1

Whole Breast Nodal Irradiation utilizing Supine VMAT and Prone 3D Planning: A Case
Study
Authors: Ashley Coffey, B.S., R.T.(T), Lisa Stevenson, B.S., R.T.(T), CMD, Ashley Hunzeker,
M.S., CMD, Nishele Lenards, M.S., CMD, R.T.(R)(T), FAAMD
Medical Dosimetry Program at University of Wisconsin- La Crosse, WI
Abstract
Introduction: The purpose of this study was to compare the variations in doses to the ipsilateral
lung and heart between supine whole breast and nodal VMAT technique and the prone 3DCRT
technique.
Case Description: In this study, 3 patients that were simulated in the prone and supine positions
were planned using a supine volumetric modulated arc therapy (VMAT) and prone threedimensional conformal radiation therapy (3DCRT) technique to compare planning target volume
(PTV) coverage as well as dose to organs at risk (OR) including the heart and ipsilateral lung.
Conclusion: While all constraints were met, the ideal treatment technique depended on the
patients anatomy and lumpectomy location. Supine VMAT provided excellent coverage to the
target structures but encountered difficulty limiting dose to the ipsilateral lung and heart. Prone
3DCRT limited dose to the ipsilateral lung and heart but provided less dose coverage to the target
volumes when compared to the supine VMAT technique.
Keywords: breast cancer, prone, supine, 3DCRT, VMAT
Introduction
The cumulative lifetime incidence of breast cancer is 1 in 8 U.S. women.1 In fact, breast
cancer is so common, it has been ranked the second most common malignancy affecting women
living in the United States after skin cancer. Although breast cancer is prevalent, the survival
rates have been steadily increasing since 1989 as a result of earlier diagnosis and the
development of more effective treatments, such as chemotherapy, hormone targeting drugs, and
radiation therapy. The increase in survival necessitates a greater need for therapies with
decreased toxicity to normal tissues, providing better cosmetic outcomes and decreasing the risk
of radiation-induced secondary malignancies.2

Supine 3DCRT, often treated with three fields monoisocentrically, was the gold standard
for many years. However, the monoisocentric technique had limitations that included less than
optimal PTV coverage, problematic junctions between the breast and nodal fields, and increased
radiation dose to surrounding healthy tissues.2 Nearly a decade ago, advances in technology
paved the way for intensity modulated radiation therapy (IMRT) techniques that increased PTV
coverage, better dose homogeneity, and decreased dose to surrounding structures.3 In 2007,
VMAT technique was introduced clinically for breast cancer treatment and has since become a
widely accepted method for treating left sided breast cancer patients with nodal involvement.
Although VMAT and IMRT are similar in PTV coverage, VMAT has better dose conformity,
decreased maximum dose to surrounding tissues, less monitor units (MUs) and decreased
treatment time.2
Recently, 3DCRT prone breast techniques have garnered interest in the radiation
oncology community. The prone position naturally pulls the breast away from the chest wall
allowing for the possibility of greater dose sparing to OR. There is currently a lack of data
comparing supine VMAT to prone nodal breast treatments. In this study, 3 patients that were
simulated in the prone and supine positions were planned using a supine VMAT and prone
3DCRT technique to compare PTV coverage as well as dose to OR including the heart and
ipsilateral lung.
Case Description
Patient Selection and Set-up
Patient selection was based on women with breast cancer and nodal involvement. All
patients in this study were diagnosed with grade 2 or 3 invasive ductal carcinoma and an intact
breast. Patients were simulated in both prone and supine positions to attempt to reduce organ
dose and deliver adequate dose to the breast and nodes. All patients were scanned in a General
Electric (GE) CT large bore scanner head first. For the supine scan, the patient was placed on a
15 tilt breast board with the ipsilateral arm up and their head turned in the opposite direction. In
the prone scan, the patient was placed on their abdomen on top of a prone breast board that
measured 24.5 cm off the CT table. There was an opening on the affected side that allowed for
the breast to fall forward and the supraclavicular area to be exposed in order to treat nodes
without obstruction. The head was also turned away from the affected side and a VacLok bag

was used to support the arms. Radio-opaque CT wires were placed on the skin to delineate the
edges of the breast tissue and tattoos were placed on the patient to reproduce daily setup. On
Board Imaging (OBI) was used daily for each patient.
Target Delineation
Target delineation was performed by the physician and medical dosimetrist on a Philips
Pinnacle v9.8 treatment planning system (TPS). Contours were created on the planning CT and
expanded following the Radiation Therapy Oncology Group (RTOG) 1304 protocol.4 The
protocol provided guidelines for contouring all target volumes and structures which included:
lumpectomy, breast, supraclavicular nodes, axillary nodes, and internal mammary nodes (IMN).
All clinical target volume (CTV) contours were delineated based on the RTOG anatomy atlas.5
Critical organs contoured included the ipsilateral and contralateral lung, contralateral breast,
heart, and thyroid.
The lumpectomy gross tumor volume (GTV) was contoured with available imaging and
included the lumpectomy cavity, lumpectomy scar, seroma, and surgical clips. Lumpectomy
CTV was created from a 1 cm expansion of the lumpectomy GTV that avoided the pectoralis
muscles, was cropped 5 mm from the skin, and did not cross midline. Lumpectomy PTV was
created by expanding the CTV by 7 mm in all directions excluding the heart.
Breast CTV was classified as all palpable breast tissue which was delineated at the time
of simulation with radio-opaque wires in the CT simulation. The Lumpectomy CTV was
included in this contour and excluded 5 mm of skin on the surface, the pectoralis, chest wall,
ribs, and lung. The Breast PTV was the Breast CTV with 7 mm expansions that avoided the heart
and did not cross midline. Breast PTV Eval was created by copying the Breast PTV and edited to
exclude air outside the patient, 5 mm skin, and anything deeper than the anterior surface of the
ribs. Breast PTV Eval was used for constraints in planning and dose volume histogram (DVH)
analysis.
Supraclavicular CTV was contoured using the RTOG Breast Cancer Atlas.5
Supraclavicular PTV was created by expanding the supraclavicular CTV by 5 mm in all
directions. The supraclavicular PTV did not include the thyroid, trachea, esophagus, lung, and
was contracted 5 mm from the skin surface.

Axillary CTV was contoured from the remaining, undissected axillary nodes. The
physician used the operative reports and other diagnostic imaging to determine what axillary
nodes were to be included in planning. Typically, level I and II axillary nodes are removed so the
level III nodes and any other remaining nodal levels must be included in the axillary CTV.
Axillary levels can be found on the RTOG Breast Cancer Atlas.5 Axillary PTV included a 5 mm
expansion of the CTV excluding lung. Internal Mammary Node CTV included the internal
mammary nodes and thoracic vessels in the first 3 intercostal spaces. The IMN PTV was 5 mm
expansion from the IMN CTV medially, laterally, superiorly, and inferiorly. The IMN PTV
excluded the sternum, lung, and heart.
Treatment Planning
Each patient used for planning had scans performed on the same day in both prone and
supine positions. A total of 3 patients were planned with 3DCRT in the prone position and with
VMAT in the supine position. Each patient were prescribed a dose of 50 Gy in 25 fractions to the
whole breast and nodal regions. The lumpectomy was to receive a boost of 10 Gy in 5 fractions.
The 3DCRT prone plans for all patients utilized single isocenter tangents for the whole breast
along with an anterior and posterior supraclavicular field.
Patient 1 was planned supine using a VMAT technique that utilized four 249 partial arcs
with split beams to allow more adequate multi-leaf collimator (MLC) range. The prescription
dose was normalized to the 100% isodose line resulting in a maximum dose point of 106.7%.
Patient 1 was also planned prone using 6 MV tangential beams included a right anterior oblique
(RAO) and a left posterior oblique (LPO). The beam angles were determined based delivering
acceptable dose to the IMN PTV and axillary level III nodes without delivering excess dose to
the heart. The prescription dose was renormalized to the 94% isodose line to achieve better dose
homogeneity resulting in a maximum dose point of 106%.
Patient 2 was planned supine using a VMAT technique that utilized four 249 partial arcs
with split beams to allow more adequate MLC range. The prescription dose was normalized to
the 100% isodose line resulting in a maximum dose of 113% with less than 1% of the volume
reaching doses higher than 108%. Patient 2 was also planned in the prone fashion with four
18MV beams to the right breast. The arrangements included medial and lateral tangents as well
as a left anterior oblique (LAO) and a right posterior oblique (RPO). The beam angles for this

patient were selected based on avoiding the contralateral breast and minimizing the amount of
lung in the field. The patient had a separation of more than 30 centimeters at the lumpectomy
cavity, which made delivering a uniform dose across that field difficult. In order to receive
adequate coverage, the prescription dose to the tangent fields was normalized to the 88% isodose
line and the prescription dose to the supraclavicular field was renormalized to the 95% isodose
line. The normalization increased the maximum dose to 112%, which is acceptable, but a slightly
higher dose than a standard 3DCRT plan.
Patient 3 was planned supine using a VMAT technique that utilized two 249 partial arcs
with split beams to allow more adequate MLC range. The prescription dose was normalized to
the 100% isodose line resulting in a maximum dose point of 110%. However only 0.01% of the
entire volume received more than 54 Gy thus meeting protocol parameters. Patient 3 was also
planned prone using 6MV and 18MV tangential beams and a RAO and a LPO. The beams angles
were decided based on coverage of the IMN PTV while avoiding as much heart as possible and
providing axillary level III coverage. The prescription dose was renormalized to the 92% isodose
line resulting in a maximum dose 108%.
Plan Analysis and Evaluation
The supine positioning for patient 1 was ideal for VMAT planning. The incline of breast
board combined with the arm on the affected side being raised above the patients head
adequately lifted the involved breast up and away from the contralateral breast resulting in
decreased dose to the contralateral breast. All protocol constraints for target and OR were either
met at the protocol or acceptable variation level; however, there was great difficulty in achieving
the ipsilateral lung constraint of 65% 5Gy without losing Breast PTV Eval per protocol
coverage. The acceptable variation of 70% 5Gy was finally attained after special contours
were created and optimized to decrease lung dose. All heart constraints met the per protocol
parameters except the mean dose which met the acceptable variation mean dose of 4.65Gy.
Patient 1 was also in ideal prone position for producing a 3DCRT plan. The positioning on the
board was straight and leveled; the contralateral breast was properly pulled away from the
affected breast and aided in clearance for tangent fields without the contralateral breast receiving
any dose. All of the constraints for the heart and lung were met exceedingly below the ideal per

protocol constraints. The dosimetric results for OR and PTV targets for both plans on Patient 1
are listed in Table 1.
Patient 2 was difficult to plan in the supine position. Although all protocol constraints for
targets and OR were either met at the protocol or acceptable variation level, there was great
difficulty achieving the ipsilateral lung to meet the 65% 5Gy without losing Breast PTV
Eval coverage. This was largely due the close proximity of the lumpectomy site to the chest
wall. The acceptable variation of 70% 5Gy was finally attained after special contours were
created and optimized to decrease lung dose. This did decrease coverage to all 4 PTV structures,
however, from per protocol to acceptable variation levels. Patient 2 provided a challenge
planning in the prone fashion as well. At the lumpectomy site, the patient had a 33.2 cm
separation making it difficult to achieve a homogeneous dose distribution with tangent fields.
After various methods such as adding smaller fields and using multiple energies, the end result
was reached with 18 MV tangents, multiple control points, and prescribing to a lower isodose
line than normal. The prone position pulled the patients breast away from the chest wall,
however, the breast then rested on the table which resulted in a bolus effect that increased dose to
the distal portion of the breast. The patient also had the contralateral breast extending off of the
opposite side of the breast positioning device. Unfortunately, this poor positioning in simulation
meant that consideration needed to be taken when selecting tangent angles in order avoid
entrance and exit dose into the contralateral breast. Even with positioning challenges, all of the
dose constraints were met for the ipsilateral lung and heart in the per protocol category while
still adequately covering all PTV structures. The dosimetric results for the OR and PTV tangents
for both plans on Patient 2 are listed in Table 2.
Patient 3 was successfully planned in the supine position. All protocol constraints for
target and OR were either met at the protocol or acceptable variation level. While trying to
decrease the mean heart dose to under 5 Gy and bring the lung dose to 70% 5Gy, IMV PTV
coverage decreased from per protocol level to the acceptable variation level. Patient 3 was not in
the most efficient position for prone planning. The contralateral breast was not positioned
correctly on the board and therefore was partially in the board opening utilized for the affected
breast. The contralateral breast was avoided by adjusting the angles utilized; however, it was not
ideally positioned for accuracy and reproducible treatments. With the prone planning, IMN PTV
coverage was compromised to only variation acceptable coverage in order to meet the acceptable

heart mean of 5 Gy. Even with the challenges presented in positioning and heart location, all
dose constraints were met and the coverage to remaining nodal PTVs and the lumpectomy site
were met. The dosimetric results for OR and PTV targets for both plans on Patient 3 are listed in
Table 3.
After evaluation of the ipsilateral lung and heart doses for each patient, there was not a
clear method as to which treatment type was better. A few factors contribute to the decision of
treating a patient in the supine or prone technique. One factor is where the patients cavity is
located. When the cavity is closer to the chest wall, it is easier to receive better target coverage
and lower OR doses in the prone conformal fashion than it is with supine VMAT. Another factor
to consider is the challenges posed in planning a prone 3DCRT plan for a patient with a larger
separation. With the breast extending away from the patient, the field becomes large to cover the
nodal targets as well as the whole breast. With larger patients, there also tends to be a roll
towards the affected breast, which causes more lung and heart to fall into the tangent fields in a
prone setup. In a study performed by Huppert et al,6 patients with left sided breast cancer were
found to have their heart extending into the prone breast field due to an axial rotation. The
solution was turning the patients head toward the affected side which straightened out the
patients roll and pulled the heart out of the tangential fields. Based off Huppert et al6 results,
NYU Trial 05-181 was developed. The NYU trial 05-181 enrolled 400 patients with breast
cancer to be simulated both supine and prone positions. The heart dose was evaluated and found
that prone position decreased the amount of heart in the field by 11cc. This complements the
results found in the current retrospective study. In figure 1, the mean heart doses were compared
for both prone and supine positions; although not significant, the prone plans all produced lower
mean heart doses than the supine plans.
The supine VMAT technique provided great target coverage of the breast and nodes but
posed challenges with increased low dose to the lung and heart. This aspect of the supine VMAT
breast technique was observed in other studies comparing supine IMRT to supine VMAT as well
and has been attributed to the increased exit dose given to the lung and heart due to the beam
emitting radiation during the entire semi-arc of the VMAT field as opposed to stationary IMRT
beams arranged to avoid these structures.7 This is of significance because each additional Gray
added to the heart mean dose can increase the rate of cardiac toxicity by as much as 4%.3 Prone
conformal technique kept the OR structure doses very low; however, the target structures were

only adequately covered. Overall, the ideal treatment method changed on a patient by patient
basis. It should be determined by the physician and the patient as to which method should be
used.

Conclusion
The purpose of this study was to compare supine VMAT versus prone 3DCRT in the
delivery of external beam radiation therapy for patients with breast and nodal disease. This study
demonstrated that while supine VMAT plans achieved better PTV coverage than prone 3DCRT,
supine VMAT techniques generated higher doses to the ipsilateral lung and heart. The most
striking variation was the increased mean dose given to the heart during supine VMAT
treatments. The increase resulted in supine VMAT technique delivering double the heart mean
dose than that of the prone 3DCRT. The average increased mean dose per patient was 1.6 Gy.
There were also limitations to this study. The study was performed retrospectively, which
was not ideal. Patient positioning could have been improved and results could be more beneficial
if the study was conducted prospectively. In the prone scans, the patients were often rolled or did
not have the contralateral breast positioned out of the field resulting in planning challenges such
as limitations of ideal gantry angles. Even with working around these issues and achieving
treatable plans, it would have been ideal to have the patients set up correctly on their initial
planning scan.

References
1.

Breast Cancer. American Cancer Society website.

http://www.cancer.org/cancer/breastcancer/detailedguide/breast-cancer-key-statistics.
Accessed June 5, 2016.
2.
Teoh M, Clark CH, Wood K, Whitaker S, Nisbet A. Volumetric modulated arc
therapy: a review of current literature and clinical use in practice. Br J Radiol. 2011; 84: 967996. http://dx.doi.org/10.1259/bjr/22373346
3.
Zhao H, He M, Cheng G, et al. A comparative dosimetric study of left sided breast
cancer after breast-conserving surgery treated with VMAT and IMRT. Radiat Oncol.
2015;10(231):e10. http://doi.org/10.1186/s13014-015-0531-4
4.
Mamounas E, White J. NRG Oncology NSABP Protocol B-51/RTOG Protocol
1304. Radiation Therapy Oncology Group (RTOG).
https://www.rtog.org/ClinicalTrials/ProtocolTable/StudyDetails.aspx?study=1304. Published
2013. Updated 2016.
5.
RTOG Breast Cancer Atlas. Radiation Therapy Oncology Group website.
https://www.rtog.org/CoreLab/ContouringAtlases/BreastCancerAtlas.aspx. Accessed June 7,
2016
6.

Huppert N, Jozsef G, DeWyngaert K, et al. The role of a prone setup in breast

radiation therapy. Front Oncol. 2011;1(31):e10. http://doi.org/10.3389/fonc.2011.00031

7.
Ali MA, Babaiah M, Madhusudhan N, George G, Jain S, Ramalingam K, Kumar
SA, Karthinkeyan K, Anantharaman A. Comparative dosimetric analysis of IMRT and
VMAT (Rapid Arc) in brain, head and neck, breast and prostate malignancies. Int J Cancer
Ther Oncol 2015; 3(1):03019. http://doi.org/10.14319/ijcto.03019

10

Figures

Figure 1. The mean heart dose for each patient in both prone and supine position. The prone 3D
consistently showed less heart dose than supine VMAT.

11

Tables
Table 1. Dose constraints for the target volumes and OR for patient 1.
Constraints

Per Protocol

Acceptable
Variation

Patient 1
Prone

Patient 1
Supine

Breast PTV Eval

95% of 47.5Gy

90% of 45Gy

95.9%

97.6%

SCL PTV

95% of 47.5Gy

90% of 45Gy

98.5%

96.8%

Axillary PTV

95% of 47.5Gy

90% of 45Gy

97.6%

99.1%

IMN PTV

95% of 45Gy

90% of 40Gy

99.0%

98.7%

Heart

<5% at 25Gy

<5% at 30Gy

0.10%

1.1%

Heart mean

4Gy

5Gy

2.1Gy

4.7Gy

Ipsilateral Lung

<30% at 20Gy

<35% at 20Gy

25.4%

23.4%

The green signifies meeting the ideal per protocol constraint and yellow correlates to the
acceptable variation.
Table 2. Dose constraints for the target volumes and OR for patient 2.
Constraints

Per Protocol

Acceptable
Variation

Patient 2
Prone

Patient 2
Supine

Breast PTV Eval

95% of 47.5Gy

90% of 45Gy

95.7%

96.8%

SCL PTV

95% of 47.5Gy

90% of 45Gy

94.9%

97.3%

Axillary PTV

95% of 47.5Gy

90% of 45Gy

98.7%

98.2%

IMN PTV

95% of 45Gy

90% of 40Gy

90.7%

99.9%

Heart

<5% at 25Gy

<5% at 30Gy

1.04%

0.5%

Heart mean

4Gy

5Gy

1.8 Gy

3.7Gy

Ipsilateral Lung

<30% at 20Gy

<35% at 20Gy

16.8%

21.0%

The green signifies meeting the ideal per protocol constraint and yellow correlates to the
acceptable variation.

Table 3. Dose constraints for the target volumes and OR for patient 3.

12

Constraints

Per Protocol

Acceptable
Variation

Patient 3
Prone

Patient 3
Supine

Breast PTV Eval

95% of 47.5Gy

90% of 45Gy

97.0%

95.7%

SCL PTV

95% of 47.5Gy

90% of 45Gy

98.7%

98.3%

Axillary PTV

95% of 47.5Gy

90% of 45Gy

98.4%

97.7%

IMN PTV

95% of 45Gy

90% of 40Gy

90.3%

99.9%

Heart

<5% at 25Gy

<5% at 30Gy

4.35%

0.8%

Heart mean

4Gy

5Gy

4.4Gy

4.7Gy

Ipsilateral Lung

<30% at 20Gy

<35% at 20Gy

15.6%

17.9%

The green signifies meeting the ideal per protocol constraint and yellow correlates to the
acceptable variation.