698-699 / 3581
N= 962
GTV=CTV (+/- 2cm SV)
No diet modification, laxatives, or rectal
enemas were used
3DCRT plan for 9 fractions (PTV 1cm
isotropic)
Prostate ITV based on planning CT +
4CBCT
Late G≥2 and G3 = 21% and 3%
No impact of RV on both late tox
and BC
Off-line adaptive RT and rectal distension (WBH)
03/01/13
Park S. et al., IJROBP 2012
The quartile values of RV, CSA, and SVP were used to subdivide
each risk group. The impact of these volumetric parameters on 5-year
biochemical control (BC, Phoenix definition) and chronic Grade
!
2
and 3 GI and GU toxicity were examined. Common Terminology
Criteria for Adverse Events V3.0 was used to score toxicity.
Statistics
Estimated likelihood of events was calculated using the Kaplan-
Meier test. The statistical significance of differences between
curves was calculated using the log-rank test. The association of
categorical variables within treatment groups was analyzed using
Fisher’s two-tailed exact test. Differences between two sample
means for continuous variables were analyzed using
t
-tests. A
two-tailed
p
value of 0.05 was considered to be statistically
significant for all tests. Statistical analyses were perfor ed using
SPSS Statistics version 17.0 (Chicago, IL).
Results
Patient, tumor, treatment, and rectal volumetric
characteristics
Group I patients comprised 41.6% of the cohort. Median ages
were 70.2 and 72.1 years for groups I and II, respectively (
p
<
0.01). Group I had a significantly lower mean PSA (5.9 vs. 12.1
ng/mL;
p
<
0.01) and Gleason score (6.1 vs. 7.1;
p
<
0.01) when
compared with group II. The majority of group I patients (77%)
were identified by elevated PSA (
"
T1c), whereas 44% of group II
were
!
T2a. IMRT was employed more frequently among group II
(47% vs. 33%,
p
<
0.01).
Table 1a
summarizes these observations.
For the entire cohort, the median (mean) follow-up was 5.5
years. The median prescription dose was 75.6 Gy (minimum dose
t cl-PTV), the median isocenter dose was 79.7 Gy, and no
statistical difference was noted between the two risk groups.
The median values for RV, CSA, and SVP on the planning CT
image were 82.8 cm
3
, 5.6 cm
2
, and 53.3 cm
3
, respectively. Offline
adaptive IGRT allowed for a mean reduction from the initial PTV
to cl-PTV of 74.4
#
30.2 cm
3
. Target margin from the planning
CTV to cl-PTV was very heterogeneous with a mean of 8.1
#
3.1
mm (calculated from the middle of the CTV in the anteroposterior
use of the planning CTV would have introduced a significant
systematic bias as determined from the CTV at the mean position.
The mean difference and the standard deviation from the planning
CTV center position to the ITV center position were 0.5
#
2.8 mm
in the anteroposterior, 0.1
#
2.1 mm in the superoinferior, and
$
0.1
#
0.7 mm in the right-left positions. Again, the systematic
bias was larger in the superior portion of the target and smaller in
the inferior portion. The volume of the rectal wall and bladder on
the initial planning CT was used for both initial planning and
adaptive modification. Therefore, their dose-volume histograms
Fig. 1.
Relationship between the initial planning clinical target
volume (CTV; red and light blue) and the ITV (purple) for 2
patients.
Volume 83
%
Number 3
%
2012
IGRT eliminates the bias of rectal diste sion in prostate cancer
949
(1)
. The 5-year BC for the 50 patients with CSA
>
12 cm
2
was
93.6%, as compared with 88.6% for those with CSA
!
12 cm
2
(
p
Z
0.468). The BC for SVP subgroups (
<
39.5, 39.5
e
53.3,
53.4
e
75.2,
>
75.2 mL) was 86.3%, 90.3%, 91.2%, and 88.6%,
respectively (
Fig. 4
;
p
Z
0.97). The lack of impact of rectal
distension on BC was observed in all subgroups stratified by RV,
CSA, and SVP.
Table 1b
is a summary of the above data.
Toxicity outcome
For the entire cohort, any chronic Grade
"
2 GI and GU toxicities
(maximum toxicity at any time point) were 21.2% and 15.5%,
respectively. The respective values for any chronic Grade
"
3 GI
and GU were 2.9% and 4.3%. No significant differences were
noted in chronic Grade
"
2 or 3 GI and GU toxicities when the
cohort was stratified by RV (
Tables 2a and 2b
).
This
the tr
adapt
volu
our c
a me
rates
by
differ
betw
adapt
cours
result
Fig. 3.
Biochemical control (BC) by cross-sectional area (CSA)
quartile (
<
4.4, 4.4
e
5.6, 5.7
e
7.9,
>
7.9 cm
2
).
Tab
RV
6
8
CS
4
5
SV
3
5
A
area
spec
pros
Patient #1
Patient #2