Brachytherapy 14 (2015) 51e55

Adjuvant brachytherapy for endometrial cancer: Advantages

of the vaginal mold technique
Clement El Khoury1,*, Isabelle Dumas2, Anne Tailleur1, Philippe Morice3,
Christine Haie-Meder1
1
Department of Radiation Oncology, Brachytherapy Unit, Gustave Roussy Cancer Campus, Villejuif, France
2
Department of Medical Physics, Gustave Roussy Cancer Campus, Villejuif, France
3
Department of Surgery, Gynecology Unit, Gustave Roussy Cancer Campus, Villejuif, France

ABSTRACT PURPOSE: Treatment of endometrial carcinoma in the adjuvant setting includes in most cases
vaginal brachytherapy. In our institution, we use the customized vaginal mold technique. Herein,
we report the advantages of this personalized applicator in terms of target coverage, normal tissue
preservation, the incidence of air pockets, and its potential impact on dosimetry.
METHODS AND MATERIALS: A total of 15 patients receiving postoperative vaginal cuff high-
dose-rate brachytherapy with the mold applicator technique were enrolled in this prospective data
collection study. Patients were treated with either two or four fractions of 5 Gy prescribed to the
clinical target volume, which consisted of an irradiation of the vaginal cuff and the upper third
of the vagina. Target coverage; dose to organs at risk, in addition to the volume; and the dosimetric
impact of air pockets surrounding the mold were evaluated.
RESULTS: In 15 patients, a total of 27 air pockets were identified. The average number of air
pockets per patient was 1.8 (range, 0e4), with the average total air pocket volume being 0.1 cc
(range, 0.01e0.54). The average dose reduction at 5 mm from the air pocket was 26% (range,
6e45%). The minimal clinical target volume coverage reported was 95% and the maximal dose
received by 2 cc of the bladder, rectum, and sigmoid never exceeded 110% of the prescribed dose.
CONCLUSIONS: Vaginal cuff high-dose-rate brachytherapy using the molded applicator pro-
vides personalized tailored treatment in terms of anatomical conformity. This translates into a dosi-
metrical advantage with smaller and fewer air pockets than reported in the literature with the use of
cylinders. Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
Keywords: Endometrial cancer; HDR brachytherapy; Vaginal mold; Air pockets

Introduction deep myometrial invasion, tumor size, tumor extension to

the cervix or vagina, and lymph node involvement (3).
Vaginal brachytherapy has been reported as an effective
Worldwide, many different applicators have been used;
prophylactic treatment in the postoperative setting with
the most common was the single-line source vaginal cylin-
minimal side effects in the prevention of vaginal cuff
der (3, 4). With the application of three-dimensional imag-
relapses of uterine carcinomas (1, 2). Vaginal cuff brachy-
ing in gynecologic brachytherapy treatment planning,
therapy can be used alone or subsequent to external
defects in conformity of cylinders to the vaginal walls
beam radiotherapy (EBRT) depending on the patient’s path-
and cuff have been reported (5, 6). Although the largest cyl-
ologic risk factors for recurrence. These include patient
inder that can be fitted into the vagina was chosen to allow
age, high tumor grade, lymphovascular space invasion,
optimal apposition of the vaginal mucosa against the cylin-
der, air pockets forming could not be prevented. Recent
Received 17 June 2014; received in revised form 25 July 2014; studies have identified the presence of an average of one
accepted 31 July 2014. air pocket per patient (5), resulting in reduced radiation
Conflict of interest: None. dose to the vaginal mucosa (6). In addition, any improper
* Corresponding author. Department of Radiation Oncology, Brachy-
therapy Unit, Gustave Roussy Cancer Campus, 114, rue Edouard Vaillant,
placement of the applicator could result in target under-
94805 Villejuif, France. Tel.: þ33-1-4211-4566; fax: þ33-1-4211-5208. and/or overdosage to normal tissue. Thus, impairing local
E-mail address: clement.elkhoury@gustaveroussy.fr (C. El Khoury). control and increasing complications.
1538-4721/$ - see front matter Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.brachy.2014.07.007
52 C. El Khoury et al. / Brachytherapy 14 (2015) 51e55
At our institution, the endocavitary brachytherapy tech-
nique is based on the molded applicator and has been used
widely in clinical practice. It is reasonable to suppose that
an applicator adapted to vaginal anatomy would be more
conformal. The present report constitutes a brief descrip-
tion of this technique with a stress on the presence of air
pockets, their effect on dosimetry, and the potential place
of molded applicators in the modern era of brachytherapy.
Methods and materials
The study population was based on 15 consecutive pa-
tients with pathologically proven carcinoma of the endome-
trium who received adjuvant high-dose-rate (HDR)
brachytherapy to the vaginal vault, either alone or subse-
quent to EBRT between May and November 2013. The
treatments studied were the first fractions only, where CT
scanner was available. As previously reported, the intraca-
vitary brachytherapy technique at our institution is based on
the vaginal mold technique, which has been already pub-
lished (7). Briefly, the first step consisted of a vaginal size
assessment using a low-dust alginate impression, then
choosing between three sizes of preset acrylic molds based
on the size of the vaginal impression. In most of the cases
where the vagina had a round shape, three catheters were
inserted in each mold, two laterally and one in posterior po-
sition. The purpose of these three catheters was to achieve a
more homogeneous dose distribution. Rarely and in case of
a smaller or flatter vagina shape, only two catheters were
designed laterally to minimize mucosal hot spots and avoid
higher dose to the bladder and rectum. Holes were made in
the walls of the applicator to prevent mold displacement by
inducing mucosal herniation (Fig. 1). The range of molds
available in use varied in width and thickness: small
Fig. 1. Molded applicator for vaginal high-dose-rate brachytherapy. Lateral catheters and posterior catheter positioning are represented.
(30  25 mm), medium (40  35 mm), and large
(50  40 mm), respectively. All were 55-mm long. The
mold was chosen to be the closest in size to the vaginal
impression. A careful clinical examination was systemati-
cally conducted before this procedure to ensure healing of
the vaginal vault and assessing for any residual tumor.
The mold was then inserted into the vagina. In-room
orthogonal radiographic localization images (anteroposte-
rior and lateral view) were taken after each application to
ensure that the mold was inserted in the same orientation.
This was achieved by evaluating the position of radio-
opaque markers placed into the catheters and on the surface
of the applicator in relation to pelvic bony structures. To
have similar bladder filling between patients, a Foley cath-
eter was inserted with 7 cc balloon inflation of diluted
contrast dye and the bladder was completely drained during
CT and treatment. The CT images through the pelvis were
obtained without intravenous contrast injection at 3-mm in-
tervals and transferred to the treatment planning system
(TPS). The mold, bladder, rectum, and sigmoid were con-
toured. Primary clinical target volume (CTV1) was defined
as 5 mm from the surface of the vaginal mold in its cranial
3-cm portion targeting the upper third of the vagina (4). The
CTV2 was defined as 5 mm from the surface of the vaginal
mucosa and thus taking into consideration the volume of
the air pockets. The CTV1-based three-dimensional plan-
ning using doseevolume parameters was performed. The
dose of 5 Gy per fraction was prescribed to the CTV1. Each
treatment plan was evaluated for the presence of air pockets
in the cranial 3 cm of the vagina. The air pockets were con-
toured in the axial plane and the corresponding volumes
were calculated by the TPS (Fig. 2). Anisotropic loading
of the three catheters was required to optimize target
coverage and protect organs at close vicinity of the vaginal
walls or any overlying small bowel (8). Relative dwell
 C. El Khoury et al. / Brachytherapy 14 (2015) 51e55
Fig. 2. Axial CT image showing the mold inserted into the vagina. Arrows
represent air pockets displacing the mucosa from the applicator’s surface.
weights were applied differently between catheters, but re-
mained identical within each catheter with an average of
five activated positions in each (Fig. 3).
Dose was calculated by Oncentra Brachy module version
4.1 (Nucletron, Elekta) using a point source approximation
of an iridium source with parameters as specified in the
report of Task Group 43 of the American Association of
Physicist in Medicine (AAPM) (9). The doseevolume
histograms were generated for the CTV1, CTV2, rectum,
bladder, and sigmoid. The maximal dose received by
2 cc of tissue (D2 cc) was also evaluated. The maximum
displacement of vaginal mucosa in addition to the relative
percent dose at the point of maximum displacement and at
5 mm from the surface of the mold were recorded for all
air pockets. This step was done by a single experienced
investigator to minimize interobserver variation.
Fig. 3. Three-dimensional representation of clinical target volume (CTV) and organs at risk. Left image: Catheter dwell loading until base of CTV. Right
image: CTV coverage by prescription isodose.
53
Treatment consisted of either two or four fractions de-
pending on if the patient received pelvic EBRT or not.
Brachytherapy was delivered one fraction weekly using
HDR 125I sources with afterloading carriers.
Followup was scheduled at 6 weeks, 3 and 6 months af-
ter completion of brachytherapy. Morbidity scoring was
performed using Common Terminology Criteria for
Adverse Events, version 4.03. Appropriate institutional
and ethics committee approval has been obtained before
study initiation.
Results
A total of 15 consecutive patients received HDR brachy-
therapy to the vaginal cuff and the upper third of the vagina
for endometrial cancer and had CT scan before the first
fraction of their treatment as by the department’s policy.
Except in one case, patients were treated for endometrioid
adenocarcinoma and 10 patients received EBRT before
HDR brachytherapy. The tumors were staged according to
the International Federation of Gynecology and Obstetrics
staging system. Table 1 summarizes the patient characteris-
tics. The median number of air pockets between the vaginal
wall and the brachytherapy mold was 1.8 (range, 0e4 air
pockets). In the 15 patients group, there were a total of
27 air pockets with only 1 patient who did not have any.
The mean average volume of air pockets was 0.09 cc
(range, 0.01e0.54 cc) affecting an average of 0.22% of
the CTV volume. About 78% of the air pockets had a vol-
ume less than 0.01 cc and only 2 of the 27 were located on
the apex of the applicator. The average distance the vaginal
mucosa was raised off the surface of the mold was 3.2 mm
(range, 1.6e7 mm). Air pocket characteristics are listed in
Table 2.
The maximal applicator/vaginal surface point dose
ranged between 145% and 163% of the prescribed dose.
54 C. El Khoury et al. / Brachytherapy 14 (2015) 51e55

Table 1 Table 3
Patients characteristics Dosimetry features
Characteristics Number of patients (%) Parameters Mean (range), %
Age, yr, mean (range) 60 (47e79) CTV1 96.06 (95.1e98.7)
FIGO clinical stage CTV2 95.6 (95.1e97.0)
IA 2 (13) D2 cc Bladder 91.6 (48.5e109.8)
IB 8 (53) D2 cc Rectum 100.0 (78.5e109.7)
II 1 (7) D2 cc Sigmoid 57.4 (19.7e97.6)
IIIC 4 (27)
CTV 5 clinical target volume; D2 cc 5 maximal dose received by 2 cc
Tumor histology of tissue.
Endometrioid adenocarcinoma 14 (93)
Papillary/clear cell carcinoma 1 (7)
Tumor grade
I 3 (20)
hysterectomy. Variations in size and shape are unlimited
II 6 (40) and can vary from larger vaginal apex in relation to introi-
III 6 (40) tus, to lateral vaginal expansions in the vault because of
Surgical lymphadenectomy surgical remnants of the fornices and termed ‘‘dog-ear.’’
Yes 10 (67) One of the frequently encountered issues is the vaginal in-
No 5 (33)
Lymphovascular space invasion
troitus being smaller in diameter than the apex. Conse-
Present 5 (33) quently, in cases of numerous air pockets with a cylinder,
Absent 10 (67) the insertion of larger one to enhance conformance with
Surgical margin status the mucosa is limited by the patient’s tolerance. This prob-
Positive 1 (7) lem was not encountered in our study; the molded appli-
Negative 14 (93)
External pelvic irradiation
cator having only the three catheters in its caudal part
Yes 10 (67) passed through the introitus (Fig. 1). The use of ovoids
No 5 (33) has been proposed to address the concern of treating irreg-
FIGO 5 International Federation of Gynecology and Obstetrics. ular vaginal vaults. However, another advantage of the
molded applicator is its proximal bulging shape that ex-
The average dose reduction to the vaginal mucosa overlying pands the vaginal walls, flattens any ‘‘dog-eared’’ configu-
the air pockets was 25.6% (range, 6e45.5%). Bladder, ration, and displaces critical organs substituting the need
rectum, and sigmoid D2 cc were 91.6%, 100%, and 57.4% for packing used with ovoids. Moreover, the holes made
of the prescribed dose, respectively. By doseevolume in the mold’s wall (Fig. 1) allow for mucosal herniation,
histogram analysis, 100% of the prescribed dose covered thus reducing the formation of large air pockets and mini-
an average of 96% of the CTV2 corresponding to the targeted mizing the risk of applicator displacement excluding the
vaginal cuff and proximal upper third of the vaginal mucosa. need of an external immobilizing device. In addition, the
When all air pockets and the effect they had on the target geometric design of the three catheters allowed adequate
coverage were considered, the median average dose to the CTV coverage without excessive hotspots on the applicator
CTV2 was 95.6% (range, 95.1e97.0). When compared with surface with a maximum point dose documented at 163%
the dose the CTV1 has received, the CTV2 was at 99.52% of the prescribed dose, and no mucosal surface was found
(Table 3). to be exposed to the 200% isodose.
Acute tolerance was recorded for all patients. Overall, Our results in terms of air pockets characteristics
the treatment was very well tolerated with only one case (Table 2) were at least as attractive as data available in
of Grade 1 cystitis that resolved by the time of review at the literature using HDR application with cylinders. Ri-
3 months. None of the patients complained of bowel chardson et al. (6) reported an average number of air
symptoms. pockets per patient of 3.6 with a mean volume of
0.34 cc (range, 0.01e1.32 cc) and an average distance
displacement of the mucosa of 3.7 mm (1.3e8 mm). In
Discussion our study, the molded applicator showed promising results
in terms of air pocket per patient (1.8), mean volume
Vaginal cuff shape is unique for every patient and is (0.09 cc), and even mucosal displacement (3.2 mm). In
highly dependent on surgical technique of closure after the same report, Richardson et al. (6) concluded that the
existence of air pockets reduced radiation dose to the
Table 2 vaginal mucosa by 27%, which is in accordance with
Air pockets summary
our data. However, the dosimetric analysis in our sample
Parameters Mean (range) showed no impact on the overall target coverage, with
Number of air pockets per patient 1.8 (0e4) the air pockets affecting 0.22% of the targeted volume
Displacement of vaginal mucosa (mm) 3.2 (1.6e7) and compromising the prescribed dose by 0.48%. This
Air pocket dose reduction (%) 25.6 (6e45.5)
was confirmed earlier in a study by Cameron et al. (5)
 C. El Khoury et al. / Brachytherapy 14 (2015) 51e55
who stated that having air pockets occupying less than 1%
of the targeted volume had low chances of having micro-
scopic disease in the area of the mucosa raised from the
surface of the applicator. Consequently, the authors
considered the presence of air pockets around cylinders
not clinically significant.
Our data showed that doses to critical structures were
acceptable in terms of toxicity: D2 cc of the rectum, bladder,
and sigmoid never exceeded 110% of the prescribed dose
per fraction. Demanes et al. (10) showed that single-
channel cylinders deliver higher doses to bladder and
rectum for a given vaginal dose compared with ovoids.
The authors advocated the use of multichannel cylinders
to optimize target coverage and to spare sensitive organs.
Herein, we suggest another technique more adapted to the
patients anatomy to address these concerns.
Moreover, scanning was performed for the first brachy-
therapy fraction only, although most patients received two
to four fractions. Yaparpalvi et al. (11) documented that in-
terfraction changes in the cylinder insertion with internal
organ motion and volume variation resulted in a compro-
mised dose distribution, therefore the author recommended
imaging for each fraction. Consequently, in our study,
assurance on accurate geometric orientation was performed
by radiographic imaging before each treatment session to
minimize dosimetric variability on rectal/bladder dose.
Limitations to our study included small sample size,
mold size approximation using three different models only,
and large CT slices intervals of 3 mm compared with the
mean average mucosal displacement of 3.2 mm. Shorter
CT imaging intervals might be more relevant in terms of
providing more detailed information about smaller air
pockets.
The molded applicator represents an example of person-
alized tailored treatment in the modern era of brachyther-
apy. It adapts at best to the patient’s anatomy and is very
well tolerated by the patient. This technique reduces
air pocket’s volume, which translates into satisfactory
dosimetric coverage of the vaginal mucosa and highly
55
acceptable doses to the bladder and rectum. As the rate
of local relapse in the upper third of the vagina is extremely
low after EBRT and/or brachytherapy, it is unlikely that any
prospective trial will show a difference between the
techniques.
References
[1] Keys HM, Roberts JA, Brunetto VL, et al. A phase III trial of surgery
with or without adjunctive external pelvic radiation therapy in inter-
mediate risk endometrial adenocarcinoma: A Gynecologic Oncology
Group study. Gynecol Oncol 2004;92:744e751.
[2] Nout RA, Smit V, Putter H, et al. Vaginal brachytherapy versus pelvic
external beam radiotherapy for patients with endometrial cancer of
high-intermediate risk (PORTEC-2): An open-label, non-inferiority,
randomised trial. Lancet 2010;375:816e823.
[3] Small WJ, Beriwal S, Demanes DJ, et al. American Brachytherapy
Society consensus guidelines for adjuvant vaginal cuff brachytherapy
after hysterectomy. Brachytherapy 2012;11:58e67.
[4] Potter R, Gerbaulet A, Haie-Meder C, et al. The GEC ESTRO hand-
book of brachytherapy. Endometrial cancer. Leuven, Belgium: ES-
TRO; 2002. p. 365e401.
[5] Cameron AL, Cornes P, Al-Booz H. Brachytherapy in endometrial
cancer: Quantification of air gaps around a vaginal cylinder. Brachy-
therapy 2008;7:355e358.
[6] Richardson S, Palaniswaamy G, Grigsby P. Dosimetric effects of air
pockets around high-dose rate brachytherapy vaginal cylinders. Int J
Radiat Oncol Biol Phys 2010;78:276e279.
[7] Magne N, Chargari C, SanFilippo N, et al. Technical aspects and per-
spectives of the vaginal mold applicator for brachytherapy of gyneco-
logic malignancies. Brachytherapy 2010;9:274e277.
[8] Li S, Aref I, Walker E, et al. Effects of prescription depth, cylinder
size, treatment length, tip space, and curved end on doses in high-
dose-rate vaginal brachytherapy. Int J Radiat Oncol Biol Phys
2007;67:1268e1277.
[9] Nath R, Luxton G, Weaver KA, et al. Dosimetry of interstitial
brachytherapy sources: Recommendations of the AAPM Radiation
Therapy Committee Task Group No. 43. Med Phys 1995;22:
209e234.
[10] Demanes DJ, Rege S, Rodriquez RR, et al. The use and advantages of
a multichannel vaginal cylinder in high-dose rate brachytherapy. Int J
Radiat Oncol Biol Phys 1999;44:211e219.
[11] Yaparpalvi R, Mutyala S, Thawani N, et al. Variance in bladder and
rectal doses in the course of fractionated cylinder high-dose-rate
brachytherapy. Brachytherapy 2009;8:145.