643 / 1082
S627
ESTRO 36 2017
_______________________________________________________________________________________________
B. Barney
1
, G. Martinez
1
, R. Hecox
1
, J. Clark
1
1
Intermountain Health Care, Radiation Oncology, Provo,
USA
Purpose or Objective
A significant percentage of female post-lumpectomy
breast cancer patients treated with whole breast
radiotherapy (WBRT) have a lumpectomy cavity seroma on
the initial CT simulation. Our purpose was to prospectively
evaluate for changes in the size of the postoperative
tumor bed during a course of WBRT, prior to the
lumpectomy cavity boost (LCB).
Material and Methods
This prospective study was approved by the IRB, and
informed consent was given by 20 women prior to study
enrollment. All patients underwent breast conserving
surgery and received a recommendation for LCB following
WBRT by the treating physician. The median patient age
was 63 years (range, 41-84). Most patients (n=19, 95%) had
Stage 0, I, or II breast cancer. There was no standardized
dose or fractionation for WBRT or the boost; these
decisions were left to the treating physician’s discretion.
Regional lymph nodes were treated as per standardized
guidelines. When chemotherapy was required, it preceded
WBRT.
Each patient underwent initial CT simulation (CT1) at a
median 39 days (range, 11-216) from surgery. Twelve
women (60%) had a lumpectomy cavity seroma on CT1,
and 8 (40%) did not. All patients underwent a second CT
simulation (CT2) approximately 1 week before the LCB
began. Median time from CT1 and CT2 was 30 days (range,
21-42). The LCB volume was immediately contoured on
CT1 based on surgical clips, presence/location of seroma,
and surgical findings. Without referencing CT1 LCB
contours, the treating physician then contoured a
modified LCB volume once CT2 was obtained, using the
same factors for CT1 LCB delineation.
We prospectively compared LCB volumes from CT1 and
CT2 across the cohort and within seroma/no seroma
subgroups. Univariate analysis of several factors
potentially associated with a change in LCB volume from
CT1 to CT2, including time from surgery to CT1 (≤40 days
vs >40 days), time from CT1 to CT2 (≤30 days vs >30 days),
and presence of seroma on CT1, was performed.
Results
The median LCB volumes on CT1 and CT2 for the entire
cohort were 20.1 and 8.5 cm
3
, respectively. Most patients
(n=17, 85%) experienced a reduction (rather than
increase) in the LCB volume from CT1 to CT2. For patients
with seromas, median LCB volumes on CT1 and CT2 were
36.0 and 8.8cm
3
, respectively, representing a volume
reduction of >75% over the course of WBRT. For patients
without seromas, median LCB volumes on CT1 and CT2
were 11.8 and 8.0cm
3
, respectively, representing a
volume reduction of 32% during WBRT. On univariate
analysis, only the presence of seroma was associated with
a significant change in LCB volume during WBRT.
Conclusion
Most patients experienced a change in the size of the LCB
volume during WBRT. Patients with seroma experienced
a more dramatic volume reduction than those
without. We recommend that women who will undergo
LCB and have a seroma at the time of initial CT simulation
undergo a re-simulation to plan the LCB boost towards the
end
of
the
WBRT course.
EP-1164 Improved accuracy in IORT with electron
beams by a new measuring system of mammary gland
thickness
P. Scalchi
1
, A. Marchesin
2
, G. Scalco
2
, S. Bacchiddu
3
, C.
Mari
3
, L. Grandin
3
, P. Francescon
1
, F. De Marchi
2
, C.
Baiocchi
3
1
Ospedale San Bortolo, MEDICAL PHYSICS, Vicenza, Italy
2
Ospedale San Bortolo, SURGERY, Vicenza, Italy
3
Ospedale San Bortolo, RADIOTHERAPY, Vicenza, Italy
Purpose or Objective
In IORT of the breast cancer using electron beams (IOERT),
the beam energy should be properly chosen, as
recommended by both ICRU 71 (2004) and AAPM TG72
(2006), to ensure that the entire PTV be covered by the
90% of the maximum dose (D
max
) and the ICRU reference
point be positioned as near as possible to D
max
. Due to the
physical characteristics of these beams, the measurement
of the mammary gland thickness can be critical. In fact,
usually it is measured before docking using a needle and a
ruler ('needle method”), or ultrasounds. Nevertheless the
measured thickness can differ from the real one after
docking completion, thus affecting the accuracy of the
subsequent dose release. To allow accurate
measurements of the gland thickness under treatment
conditions, a new measurement system (MARK's) was
developed at Vicenza Hospital. The aim of this work is to
compare the needle method to MARK's in terms of surgeon-
surgeon variability and dosimetry impact.
Material and Methods
A mobile IOERT-dedicated linac (LIAC,SIT) with four
electron energies (4 to 10 MeV) is used at Vicenza
Hospital. MARK’s is a sterilizable manual pointer with
integrated ruler. After radioprotective disk positioning,
the surgeon stitches the mammary gland to prepare the
PTV. Then he inserts the terminal part of the applicator,
after applying a thin patch layer underneath to prevent
target herniation and, while keeping it pressed, he inserts
the pointer inside the applicator allowing direct thickness
measurements in treatment conditions.
14 patients were studied. The measurements were taken
first by the needle method, and then by MARK's. Five
measurements points were always taken, one at the
center of the PTV and four marginal positions (cranio-
caudal and lateral). The electron energies were chosen
based on the resulting thickness. The two systems were
compared in terms of both the choice of the electron
energy, as resulting by following ICRU and AAPM
recommendations, and the surgeon-surgeon variability.
Results
As shown in the following Table, the needle method
systematically overestimates the PTV thickness and
surgeon-surgeon reproducibility is better for MARK’s.
Following ICRU71 and AAPM TG72 the needle method
would cause 11 erroneous energy choices and 5 treatments
to be wrongly canceled.
N.
of
erroneous
energy
choices
(needle
method)
N.
of
possible
treatment
cancelations
following
ICRU 71 and
Surgeon-
surgeon
variability
(needle
method)
Surgeon-
surgeon
variability
(MARK's)
Thickness
difference
between
methods