differences. Bladder and rectal volume varied during
treatment. The range of maximal % change was 9- 260%
for bladder and 32-249% for rectum. Grand mean ± SD
(range) (cm) for displacement of the DVF within the CTVv
annular structure were LR 0.04± 0.28 (-2.11-2.27), AP
0.19± 0.54 (-2.99- 2.92) and SI -0.15± 0.26 (-2.17-2.00).
Conclusion
Simulation of delivered dose using deformable registration
reveals significant differences in the planned and
delivered dose. Target and OAR motion may not be
accounted for with standard margins. Changes in rectal
volume can lead to significant under dosing to target and
AP margins of over 2cm may be required in some patients
as demonstrated by the DVF displacement.
PV-0133 Re-irradiation of pelvic recurrence of rectal
cancer: Developing an adaptive plan selection strategy
L. Nyvang
1
, C.S. Byskov
1
, M.G. Guren
2
, L.P. Muren
1
,
K.L.G. Spindler
3
1
Aarhus University Hospital, Dept. of Medical Physics,
Aarhus, Denmark
2
Oslo University Hospital, Dept. of Oncology and K.G.
Jebsen Colorectal Cancer Research Centre, Oslo, Norway
3
Aarhus University Hospital, Dept. og Oncology, Aarhus,
Denmark
Purpose or Objective
Radiotherapy (RT) of rectal cancer is challenged by
potentially large inter-fractional changes in internal
anatomy, of the tumour site as well as surrounding normal
tissues. Adaptive RT strategies have so far not been
applied clinically for rectal cancer. The aim of this study
was to develop an adaptive plan selection strategy based
on assessment of CBCTs in patients receiving RT for
recurrent rectal cancer, and to show its clinical feasibility
as well as its normal tissue sparing potential.
Material and Methods
Five patients previously treated with pre-operative
chemo-RT followed by surgery for rectal adenocarcinoma,
received pelvic re-irradiation according to a re-irradiation
protocol comprising - 40.8 Gy delivered in 34 fractions
with two fractions per day and concomitant capecitabine.
Daily CBCTs were acquired prior to each fraction with a
Varian Truebeam accelerator. A plan selection strategy
with a library of three plans was investigated, with the
target volume in Plans A, B and C covering the CTV with a
margin of 5 mm, 10 mm and 15 mm, respectively (Plan C
is close to the current non-adaptive plan). The CBCT of
each fraction was matched on the CTV of the planning CT
and analysed in order to determine which plan would
cover the CTV at the specific treatment fraction. The
‘effective PTV’ (PTVeff), a weighted mean of the volumes
treated throughout the treatment course using the plan
selections, was calculated for each patient in order to
quantify the potential reduction of the treated volume
compared to the standard non-adaptive PTV.
Results
Evaluations of all CBCTs were possible for all five patients.
For three patients, the CTV was included in Plan A in all
34 CBCTs. For one patient (patient 5), Plan A could have
been used in 25 fractions and Plan B in the remaining nine
fractions. One patient (patient 4) was more challenging
than the others due to a systematic change in the position
of the CTV resulting in Plan B to be chosen for all fractions.
In this case a re-scan and a new treatment plan would have
accounted for the systematic change. Overall, a
considerable potential for reduction of the treated
volumes is evident from table 1.