Purpose: To quantify the variability in prostate and seminal vesicle positi
on during a course of external beam radiotherapy, and to measure the propor
tion of target variability due to setup error.
Methods and Materials: Forty-four weekly planning computerized tomography (
CT) studies were obtained on six patients undergoing radiotherapy for prost
ate cancer, All patients were scanned in the radiotherapy treatment positio
n, supine with an empty bladder, with no immobilization device. All organs
were outlined on 3-mm-thick axial CT images. Anterior and lateral beam's ey
e view digitally reconstructed radiographs and multiplanar reformatted imag
es were generated, The position of the prostate and seminal vesicles relati
ve to the isocenter location as set that day was recorded for each CT study
. Target position relative to a bony landmark was measured to determine the
relative contribution of setup error to the target position variability,
Results: The seminal vesicle and prostate position variability was most sig
nificant in the anterior-posterior (AP) direction, followed by cranial-caud
al (CC) and mediolateral (ML) directions, Setup error contributed significa
ntly to the total target position variability. Rectal filling was associate
d with a trend to anterior movement of the prostate, whereas bladder fillin
g was not associated with any trends, Although most deviations from the tar
get position determined at the initial planning CT scan were within 10 mm,
deviations as large as 15 mm and 19 mm were seen in the prostate and semina
l vesicles respectively. Target position variations were evenly distributed
around the initial target position for some patient studies, but unpredict
able patterns were also seen. From a simulation based on the observed varia
bility in target position, the AP, CC, and ML planning target volume (PTV)
borders around the clinical target volume (CTV) required for target coverag
e with 95% certainty are 12.4 mm, 10.3 mm, and 5.6 mm respectively for the
prostate and 13.8 mm, 8.6 mm, and 3.9 mm respectively for the seminal vesic
les,
Conclusion: Target position variability is significant during prostate radi
otherapy, requiring large PTV borders around the CTV. This target position
variability may be potentially seduced by improving the setup accuracy, (C)
1998 Elsevier Science Inc.