ESTRO 36 Abstract Book

S557

ESTRO 36

_______________________________________________________________________________________________

Institute of Oncology, Department of Radiotherapy,

Warsaw, Poland

Purpose or Objective

On-board kilovoltage (kV) Cone Beam Computed

Tomography (CBCT) imaging is being used predominantly

for patients' positioning to improve the precision and

accuracy of treatment delivery. Moreover, volumetric

CBCT images can be used to evaluate anatomy changes

like tumour shrinking or weight loss as well as changes in

organ volume, deformation or position. Also

reproducibility of accessories used in radiotherapy (the

vacuum bag, bolus position, etc.) can be monitored.

Additionally, the CBCT data set can be used for calculation

of dose distribution. The aim of our study was to analyze

reproducibility of bolus’s position based on kV CBCT

imaging and compare planned and delivered dose

distribution in Clinical Target Volume (CTV).

Material and Methods

For 10 post-mastectomy patients, 35 sarcoma patients and

5 patients with vulva cancer the treatment CT based plans

with bolus were prepared (Eclipse, Varian). For the post-

mastectomy patients the planning CT was acquired with

bolus. For the two other groups the planning CT was made

without bolus. Bolus was drawn in the treatment planning

system. For each patient CBCTs were acquired in the first

and mid fraction and also at the end of treatment. CBCTs

were co-registered offline (automatic rigid match) to the

planning CT. The correctness of boluses positions were

evaluated. Also dose distributions were calculated with

CBCT images and compared with dose distributions

obtained with planning CT. For each patient we took a

photo to document the bolus’s position. To compare dose

distribution calculated on CT and CBCT, a new HU-

density

calibration curve was measured and introduced

into treatment planning system.

CatPhan®503 phantom

was used.

Results

Fusions of CTs and CBCTs showed that there are several

different problems with reproducibility of bolus position.

First of all, bolus generated in TPS will never adhere to

skin like it is presented in TPS, especially when a

irregularity of patient surface is high (Fig. 1a). Moreover,

usually air gaps occur even when there is a smooth surface

of the patients’ body (Fig. 1b). Another problem is the

compatibility of bolus position relative to a field edge (Fig.

1c) which is difficult to reproduce despite it is accurately

described in patient folder. The last two discrepancies

appear regardless if bolus was placed on patient’s skin

during CT scanning or generated in TPS. Preliminary

calculations for 10 soft tissue patients treated with 3D-

CRT plan show, that there is 5,9% ±2,0% a discrepancy

between D98% calculated on CT and CBCT. More advanced

data analysis will be presented for each location

separately, for treatment planning techniques and

information about taken/not taken photo before CT

scanning.

Conclusion

CBCT is a very useful method for accuracy of treatment

planning verification. It allows not only patient position

verification but also for evaluation of bolus positioning

accuracy. Based on CBCT several mistakes influencing

dose deposited to CTV were revealed, what was not

noticeable during a routine verification based on two-

dimensional orthogonal images.

PO-1011 Is it safe to omit a setup correction

validation scan for central lung lesions treated with

SBRT?

M.M.G. Rossi

, H.M.U. Peulen

, J.S.A. Belderbos

, J.J.

Sonke

Netherlands Cancer Institute Antoni van Leeuwenhoek

Hospital, Radiation Oncology, Amsterdam, The

Netherlands

Purpose or Objective

Our standard IGRT protocol for SBRT of pulmonary lesions

consists of a pre correction CBCT to determine the couch

shift that aligns the tumor, post correction (PCorr) scan

for verification and two intra-arc CBCT scans to monitor

position stability. The intra-arc scans are acquired

simultaneously during VMAT delivery. To limit the number

of scans for patients with centrally located lesions treated

with 8x7.5 Gy, the pCorr CBCT is omitted and the 1

intra-

arc CBCT is used as the verification scan. In this study we

evaluate the positional accuracy of this protocol for the

8x7.5Gy patient cohort and compare this to a 3x18Gy

cohort.

Material and Methods

All 16 patients treated since the implementation of a

setup protocol in April 2016 without a validation scan for

a single centrally located tumor with 8x7.5Gy(Gp1), were

included. Fifty patients were randomly selected from our