Alyssa Olson

Comparative Article Analysis Part I: Trade Journal

There are stark differences in the format and design of a trade magazine versus a peerreviewed journal. A quick browse through the trade magazine will entice the reader through
colorful and interesting graphics, short-read style articles, and advertisements embedded within
each page.1 A quick browse through the latter will reveal just the opposite. An article from the
trade magazine, RT Image, entitled In the Right Direction: Improving Quality and Safety One
Measure at a Time by Todd Pawlicki captured my attention through a front page feature of an
intricate radiation therapy treatment set-up and will be the subject of this article analysis. Areas
of discussion within the analysis will include a brief summary, relevance and accuracy of
presented information, and personal opinions of its strengths and weaknesses.
The overall message of the article focused on the importance of quality assurance
measures in the radiation oncology setting.2 Due to the relatively recent life-threatening
treatment mishaps and the general public concern of radiation exposure, the rapport of radiation
therapy treatment may be potentially jeopardized in the public eye. While no two treatments or
linear accelerators are alike, Pawlicki emphasizes the importance of establishing a strong quality
assurance program amongst radiation oncology treatment centers. This process, commonly
referred to as statistical process control (SPC), begins by determining an area of improvement in
which data can be gathered, measured, and assessed. With the implementation of departmental
policies, radiation therapy facilities should be able to maintain a high level of care through use of
good judgment, critical thinking skills, and recommendations of quality practice through the
American Association of Physicists in Medicine (AAPM) task group reports.
Quality assurance assessments are daily common practices completed by many medical
dosimetrists. In addition to designing a safe and accurate treatment plans, dosimetrists will also
perform their own quality assurance check of treatment planning dose verification. Knowing
what action to take when quality assurance dose verifications are out of tolerance is crucial. For
that reason alone, this article is quite relevant to any individual in the medical dosimetry
profession. Although the article is authored by a medical physicist and is geared towards the
medical physics community, dosimetrists and students alike should still be able to apply the

Olson 2
same concepts of utilizing good judgment and policy compliance when treatment planning.
Aside from relevance, the validity of information presented was more difficult to assess due its
informal editorial-type format. The author focused more on presenting his opinion of why
quality assurance procedures are valuable rather than providing any statistical data to support his
reasoning. He does, however, encourage departments to follow the recommendations of the
internationally renowned organization, the AAPM , when designing quality assurance policies.1
While it is apparent this article is more opinion based rather than statistically supported,
there are positive aspects that should be noted. The organizational structure and flow of the
article was easy to follow as the article began with defining quality and progressively introduced
his suggested step-by-step process for measurement and assessment. The goal of professional
journals is to keep professionals engaged within their profession.1 Pawlicki did just that, as it was
obvious of his intentions to improve the standard of care in the field of radiation oncology. On
the contrary, despite the article being published in a radiology professional journal, the target
audience of this article was in radiation oncology. Because of my background in radiation
therapy, I was able to relate to and understand the brief descriptions of equipment and dose
verification processes discussed. However, as an outsider, I could see a diminishing interest in
reading the entire article due to the inability to relate. This article had a powerful message of how
to provide a high standard of care to patients and may have had more of an impact on the
professional audience if radiation therapy were not the primary focus. This topic could be
applicable to any profession because quality assurance is needed in all aspects of healthcare for
continuous improvement.
Although Pawlicki focused on quality assurance assessments in radiation therapy, it
should not deter readers from understanding the authors powerful message. Frequent deviations
or variations in a daily task, such as patient treatment positioning, is a key identifier to a process
that could be improved.2 Having a departmental policy in place for quality assurance measures,
such as tolerance limits on patient verification checks for medical physicists, will be helpful in
making appropriate decisions when unpredictable situations arise. There are several resources
available to facilities to help create effective quality assurance plans, such as the task group
reports from the AAPM. These are continuously updated and encourage facilities to maintain a
high level of care provided to patients.

Olson 3
References
1. Lenards N, Weege M. Reading & Writing in Radiation Therapy & Medical Dosimetry.
[SoftChalk]. La Crosse, WI: UW-L Medical Dosimetry and Radiation Therapy Program;
2016.
2. Pawlicki T. In the right direction: improving quality and safety, one measure at a time. RT
Image. 2010; 23(18):12-15. Retrieved from
http://www.onlinedigitalpubs.com/publication/?i=48439&page=1. Accessed February 28,
2016.

Olson 4
Comparative Article Analysis Part II: Peer-Reviewed Journal
The peer-reviewed professional journal, Medical Dosimetry, is a prestigious membershipbased journal which strives to improve the standard of care in treatment planning through
research and statistical studies. Peer review articles undergo an extensive peer-review process
prior to publication and have strict formatting requirements.1 One article in particular within this
journal peaked my interest: Dosimetric evaluation of whole-breast radiation therapy: clinical
experience written by Osei et al.2 In addition to a brief summary, the following article analysis
will include an evaluation of the literature review, research design, reported results and
conclusions, and an overall personal impression.
The proposed question to the research study completed by Osei and colleagues2 was to
evaluate the efficacy of a standardized dose and target coverage regimen specifically designed
for their treatment facilities for whole-breast irradiation. According to Osei et al,2 a total of 621
patients who received whole-breast radiation in courses of either 4256 cGy in 16 fractions or
5000 cGy in 25 fractions were evaluated over a period of 18 months. Specific areas of
assessment included treatment anatomical measurements, treatment technique type, organ at risk
(OR) constraints, and target coverage outcomes. Three whole-breast irradiation techniques were
evaluated: three-dimensional (3D) conformal, deep-inspiration breath-hold (DIBH) for left-sided
breast cancer, and hybrid 3D conformal with intensity modulated radiation therapy (IMRT).
Through research, it was hoped that the study outcome would shed light on validating the
standardized regimen or find ways to improve patient care.
In order to help readers understand the purpose for the study, background information
was provided. The literature review, although quite vague and brief, provided supporting
information on other published research studies which also analyzed the efficacy of different
breast irradiation techniques.2 The primary focus of each externally presented study was an
evaluation of the dose distribution and hot spots using different treatment techniques. This made
a smooth transition into how and why the research study was conducted by Osei et al.
The methods section clearly explained the treatment process for each whole-breast
irradiation technique studied. Over 18 months, 621 whole-breast patient treatments were
evaluated.2 Osei et al delved into detail with regard to the simulation process, target delineation,
and treatment planning development. In addition, field arrangements were described for both 3D

Olson 5
and hybrid techniques, which allowed for potential reproducibility if the reader so desired. The
treatment technique utilized for each patient was specific to their body habitus and staging. All
techniques were planned using the same field borders and target volumethe PTV_eval.
Following the methods section, a discussion of the results and conclusions were provided.
Tables and graphs were created using a compilation of results from each treatment
technique. Total coverage of the PTV_eval was assessed in addition to the doses received by OR
using V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy. The results validated and achieved the desired
outcome for the standardized dose and coverage regimen in all three techniques: at least 92% of
prescription dose covered 99% of the PTV_eval, at least 95% of prescription dose covered 97%
of the PTV_eval, and less than 1% of PTV_eval received above 105% of prescription dose.
Results determined that there was a statistical difference between the 3D conformal and DIBH
technique regarding total treated heart volume in patients with left-sided breast cancer, totaling
606.5 cm3 and 543.7 cm3 respectively. Finally, the hybrid technique proved to be superior to the
traditional wedged-tangent fields in cases of larger breasted women, providing optimal coverage
with minimal hot spots. From these results, this research group concluded that they will continue
to utilize their standard dose and coverage regimen, knowing that all constraints and dose
objectives can be achieved for any whole-breast technique utilized.
My overall impression of the research article was quite positive. It is apparent that the
research group conducted their study with thoroughness and precision. All tables and literature
presented were easy for any medical dosimetry professionalincluding students--to follow and
understand. On the contrary, while the priority of the study was to evaluate the current
standardized regimen, there was more focus on the use of the hybrid technique in the results and
concluding statements. This issue, however, should not distract readers from understanding the
main focus of the publication; it was simply an observation.
This publication is an excellent example of a high quality peer-reviewed journal article.
Like other peer-reviewed articles, it was obvious that the authors intentions of conducting this
study was to evaluate current standards for process-improvement opportunities.1 Achieving
good dose distribution with minimal hot spots when designing breast plans can be difficult.
Because this study demonstrated optimal results, I may refer back to this for future planning
considerationsespecially for larger breasted women.

Olson 6

References
1. Lenards N, Weege M. Reading & Writing in Radiation Therapy & Medical Dosimetry.
[SoftChalk]. La Crosse, WI: UW-L Medical Dosimetry and Radiation Therapy Program;
2016.
2.

Osei E, Darko J, Fleck A, et al. Dosimetric evaluation of whole-breast radiation therapy:

clinical experience. Med Dos. 2015;40: 355-365.
doi:http://dx.doi.org/10.1016/j.meddos.2015.05.001

In the

RIGHT
D IRECTION

Improving quality
and safety, one
measurement at a time

istockphoto.com/Mark Kostich

Todd Pawlicki, PhD

|12|

S e p t e m b e r 2 0, 2 0 1 0

EFFORTS TOWARDS QUALITY IMPROVEMENT AND ERROR REPORTING HAVE A

long history in medicine. However, a 1999 report To Err Is Human from the National
Academy of Sciences' Institute of Medicine cited studies that showed between
44,000 and 98,000 people die each year because of mistakes by medical professionals.
In the interim, significant progress has been made toward quality improvement in
medicine by considering health care as a process that happens within a system.

Quality and safety in medicine have fallen short of levels

realized in other industries. The answer is to adapt quality and
safety techniques from other industries to health care, and the
field of radiation oncology is no different.
Radiation oncology has a unique confluence of health care
professionals; physicians, physicists, dosimetrists, therapists, and
nurses. Radiation oncology is technological, and along with the
rest of medicine, has had a long history of quality and safety
efforts. However, recent publicized errors in radiation oncology,
some of which have cost patients their lives, indicate that opportunities for quality improvements still exist.
The seminal question is: How can we achieve improvement?
It begins by thinking about the meaning of quality.

DETERMINING QUALITY
What is quality and what is the goal of a quality assurance
(QA) program? Crosby, Feigenbaum, Juran, and Taguchi offered
the following descriptions: conformance to requirements; whats
best for patient use and value; fitness for use or purpose; meeting the patients requirements; zero defects; or the loss imparted
to the patient from the time the service is provided.
In radiation therapy, there are two types of QA: inter-patient QA;
and intra-patient or process/equipment QA. Inter-patient QA is the
detailed investigation of any patients experience, from consultation
to treatment. Intra-patient QA is the detailed investigation of the
treatment processes that are general to all patients or patient types.
There are different degrees of quality; it is not only good or
bad. What leads to different degrees of quality is variation in the
intra- or inter-processes of patient treatments. Any definition of
quality should, therefore, include a component that can be measured. Without a quantitative value for quality, you would never
be able to determine if radiation treatments in a department are
optimal or how treatments compare to other departments.

TRACKING THE TREATMENT PROCESS

Some variation can be perceived as necessary for good
quality treatments because each patient is different and a
patients response to treatment is unique. Following the lead of
modern data- and process-based quality improvement, the key to
improving quality is to understand and minimize variation. This
applies to repetitive processes that have a defined goal and
produce data (metrics). The issue of variation is not relevant to
process design or similar activities, such as deciding if a patient
should be treated with radiation.
There are steps in a patients treatment where different
approaches allow one to achieve the same goal. This means that
those different approaches are in the noise of the system. Any
effort to standardize or improve approaches that are in the
system noise is a waste of time because you are then trying to
affect results where the optimal path to achieve a goal is not, or
cannot, be defined.
All steps in patient treatment have some limits on what is
acceptable. These limits are clinical guidelines, best practices, or
specifications found in American Association of Physicists
in Medicine (AAPM) task group reports. Variation outside
acceptable limits is obviously a bad thing.
What is not so obvious is that large variation, even within
acceptable limits, is a detriment to quality improvement. Large
variation makes it impossible to correlate a poor result with any
input parameters that could be changed to produce a better result
in the future. It is just a matter of applying a scientific approach
to quality improvement.
Quality cannot be poor because you are too
precise. Quality is poor because you may be focused on
improving a result that is in the noise in the system.
Something cannot be improved that does not have a
well-defined target value or goal.
| w w w. r t - i m a g e .c o m |

S e p t e m b e r 2 0, 2 0 1 0

|13|

CONTROLLING A PROCESS
Statistical process control (SPC) is an idea that is supported by
established research in fields outside of health care. The benefits
of SPC in manufacturing, engineering, and the service industries
are undeniable. However, SPC is new to radiation therapy and
deserves an explanation.
To describe SPC, we first note that no two things are the same.
This is true of a radiation measurement from a linear accelerator.
It is also true for two identical treatments two subsequent
fractions in the course of radiation therapy for the same patient.
SPC is a quantitative method that allows you to identify cases
where the difference is due to an assignable cause.
Radiation treatments are technical in nature and require
complex delivery machines, radiation measurement equipment,
and procedures. Furthermore, you are required to follow the
same steps for each patient once a physician prescribes the dose.
In the past, the paradigm of manufacturing QA has followed
a predictable path: make, inspect, sort, and act. It is an intuitive
procedure. If you are making the same part, you want to know if
the specifications of that part are met so it will work with the other
parts, like assembling a car. If the part being tested meets
specifications, then it is OK to be used. If the part does not meet
specifications, then it is sent back for repairs. If it cannot be fixed,
it is scrapped and you have to start over.
The medical physics community routinely uses the make,
inspect, sort, and act approach for QA. Physicists obtain a
measurement to determine if it is out of some pre-set limits. If the
measurement is out of limits, you can take another measurement

|14|

S e p t e m b e r 2 0, 2 0 1 0

By the nature of physics training, a

medical physicist believes that
there is an understandable reason
for the results that he obtains.

to check it again. If the measurement is still out of limits, you can

proceed to identify the underlying cause.
At this point, different things may happen, depending on the
magnitude of the error. Lets consider a hypothetical situation to
measure the output of the linear accelerator (where limits are 2
percent). If the output is 10 percent high, the decision is easy
no treatments until the problem is identified and corrected. If the
output is 0.2 percent high, the decision is also easy fill out the
QA documentation and allow patients to be treated.
However, what if the output is 2.1 percent high? Now the
physicist has to make a decision. Should you let it go and
measure again tomorrow? Should you adjust the linear accelerator
to reduce the output? Or, is this result within the stability limits of
your linear accelerator output and measurement system, so
nothing can be done without re-engineering the system?
This last question is where SPC can provide a scientific-based
answer. Physicists deal with these types of questions regularly in
radiation therapy. Can the difference be resolved between a
calculated value versus a measured value, or the difference
between a Monte Carlo calculated dose value versus a conventionally calculated dose value, or hand-calculated monitor units
versus computer calculated monitor units? Usually, a physicist has
to use his own clinical judgment to determine to accept the
difference and proceed with treatment.
Clinical judgment is not completely on the shoulders of the
individual physicist making the decision. Guidelines within the
medical physics community give physicists some idea of what to
do. However, guidelines dont give physicists the limits for
your linear accelerator. For example, it is possible that the linear
accelerator at your institution is stable and the output can be kept
to within 0.5 percent. Techniques of SPC can answer this question.
By the nature of physics training, a medical physicist believes
that there is an understandable reason for the results that he
obtains. For example, if the hand calculated monitor units
differ by 5 percent from what the treatment planning computer
calculated, then start looking for answers. Transcription errors
and other silly mistakes are ruled out first. Then, you are left to
consider the approximations of the hand calculation compared
with the computer calculations to determine if the error is of an
acceptable magnitude and in the right direction.
Once this parameter is justified, you can approve the
treatment plan. Given the time it takes, it is impractical to system-

atically investigate each occurrence like the one just described. This highlights the point that
clinical medical physics is a product and a service. That is to say, clinical medical physics is
not a scientific research project. To treat it as such would be to the detriment of many patients
waiting for treatment.
You have to address two pertinent questions. What is an acceptable level of variability?
When is it necessary for the physicist to address that variability?

It is important to focus on input metrics rather than

outcome metrics. Variability in a patients
response to treatment can obscure the
effects of changes to process inputs geared
toward quality improvement.

Now an 80% dose reduction

can mean...

MEASURING UP
It is important to focus on input metrics rather than outcome metrics. Variability in a patients
response to treatment can obscure the effects of changes to process inputs geared toward
quality improvement. While focusing on the input metrics to improve quality, you must
separate noise from true signals that can cause a problem, error, or an even better result (not
all signals are negative).
The way to do this is by taking the following steps:

i
i

define goals of clinical performance for different steps (targets)

i
i

collect data to see how youre doing

set specification limits on targets, such as defining a clinically acceptable range of

performance

adopt causes of signals that take you closer to the goal, and minimize variation and
eliminate those causes that take you farther away.
Although it seems simple, successful implementation requires a high degree of domain
expertise, coupled with using quality and safety tools from the industry.
| Todd Pawlicki, PhD, is associate professor and director of the division of medical
physics, department of radiation oncology, at the University of California, San Diego in La
Jolla, Calif. He is also co-founder of a new organization created to broaden the implementation of process improvement strategies to minimize errors and maximize quality in radiation medicine. The organization provides workshops and educational materials for clinicians and institutions to implement functional quality improvement and error management programs within organizations. For more information and course schedules, go to
www.aqusi.org. Direct comments and questions to editorial@rt-image.com.

http://www.facebook.com/rt.image
| w w w. r t - i m a g e .c o m |

S e p t e m b e r 2 0, 2 0 1 0

|15|

Medical Dosimetry 40 (2015) 355365

Medical Dosimetry
journal homepage: www.meddos.org

Dosimetric evaluation of whole-breast radiation therapy:

Clinical experience
Ernest Osei, Ph.D.,* Johnson Darko, Ph.D.,* Andre Fleck, M.Sc.,* Jana White, B.Sc., M.R.T.T.,
Alexander Kiciak, Rachel Redekop, and Darin Gopaul, M.D.
Department of Medical Physics, Grand River Regional Cancer Centre, Kitchener, Ontario, Canada; Department of Physics and Astronomy, University of
Waterloo, Waterloo, Ontario, Canada; Department of Systems Design, University of Waterloo, Waterloo, Ontario, Canada; Department of Radiation Therapy,
Grand River Regional Cancer Centre, Kitchener, Ontario, Canada; and Department of Radiation Oncology, Grand River Regional Cancer Centre, Kitchener,
Ontario, Canada

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 25 November 2014
Received in revised form
15 April 2015
Accepted 9 May 2015

Radiation therapy of the intact breast is the standard therapy for preventing local recurrence of earlystage breast cancer following breast conservation surgery. To improve patient standard of care, there is a
need to dene a consistent and transparent treatment path for all patients that reduces signicance
variations in the acceptability of treatment plans. There is lack of consistency among institutions or
individuals about what is considered an acceptable treatment plan: target coverage vis--vis dose to
organs at risk (OAR). Clinical trials usually resolve these issues, as the criteria for an acceptable plan
within the trial (target coverage and doses to OAR) are well dened. We developed an institutional
criterion for accepting breast treatment plans in 2006 after analyzing treatment data of approximately
200 patients. The purpose of this article is to report on the dosimetric review of 623 patients treated in
the last 18 months to evaluate the effectiveness of the previously developed plan acceptability criteria
and any possible changes necessary to further improve patient care. The mean patient age is 61.6 years
(range: 25.2 to 93.0 years). The mean breast separation for all the patients is 21.0 cm (range: 12.4 to
34.9 cm), and the mean planning target volume (PTV_eval) (breast volume for evaluation) is 884.0 cm3
(range: 73.6 to 3684.6 cm3). Overall, 314 (50.4%) patients had the disease in the left breast and 309
(49.6%) had it in the right breast. A total of 147 (23.6%) patients were treated using the deep inspiration
breath-hold (DIBH) technique. The mean normalized PTV_eval receiving at least 92% (V92% PD) and 95%
(V95% PD) of the prescribed dose (PD) are more than 99% and 97%, respectively, for all patients. The mean
normalized PTV_eval receiving at least 105% (V105% PD) of the PD is less than 1% for all groups. The mean
homogeneity index (HI), uniformity index (UI), and conformity index (CI) for the PTV_eval are 0.09
(range: 0.05 to 0.15), 1.07 (range: 0.46 to 1.11), and 0.98 (range: 0.92 to 1.0), respectively. Our data conrm
the signicant advantage of using DIBH to reduce heart dose when compared with the free-breathing
technique. The p values analyses of the results for the V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy for the heart
comparing DIBH and free-breathing techniques are well less than 0.05 (i.e., p o 0.05). However, similar
analyses for the lung give values greater than 0.05 (i.e., p 4 0.05), indicating that there is no signicant
difference in lung dose comparing the 2 treatment techniques.
& 2015 American Association of Medical Dosimetrists.

Keywords:
Dose-volume histogram
Breast treatment
Breath-hold technique
Free-breathing technique
Hybrid treatment
Radiation therapy

Introduction
Radiation therapy of the intact breast is the standard therapy
for preventing local recurrence of early-stage breast cancer

Reprint requests to Ernest Osei, Ph.D., Department of Medical Physics, Grand

River Cancer Center, 835 King Street West, Kitchner, Ontario, Canada N2G1G3.
E-mail: ernest.osei@grhosp.on.ca
http://dx.doi.org/10.1016/j.meddos.2015.05.001
0958-3947/Copyright 2015 American Association of Medical Dosimetrists

following breast conservation surgery. To improve patient standard of care, we have developed an evaluation process to dene a
consistent and transparent treatment path for all patients that
reduces signicant variations in the acceptability of treatment
plans. Over the past few years, many studies have investigated the
dosimetric differences between traditional treatment planning
techniques using wedges and more recently developed methods
such as the eld-in-eld (FnF), irregular surface compensation,

356

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

hybrid technique, and inverse planned intensity-modulated radiation therapy (IMRT).1-28 Target coverage and dose uniformity
parameters such as the planning target volume (PTV) doses, V95,
V100, V105, and V107 (PTV volume receiving 95%, 100%, 105%, and
107% of the prescribed dose, respectively) as well as doses to the
organs at risk (OAR) such as the ipsilateral lung, heart, and
contralateral breast have been compared.1-4,24 A number of
studies have reported signicantly better dose distributions with
FnF2,4,7 and signicant dose reduction to the OAR.3,7,24 Gursel
et al.3 attributed these improvements to the fact that the FnF
technique reduces both the scatter and treatment time. On the
contrary, Sun et al.5 reported that the use of the FnF technique
is not superior to the physical wedges technique, as it does
not improve the dosimetric results. They found that although
the number of monitor units delivered was lower and the
uniformity index (UI) was higher for the FnF technique, the
physical wedges technique had a decreased homogeneity index

(HI) and maximum dose to the OAR, along with an increased

conformity index (CI).5
The effect of respiratory motion on breast radiotherapy treatment has also been investigated.6,9 Tanaka et al.9 conducted an
investigation to determine whether the FnF technique was more
susceptible to respiratory motion than the physical wedges technique. By comparing data from the free-breathing and the deep
inspiration breath-hold (DIBH) methods, they concluded that the
FnF technique is less affected by respiratory motion when compared with the physical wedges technique.9 The dose-volume
histogram (DVH) data showed that the difference in the volume
of PTV receiving 90%, 95%, and 107% of the prescribed dose
between the free-breathing and the DIBH techniques was less
pronounced for the FnF technique than for the physical wedges
technique.9 The dosimetric aspect involved in the use of an
irregular surface compensator has been compared with treatment
plans involving physical wedges.11 The use of the irregular surface

Fig. 1. Structure segmentation showing PTV_eval, lung, and heart volumes. The 50% isodose line is converted to a structure known as the treated volume. The breast contour
is created by removing the lung and heart (where applicable) overlap from the treated volume contour. The PTV_eval contour is a contraction of 5 mm (in all directions) of
the breast contour. (A) A transverse slice through the isocenter and digitally reconstructed radiograph (DRR) showing the projected (B) medial eld and (C) lateral eld and
showing the PTV_eval, lung, and heart volumes. (Color version of gure is available online.)

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

Table 1
Summary of the statistical analysis of the homogeneity Index, the uniformity Index,
and the conformity index for the PTV_eval for all the patients studied

Age (years)
Homogeneity index
Uniformity index
Conformity index

Minimum

Maximum

Mean

Standard deviation

25.20
0.05
0.46
0.92

93.0
0.15
1.11
1.00

61.56
0.09
1.07
0.98

11.49
0.01
0.03
0.01

compensator technique has been found to result in a decrease in

the HI, maximum dose, lung dose, heart dose, and PTV V105 while
keeping the number of monitor units delivered constant.11 According to Gursel et al,3 the use of the irregular surface compensator

357

algorithm in Eclipse treatment planning system (TPS) is a promising solution for patients with a larger breast size, where the use of
the FnF technique or IMRT alone may be less favorable.3 Other
researchers have investigated the use of IMRT and hybrid (open
beams plus optimized beams) techniques for the treatment of
breast cancer.12-19,22-28 The hybrid treatment planning uses open
and optimized beams to produce a homogeneous dose distribution, and to decrease the effect of respiratory motion, the open
beams are usually assigned weights that are as high as possible,
and the tangential posterior eld edges are matched to spare lung
dose.20,21
There is a lack of consistency among institutions or individuals
on what is considered an acceptable treatment plan: target coverage vis--vis dose to OAR. Clinical trials usually resolve some of

Fig. 2. Plot of (A) the PTV_eval volumes, (B) breast separation, and (C) a plot of PTV_eval against separation for all patients.

358

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

Table 2
A summary of the statistical analysis of the breast separation, PTV_eval volume, and normalized PTV_eval volume receiving 90%, 92%, 95%, 100%, 105%, 107%, 108%, and 110%
of the prescribed dose for all patients. Also shown are the results when patients have been stratied by breast separation into 3 main groups (o 20 cm, Z 20
r 25 cm, and
4 25 cm)
All prescriptions

Breast separation
(cm)

PTV_eval volume
(cm3)

Normalized PTV_eval volume receiving the percentage of the prescribed dose

V90%
(%)

Small (o 20 cm, n 229)

Mean
17.98
Standard
1.52
deviation
Maximum
19.99
Minimum
12.40

All patients (n 623)

Mean
20.99
Standard
3.04
deviation
Maximum
34.91
Minimum
12.4

V92%
(%)

PD

V95%
(%)

PD

V100%
(%)

PD

V105%
(%)

PD

V107%
(%)

PD

V108%
(%)

PD

V110%
(%)

551.33
271.6

99.95
0.08

99.76
0.29

98.33
1.15

76.84
10.28

0.24
0.75

1557.78
73.60

100
99.6

100
98.23

100
92.98

92.89
29.47

4.95

0.02

99.87
0.34

99.58
0.55

98.0
1.38

76.73
7.66

0.48
0.93

100
94.92

100
94.67

99.99
92.14

95.93
45.33

4.81

0.02

1659.69
493.03

99.59
0.32

98.95
0.50

96.28
0.99

73.03
8.15

1.06
1.43

0.06
0.28

0.03
0.13

3684.61
1017.97

99.98
98.74

99.81
97.7

98.13
94.32

86.35
51.53

4.94

1.71

0.83

0.01

884.0
458.07

99.87
0.29

99.59
0.51

97.98
1.38

76.46
8.80

0.44
0.95

0.01
0.08

0.04

3684.61
73.6

100
94.92

100
94.67

100
92.14

95.93
29.47

4.95

1.71

0.83

0.01

Medium (Z 20 cm and r 25 cm, n 342)

Mean
22.09
988.82
Standard
1.37
346.36
deviation
Maximum
24.93
2456.60
Minimum
20.00
129.13
Large (4 25 cm, n 52)
Mean
27.03
Standard
2.03
deviation
Maximum
34.91
Minimum
25.04

PD

PD

Indicates that values are less than 0.01%.

these issues, as the criteria for an acceptable plan within the trial
(target coverage and doses to OAR) are well dened. Any plan
fullling the criteria is considered acceptable, whereas any plan
not fullling all the criteria may be considered unacceptable. In
such cases, there is relatively less stress on treatment planners, as

they can present treatment plans to radiation oncologists, which

are less likely to be rejected. This more likely improves condence
in planners, reduces variation in treatment plans, and improves
workow and patient care. Despite these benets, some institutions have still not developed local institutional criteria for

Table 3
A summary of the statistical analysis of the breast separation, PTV_eval volume, and normalized PTV_eval volume receiving 90%, 92%, 95%, 100%, 105%, 107%, 108%, and 110%
of the prescribed dose for all patients receiving 42.5-Gy dose in 16 fractions. Also shown are the results when patients have been stratied by breast separation into 3 main
groups (o 20 cm, Z 20
r 25 cm, and 4 25 cm)
42.5 Gy/16
Prescription

Breast separation
(cm)

PTV_eval volume
(cm3)

Normalized PTV_eval volume receiving the percentage of the prescribed dose

V90%
(%)

Small (o 20 cm, n 212)

Mean
16.2
Standard deviation
1.52
Maximum
19.99
Minimum
12.4

PD

V92%
(%)

PD

V95%
(%)

PD

V100%
(%)

PD

V105%
(%)

PD

V107%
(%)

PD

V108%
(%)

PD

V110%
(%)

550.93
272.72
1557.78
108.57

99.95
0.08
100
99.6

99.77
0.28
100
98.23

98.36
1.12
100
92.98

76.89
10.32
92.89
29.47

0.23
0.74
4.95

0.02

970.58
331.5
231.20
129.13

99.87
0.34
100
94.92

99.59
0.54
100
94.67

98.05
1.35
99.99
93.1

76.85
7.50
95.93
45.33

0.47
0.91
4.81

0.02

Large (4 25 cm, n 38)

Mean
26.77
Standard deviation
1.61
Maximum
32.60
Minimum
25.28

1468.37
299.51
2313.34
1017.97

99.59
0.32
99.98
98.74

98.95
0.51
99.81
97.7

96.31
0.92
97.91
94.34

73.22
7.19
83.68
53.65

0.86
1.34
4.75

0.01
0.09

Total (n 570)
Mean
Standard deviation
Maximum
Minimum

847.69
402.56
2313.34
108.57

99.88
0.28
100
94.92

99.62
0.49
100
94.67

98.05
1.34
100
92.98

76.62
8.68
95.93
29.47

0.41
0.90
4.95

0.09

Medium (Z 20 cm and r 25 cm, n 320)

Mean
22.09
Standard deviation
1.37
Maximum
24.93
Minimum
20.00

20.88
2.89
32.60
12.4

Indicates that values are less than 0.01%.

PD

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

accepting treatment plans. It was for these reasons that we

developed an institutional criterion for accepting volume-based
breast treatment plans in 2006 after analyzing treatment data of
approximately 200 patients. The purpose of this article is to report
on the dosimetric review of 623 patients treated in the last 18
months to evaluate the effectiveness of the previously developed
breast plan acceptability criteria and any possible changes necessary to further improve patient care.

Methods and Materials

We reviewed the treatment plans of 623 randomly selected patients with
breast cancer who were treated at our cancer center from January 2013 to June
2014. Based on staging criteria, patients are treated with either a prescription dose
of 50 Gy in 25 fractions or a prescription dose of 42.5 Gy in 16 fractions. For the
data reviewed, 53 (8.5%) patients were treated with a prescription dose of 50 Gy in
25 fractions and 570 (91.5%) were treated with a prescription dose of 42.5 Gy in 16
fractions. Based on patient suitability and tolerance, 147 (23.6%) patients were
treated using the DIBH technique and 476 (76.4%) were treated with free-breathing
technique.

359

same level 1.0 cm posterior to palpable breast tissue, and a third BB is placed on the
contralateral side at this same level. A wire is placed to delineate the scar, and
another wire is placed around the entire breast to delineate the edge of palpable
breast tissue. The patient is scanned using the center's breast CT scan protocol,
with a slice thickness of 0.3 cm. The scan extends superiorly from just below the
mandible to inferiorly to include lungs and breast tissue. The scan data are
exported to the virtual simulator, where the laser origin is marked at the location
of the BBs. If the BB location is acceptable for breast tissue coverage and lung
volume, then this point is used for marking. If the BB position is not acceptable,
then a new tattoo point is created and placed as appropriate. This point is then
marked on the patient. The patient is tattooed at the anterior, right lateral, and left
lateral setup points. Additional tattoos can be placed inferiorly for leveling if the
patient's anatomy requires a more stable setup point. Marks are also placed on the
Vac-Lok for straightening (a mark in line with the AP tattoo and lateral marks).

Field borders placement

The superior eld border is set at the level of the second intercostal space but
may be adjusted to clinically cover the entire breast tissue. The inferior border is set
at 1.5 cm inferior to the inframammary fold. The medial border is set at midline and
the lateral border is set at the midaxillary line, or 1.0 cm posterior to palpable
breast tissue.

Patient positioning

Radiation treatment planning

As part of the standard protocol for breast treatment at our institution, the
patients were positioned supine, with both the arms raised above the head. A VacLok cushion is used to support the arms, and a kneex is placed under the knees.
This positioning can be modied if required. If the positioning of the patient results
in breast tissue falling superiorly to the level of the second intercostal space, a
breast board is used at an appropriate inclined angle, and a headrest is placed
under the head. In such a case, the ipsilateral arm is raised above the head, the
contralateral arm is placed by the patient's side, and a kneex remains under the
knees. In the case where a pendulous breast falls too far laterally, resulting in a
large volume of lung being included in the treatment eld, a breast sling may be
used. A sling may also be used if the breast falls inferiorly and creates a fold.

The CT scan data set and all points are exported into the Eclipse (Version 10:
Varian Medical Systems) TPS. The scar is contoured, and the CT value is set to zero
Hounseld units. The isocenter is placed at the International Commission on
Radiation Units and Measurements point at midbreast separation, and the gantry,
collimator, couch, and jaws/multi-leaf collimator of the medial eld are adjusted
such that the entire breast is covered and the lung volume included in the eld is
acceptable. An opposing lateral eld is created from the approved medial eld, and
the lateral gantry angle is adjusted for divergence on the posterior border.

CT simulation
Before CT scan, the second intercostal space is palpated, and a wire is placed on
the patient's skin, midway along the breast at that level. Another wire is also placed
1.5 cm inferior to the inframammary fold, and a small ball bearing (BB) is placed
midline at the level middistance between the 2 wires. A second BB is placed at the

Target volumes and OAR

The body and lung contours are created using an automated contouring feature
of the TPS. To ensure less cleanup and maintain consistent PTV_eval (breast volume
for evaluation) creation, a treated volume (50% isodose line) is rst created by
reducing the eld lengths by 1.0 cm in the superior and inferior directions and
calculating the dose distribution. The 50% isodose line is then converted to a
structure known as the treated volume. The ipsilateral lung and heart (for left-sided
treatment) are contoured. The breast contour is created by removing the lung and

Table 4
A summary of the statistical analysis of the breast separation, PTV_eval volume, and normalized PTV_eval volume receiving 90%, 92%, 95%, 100%, 105%, 107%, 108%, and 110%
of the prescribed dose for all patients receiving 50-Gy dose in 25 fractions. Also shown are the results when patients have been stratied by breast separation into 3 main
groups (o 20 cm, Z 20
r 25 cm, and 4 25 cm)
50 Gy/25 Prescription Breast separation
(cm)

PTV_eval volume
(cm3)

Normalized PTV_eval volume receiving the percentage of the prescribed dose

V90%
(%)

PD

V92%
(%)

PD

V95%
(%)

PD

V100%
(%)

PD

V105%
(%)

PD

V107%
(%)

PD

V108%
(%)

PD

V110%
(%)

Small (o 20 cm, n 17)

Mean
17.86
Standard deviation
1.56
Maximum
19.82
Minimum
14.37

556.23
265.12
1010.98
73.60

99.95
0.09
100
99.66

99.71
0.36
100
98.84

98.04
1.46
99.88
94.85

76.22
9.92
91.75
57.32

0.37
0.86
2.97

Medium (Z 20 cm and r 25 cm, n 22)

Mean
22.07
Standard deviation
1.37
Maximum
24.78
Minimum
20.32

1254.12
447.09
2456.60
567.45

99.76
0.37
100
98.62

99.34
0.68
99.96
97.37

97.25
1.50
99.18
92.14

74.86
9.64
86.41
56.76

0.66
1.21
4.06

0.01

Large (4 25 cm, n 14)

Mean
27.74
Standard
2.82
deviation
Maximum
34.91
Minimum
25.04

1713.82
81.70

99.59
0.33

98.94
0.50

96.20
1.20

72.53
10.62

1.60
1.57

0.21
0.53

0.09
0.25

3684.61
1654.20

99.91
98.84

99.58
97.86

98.13
94.32

86.35
51.53

4.94

1.71

0.83

0.01

Total (n 53)
Mean
Standard deviation
Maximum
Minimum

1274.58
752.78
3684.61
73.60

99.77
0.32
100
98.62

99.36
0.62
100
97.37

97.23
1.56
99.88
92.14

74.68
9.90
91.75
51.53

0.81
1.30
4.94

0.06
0.28
1.71

0.02
0.13
0.83

0.01

22.21
4.23
34.91
14.37

Indicates that values are less than 0.01%.

PD

360

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

heart (where applicable) overlap from the treated volume contour. The PTV_eval
contour is a contraction of 5 mm (in all directions) of the breast contour (Fig. 1).

Hybrid IMRT technique

The hybrid IMRT technique uses both open and optimized beams for
generating the treatment plans of the patients with breast cancer. This technique
has the potential to produce a homogeneous dose distribution. To decrease the
effect of respiratory motion, the open beams are usually assigned weights that are
as high as possible. The superior and inferior eld borders of the open elds are

now reset to their original positions, and the dose distribution is recalculated using
the desired energy. The prescription of this open-eld plan is set to 70% of the
prescription dose (i.e., 3000c-Gy dose for a prescription dose of 4250 Gy in 16
fractions or 3530c-Gy dose for a prescription dose of 5000c Gy in 25 fractions). The
plan is normalized to the isocenter or a normalization point (if required). The
open-eld plan is copied to create a plan for optimization. We usually use lowenergy beams for the optimized eld, and the prescription is set to the remaining
30% of the prescription dose (i.e., 1250c-Gy dose for prescription dose of 4250c Gy
or 1470c-Gy dose for a prescription dose of 5000c Gy). The plan is optimized using
the open-eld plan as a base dose plan and is calculated using xed jaws and
sliding window settings in the TPS. This optimized plan is also normalized in the

Fig. 3. A plot of the dose-volume histograms (DVHs) for all PTV_eval for patients treated with (A) 42.5-Gy dose in 16 fractions with regular breathing, (B) 42.5-Gy dose in 16
fractions with the deep inspiration breath-hold (DIBH) technique, and (C) 50-Gy dose in 25 fractions with the regular-breathing technique. Also shown are the mean DHVs.
(Color version of gure is available online.)

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

same manner as the open-eld plan. The open-eld and the optimized plans
are then merged to create the nal plan. The PTV_eval DVH is evaluated (the
merged plan may be renormalized if necessary) to ensure that the plan meets the
following criteria for the PTV_eval: V92% PD 4 99%, V95% PD 4 95%, and V105% PD
o 1%.
Indices for PTV_eval
In addition to the evaluation of the DVH for the PTV_eval, the HI, UI, and the CI
are also calculated for the PTV_eval. The HI, CI, and UI evaluate the dose
homogeneity, conformity, and uniformity, respectively, within the PTV_eval and
are calculated as
D2 D98
HI
DPD
D5
UI
D95
VRI
CI
TV
where D2, D5, D95, and D98 are the doses received by 2%, 5%, 95%, and 98% of the
PTV_eval, respectively. DPD is the prescribed dose, VRI is the volume of PTV_eval
covered by the reference isodose line (in this case the 95% isodose line), and TV is
the target volume (in this case the PTV_eval). The values of CI and UI close to unity
indicate greater conformity and uniformity and values of HI close to zero indicate
greater homogeneity.

Results and Discussions

We have evaluated the dose distribution of the treatment
plans of 623 patients with breast cancer treated at our center
from January 2012 to June 2014 to assess the efcacy of a breast
plan criterion that was already developed and institutionally
acceptable (target coverage and doses to OAR). The Eclipse TPS
version 10.0 (Varian Medical Systems, Palo Alto, CA) and the
analytical anisotropic algorithm dose calculation algorithm version 10.0 were used for all treatment plans dose calculations. All
the patients were randomly selected. For each patient, we
determined the age; breast separation; and volumes of the
PTV_eval, ipsilateral lung, and heart (for patients whose leftsided breast was treated for which the heart is usually
contoured).
For each treatment plan dose distribution, we determined the
mean, maximum, minimum, and standard deviation of the
PTV_eval, ipsilateral lung, and heart and analyzed the DVHs for
all the structures. We also calculated the HI, UI, and the CI for the
PTV_eval. Table 1 shows the summary of the statistical analysis of
the HI, UI, and the CI for all 623 patients' PTV_eval structures. The

361

mean patient age was 61.6 years (range: 25.2 to 93.0 years).
Overall, 314 (50.4%) patients had the disease in the left breast
and 309 (49.6%) had it in the right breast. Based on patient
suitability and tolerance, 147 (23.6%) patients were treated using
the DIBH technique and all others were treated with the freebreathing technique.
PTV_eval dose evaluation
Figure 1 displays the structure segmentation showing the
PTV_eval, lung, and heart volumes. Figure 2 shows a plot of the
PTV_eval for all the patients (Fig. 2A), a plot of breast separations
for all the patients (Fig. 2B), and a plot of PTV_eval against breast
separation (Fig. 2C) for all patients. Tables 2 to 4 show a summary
of the statistical analysis of breast separations, PTV_eval, and dosevolumetric analysis for the PTV_eval for all patients (Table 2).
Tables 3 and 4 show the dose-volumetric analysis for the PTV_eval
when the patients were grouped into those treated with 42.5-Gy
dose in 16 fractions (Table 3) and those treated with 50-Gy dose in
25 fractions (Table 4). Also shown in Tables 2 to 4 are similar
statistical analysis after stratifying the patients by breast separation into 3 main groups: small (o 20 cm), medium (20 r x r 25),
and large (4 25 cm). We stratied the patient population into
these 3 different breast separations and prescription doses and
analyzed the PTV_eval dosimetry to assess if different criteria are
required for each subgroup.
The mean breast separation for all the patients was 21.0 cm
(range: 12.4 to 34.9 cm). The mean breast separation for the smallbreast group was 18.0 cm (range: 12.4 to 20.0 cm), the mediumbreast group was 22.1 cm (range: 20.0 to 24.9 cm), and the largebreast group was 27.0 cm (range: 25.0 to 34.9 cm). The corresponding mean PTV_eval for all patients was 884.0 cm3 (range:
73.6 to 3684.6 cm3), small-breast group was 551.3 cm3 (range: 73.6
to 1557.8 cm3), medium-breast group was 988.8 cm3 (range: 129.1
to 2456.6 cm3), and large-breast group was 1659.7 cm3 (range:
1018.0 to 3684.6 cm3). The mean normalized PTV_eval receiving at
least 92% (V92% PD) and 95% (V95% PD) of the prescribed dose are all
more than 99% and 97%, respectively, for all groups. Mean
normalized PTV_eval receiving at least 105% (V105% PD) of the
prescribed dose are less than 1% for all groups. The mean HI, UI,
and CI for the PTV_eval are 0.09 (range: 0.05 to 0.15), 1.07 (range:
0.46 to 1.11), and 0.98 (range: 0.92 to 1.0), respectively. A plot of
the DVHs for all PTV_eval for all patients for both 42.5-Gy dose in

Table 5
A summary of the statistical analysis of the ipsilateral lung volumetric doses for all patients. Data have been separated into patients who were treated with the free-breathing
technique and those treated with the deep inspiration breath-hold technique. nFB and nDIBH are the number of patients being treated with the free-breathing and the deep
inspiration breath-hold techniques, respectively. The p value analysis of the tabulated results for the various doses when comparing the DIBH and the free-breathing
techniques is also shown in the Table
Free-breathing technique (n 476)

Deep inspiration breath-hold (DIBH) technique (n 147)

p Value

Minimum

Maximum

Mean

Standard deviation

Minimum

Maximum

Mean

Standard deviation

598.20

2737.59

1383.37

346.45

892.65

3091.41

2127.48

394.65

Ipsilateral lung dose parameters for 42.5-Gy dose in 16 fractions prescription dose (nFB 431 and nDIBH 139)
2.42
32.37
18.02
4.57
10.86
29.95
V5 Gy (%)
V10 Gy (%)
1.37
22.72
11.48
3.49
3.76
19.15
1.04
19.81
9.45
3.15
2.75
15.98
V15 Gy (%)
V20 Gy (%)
1.31
18.43
8.42
2.94
2.23
14.55
V30 Gy (%)
0.71
15.93
6.5
2.59
1.36
11.92
Maximum lung dose (%PD) (%)
90.48
103.88
97.67
2.35
90.33
100.96

18.58
11.07
9.04
8.05
6.22
96.03

3.85
2.86
2.52
2.35
2.00
2.15

0.200
0.218
0.159
0.201
0.236
o 0.001

Ipsilateral lung dose parameters for 50-Gy dose in 25 fractions prescription dose: The free-breathing technique (nFB 45 and nDIBH 8)
V5 Gy (%)
9.28
30.5
21.57
5.15
16.31
26.57
20.72
3.71
3.89
20.2
13.09
3.95
8.19
15.23
12.13
2.31
V10 Gy (%)
V15 Gy (%)
2.49
16.91
10.31
3.48
6.06
12.05
9.48
1.93
V20 Gy (%)
1.93
15.41
8.89
3.22
5.31
10.76
8.40
1.75
1.36
13.48
7.30
2.96
4.33
9.25
6.99
1.62
V30 Gy (%)
Maximum lung dose (%PD) (%)
92.19
101.81
97.37
2.12
93.19
101.55
96.36
2.94

0.660
0.511
0.520
0.681
0.776
0.249

Lung volume (cc)

362

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

16 fractions with free-breathing, DIBH technique and 50-Gy dose

in 25 fractions with free-breathing technique are shown in Fig. 3,
respectively. The mean PTV_eval DVH is also shown in Fig. 3.
PTV_eval DVHs for new treatment plans can be evaluated along
these mean DVH values to ensure consistency in treatment plans.
Based on current data, it is possible to develop breast treatment
plans that achieve dose coverage for the PTV_eval volume of 92%
(V92% PD) and 95% (V95% PD) of the prescribed dose to at least 99%
and 97% of the normalized volume, respectively, while at the
same time, restricting the normalized volume of the PTV_eval
receiving a hot spot of 105% (V105% PD) of the prescribed dose to
less than 1%.

Ipsilateral lung dose evaluation

The dose to the ipsilateral lung (i.e., the main OAR for breast
and chest wall radiation treatment) was indexed by the percentage volume of ipsilateral lung receiving at least doses of 5 (V5 Gy),
10 (V10 Gy), 15 (V15 Gy), 20 (V20 Gy), and 30 Gy (V30 Gy). Table 5
shows the dose-volumetric analysis of the ipsilateral lung comparing patients treated with 42.5-Gy dose in 16 fractions and 50Gy dose in 25 fractions. Table 5 is further subdivided into
patients treated using the free-breathing and the DIBH techniques. The mean ipsilateral lung volume for patients treated with
the free-breathing and the DIBH techniques were 1383.4 cm3

Fig. 4. A plot of the dose-volume histograms (DVHs) for the ipsilateral lung for all patients treated with (A) 42.5-Gy dose in 16 fractions with regular breathing, (B) 42.5-Gy
doses in 16 fractions with the deep inspiration breath-hold (DIBH) technique, and (C) 50-Gy dose in 25 fractions with the regular-breathing technique. Also shown are the
mean DHVs. (Color version of gure is available online.)

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

(range: 598.2 to 2737.6 cm3) and 2127.5 cm3 (range: 892.7 to

3091.4 cm3), respectively. The p value analyses of the results for
the V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy for the lung when
comparing the DIBH and free-breathing techniques are all more
than 0.05 (i.e., p 4 0.05). This indicates that there is no
signicant difference in lung dose between the free-breathing
and the DIBH techniques. The DVHs for the ipsilateral lung for all
patients including the mean DVH are shown in Fig. 4. Lung DVHs
for new treatment plans can be evaluated along these mean DVH
values to ensure consistency in treatment plans. Based on current
data, breast treatment planning can aim at achieving the mean
values of V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy for the lung.
Heart dose evaluation
The main OAR for patients being treated with the disease in the
left breast is the heart. Data are only available for patients who had
the diseases in the left breast, because normally, it is only for those
patients that the heart is contoured for planning. The dose to the
heart was also indexed by the percentage volume of ipsilateral
heart receiving at least doses of 5 (V5 Gy), 10 (V10 Gy), 15 (V15 Gy),
20 (V20 Gy), and 30 Gy (V30 Gy). Table 6 shows the dose-volumetric
analysis of the heart comparing patients treated with 42.5-Gy dose
in 16 fractions and 50-Gy dose in 25 fractions. Table 6 is further
subdivided into patients treated with the free-breathing and the
DIBH techniques. The mean heart volume for patients treated with
the free-breathing and the DIBH techniques is 606.5 cm3 (range:
353.3 to 999.5 cm3) and 543.7 cm3 (range: 313.0 to 1169.7 cm3),
respectively. Our results conrm the signicant advantage of using
the DIBH technique to reduce heart dose when compared with the
free-breathing technique. The p value analyses of the results for
the V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy for the heart comparing
the DIBH and the free-breathing techniques are well less than 0.05
(i.e., p o 0.05). The DVHs for the heart for all patients including the
mean DVH are shown in Fig. 5. Heart DVHs for new treatment
plans can be evaluated along with these mean DVH values to
ensure consistency in treatment plans. Based on current data,
breast treatment planning can aim at achieving the mean values of
V5 Gy, V10 Gy, V15 Gy, V20 Gy, and V30 Gy for the heart.
The current data indicate that using the hybrid IMRT technique for tangential intact breast radiation therapy can result in

363

smaller hot spots while maintaining improved dose coverage

and dose uniformity throughout the whole breast. This has also
been demonstrated by other groups22-28 using similar techniques.
Hot spots and dose inhomogeneity can lead to poor cosmetic
outcomes and more skin reactions, especially in women with
larger breasts. One such reactionmoist desquamationhas been
correlated with increased pain and reduction in quality of life.22
Kestin et al.12 used IMRT to improve dose uniformity and
potentially to reduce acute toxicity with tangential wholebreast radiotherapy. They concluded that the use of intensitymodulation techniques for tangential breast radiotherapy is an
efcient and effective method for achieving uniform dose
throughout the breast, and it is dosimetrically superior to treatment techniques that use wedges. They found lower overall areas
of hot spots and lower maximum dose with IMRT. Abo-Madyan
et al.23 evaluated intensity-modulated irradiation and other
techniques by comparing dose homogeneity, target coverage,
feasibility, and dosimetric reliability in patients with large
breasts. They observed that aperture-based IMRT using 2 tangential beams with inverse optimization provides a better alternative
to the standard wedged tangential beams for patients with large
breasts while maintaining the efciency of the treatment planning and delivery process.
Pignol et al.22 conducted a multicenter, double-blind, randomized clinical trial to test if breast IMRT would reduce the rate of
acute skin reaction, decrease pain, and improve quality of life
when compared with standard radiotherapy using wedges. They
concluded that breast IMRT technique signicantly reduced the
occurrence of moist desquamation compared with the standard
wedged technique presumably because of lower hot spots and
better homogeneity. We do not use hard wedges at our center for
breast radiotherapy, as it has been recognized that the use of hard
wedges increases scatter to the contralateral breast, which leads to
an increased risk of radiation malignancies.1 Bhatnagar et al.24
compared the dose received by the contralateral breast during
breast irradiation using IMRT and conventional tangential wedged
eld techniques and concluded that breast irradiation with tangential IMRT technique signicantly reduces the dose to the
contralateral breast when compared with conventional tangential
wedged eld techniques. Ludwig et al.26 compared 2 irradiation
techniques for whole-breast irradiationtangential wedged beams

Table 6
A summary of the statistical analysis of the heart volumetric doses for all patients. The data for the heart are only for those patients whose left breast was treated. Data have
been separated into patients who were treated with the free-breathing technique and those treated with the deep inspiration breath-hold technique. nFB and nDIBH are the
number of patients being treated with the free-breathing and the deep inspiration breath-hold techniques, respectively. The p value analysis of the tabulated results for the
various doses when comparing the DIBH and the free-breathing techniques is also shown in the Table
Free-breathing technique (n 167)

Heart volume (cc)

Deep Inspiration breath-hold technique (DIBH) (n 147)

Minimum

Maximum

Mean

Standard deviation

Minimum

Maximum

Mean

Standard deviation

353.34

999.51

606.52

98.75

312.98

1169.68

543.66

109.41

Heart dose parameters for 42.5-Gy dose in 16 fractions prescription dose (nFB 150 and nDIBH 139)
V5 Gy (%)

10.16
2.37
1.76

7.68
1.13
1.17

V10 Gy (%)
V15 Gy (%)

6.92
0.81
0.98

V20 Gy (%)

6.35
0.62
0.85

V30 Gy (%)

5.21
0.35
0.63

Maximum heart dose (%PD)

8.35
101.19
79.06
26.67
7.67

5.42
3.55
2.72
2.14
1.21
103.47

0.52
0.14
0.09
0.06
0.02
31.45

Heart dose parameters for 50-Gy dose in 25 fractions prescription dose: The free-breathing technique (nFB 17 and nDIBH 8)
0.64
5.95
2.53
1.48
0.15
1.30
0.54
V5 Gy (%)
V10 Gy (%)

1.59
0.70
0.50

0.21
0.03
V15 Gy (%)

0.95
0.37
0.31

0.10
0.01

0.76
0.23
0.22

0.06
0.01
V20 Gy (%)
V30 Gy (%)

0.47
0.09
0.12

0.02

Maximum heart dose (%PD)

15.49
96.09
75.58
22.85
13.40
86.34
26.15
Indicates that values are less than 0.01%.
n

The mean values for the V30 Gy are so small that a calculated p value is not reliable.

p Value

0.84
0.44
0.32
0.24
0.12
27.27

o0.001
o0.001
o0.001
o0.001
o0.001
o0.001

0.39
0.07
0.04
0.02
0.01
24.63

0.001
0.001
0.003
0.008
*

o 0.001

364

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

Fig. 5. A plot of the dose-volume histograms (DVHs) for the heart volumes for all patients treated with (A) 42.5-Gy dose in 16 fractions with regular breathing, (B) 42.5-Gy
dose in 16 fractions with the deep inspiration breath-hold (DIBH) technique, and (C) 50-Gy dose in 25 fractions with the regular-breathing technique. Also shown are the
mean DHVs. (Color version of gure is available online.)

vs open elds (without wedges) with forward planned segments

and indicated that partial volume segments can replace wedges
for improved dose coverage and homogeneity in the PTV. Our data
indicate very low volumes of V105%, V107%, V108%, and V110% using
the hybrid IMRT technique, and although we have not correlated
the dosimetry with clinical outcomes, the advantages of improved
dose homogeneity in breast radiotherapy has been demonstrated
by several authors.12,13,22,25 According to Vicini et al.,13 the breast
volume receiving 105% (V105%) and 110% (V110%) of the prescribed

dose is signicantly associated with an increased risk of developing acute skin toxicity. They observed that for patients with V110%
o 200 cm,3 the risk of developing grade II or grade III acute skin
toxicity was 31% vs 61%, respectively, in patients with V110% Z
200 cm.3 The use of IMRT in the treatment of the whole breast
results in a signicant decrease in acute dermatitis, edema,
and hyperpigmentation and a reduction in the development of
chronic breast edema compared with conventional wedge-based
radiotherapy.25

E. Osei et al. / Medical Dosimetry 40 (2015) 355365

Our treatment planning process including beam alignment,

collimation, and optimization takes approximately 20 to 30
minutes. The planner focuses on examining the 3-D dose distribution including the DHVs and making minor adjustments, if
necessary. Treatment delivery, including setup, takes approximately 10 to 15 minutes on both the Varian TrueBeam and EX
treatment machines (Varian Oncology Systems) equipped with
120-millennium MLC. It is therefore possible to standardize the
implementation of this technique and signicantly improve the
quality of radiotherapy delivery to all patients receiving intact
breast radiotherapy without a negative effect on clinic time and
economic resources. In addition, acute toxicity can in theory be
reduced through a reduction in the breast volume receiving high
doses.12,13 Vicini et al.13 presented their clinical experience with an
intensity-modulated technique to improve dose uniformity and
treatment efcacy in patients with early-stage breast cancer and
concluded that the intensity-modulation technique is an efcient
method for achieving a uniform and standardized dose throughout
the whole breast. They also indicated that a widespread implementation of this technology can be achieved with minimal
imposition on clinic resources and time constraints.

Conclusion
The use of hybrid IMRT technique for tangential intact breast
radiation therapy is an efcient and reliable method for achieving
dose uniformity throughout the whole breast. Predened dosevolume constraints and objectives can be achieved, resulting in
improved dose coverage of target breast tissue, reduction in breast
volume receiving high doses and dose to adjacent normal tissue,
and therefore with the potential to reduce the rate of acute skin
reaction, decrease pain, and improve quality of life.22 Based on
current data, it is possible to develop breast treatment plans that
achieve dose coverage for the PTV_eval volume of 92% (V92% PD)
and 95% (V95% PD) of the prescribed dose to at least 99% and 97% of
the normalized volume, respectively, while at the same time,
restricting the normalized volume of the PTV_eval receiving a
hot spot of 105% (V105% PD) of the prescribed dose to less than 1%.
Treatment planning of intact breast should also aim at minimizing
doses to both the heart and the lung. The clinical implementation
of this technology for patients with breast cancer can be achieved
with minimal or no imposition on resources and time constraints.

Acknowledgment
We gratefully acknowledge the support from all the treatment
planners and radiation therapists at the cancer center for their
dedication and commitment to patient care and ensuring a smooth
implementation of the hybrid IMRT technology.
References
1. Asbury K. 3D versus hybrid IMRT breast treatment planning: data, tips, and
techniques. University of Maryland School of Medicine, 2012.
2. Falahatpour, Z.; Aghamiri, S.; Anbiaee, R. External radiotherapy of intact breast:
a comparison between 2D (single CT-slice) and 3D (full CT-slices) plans. Int. J.
Radiat. Res. 9(2):1215; 2011.
3. Gursel, B.; Meydan, D.; Ozbek, N.; et al. Dosimetric comparison of three
different external beam whole breast irradiation techniques. Adv. Ther. 28
(12):111425; 2011.

365

4. Onal, C.; Sonmez, A.; Arslan, G.; et al. Dosimetric comparison of the eld-ineld technique and tangential wedged beams for breast irradiation. Jpn. J.
Radiol. 30(3):21826; 2012.
5. Sun, L.; Meng, F.; Yang, T.; et al. Field-in-eld plan does not improve the
dosimetric outcome compared with the wedged beams plan for breast cancer
radiotherapy. Med. Dosim. 39(1):7982; 2014.
6. Yao, W. A two-point scheme for optimal breast IMRT treatment planning. J.
Appl. Clin. Med. Phys. 14(6):4525; 2013.
7. Yavas, G.; Yavas, C.; Acar, H. Dosimetric comparison of whole breast radiotherapy using eld in eld and conformal radiotherapy techniques in early
stage breast cancer. Int. J. Radiat. Res. 10(3-4):1318; 2012.
8. van der Laan, H.; Hurkmans, C.; Kuten, A.; et al. Current technological clinical
practice in breast radiotherapy; results of a survey in EORTC-radiation oncology
group afliated institutions. Radiother. Oncol. 94(3):2805; 2010.
9. Tanaka, H.; Hayashi, S.; Ohtakara, K.; et al. Impact of respiratory motion on
breast tangential radiotherapy using the eld-in-eld technique compared to
irradiation using physical wedges. Radiother. Oncol. 48(1):948; 2014.
10. Emmens, D.; James, H. Irregular surface compensation for radiotherapy of the
breast: correlating depth of the compensation surface with breast size and
resultant dose distribution. Br. J. Radiol. 83(986):15965; 2010.
11. Hideki, F.; Nao, K.; Hiroyuki, H.; et al. Improvement of dose distribution with
irregular surface compensator in whole breast radiotherapy. J. Med. Phys. 38
(3):1159; 2013.
12. Kestin, L.; Sharpe, M.; Frazier, R.; et al. Intensity modulation to improve dose
uniformity with tangential breast radiotherapy: initial clinical experience. Int. J.
Radiat. Oncol. Biol. Phys. 48(5):155968; 2000.
13. Vicini, F.; Sharpe, M.; Kestin, L.; et al. Optimizing breast cancer treatment
efcacy with intensity-modulated radiotherapy. Int. J. Radiat. Oncol. Biol. Phys.
54(5):133644; 2002.
14. Chin, L.; Cheng, C.; Siddon, R.; et al. Three-dimensional photon dose distributions with, and without lung corrections for tangential breast intact treatments.
Int. J. Radiat. Oncol. Biol. Phys. 17(6):132735; 1989.
15. Buchholz, T.; Gurgoze, E.; Bice, W.; et al. Dosimetric analysis of intact breast
irradiation in off-axis planes. Int. J. Radiat. Oncol. Biol. Phys. 39(1):2617; 1997.
16. Cheng, C.; Das, I.; Baldassarre, S. The effect of the number of computed
tomographic slices on dose distributions and evaluation of treatment planning
systems for radiation therapy of intact breast. Int. J. Radiat. Oncol. Biol. Phys. 30
(1):18395; 1994.
17. Solin, L.; Chu, J.; Sontag, M.; et al. Three-dimensional photon treatment
planning of the intact breast. Int. J. Radiat. Oncol. Biol. Phys. 21(1):193203;
1991.
18. Fraass, B.; Lichter, A.; McShan, D.; et al. The inuence of lung density
corrections on treatment planning for primary breast cancer. Int. J. Radiat.
Oncol. Biol. Phys. 14(1):17990; 1988.
19. Das, I.; Cheng, C.; Fosmire, H.; et al. Tolerances in setup and dosimetric errors in
the radiation treatment of breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 26
(5):88390; 1993.
20. Jassem, J. Favourable and unfavourable effects on long-term survival of radiotherapy for early breast cancer: an overview of the randomized trials. Lancet
355:175770; 2000.
21. Canney, P.; Deehan, C.; Glegg, M.; et al. Reducing cardiac dose in post-operative
irradiation of breast cancer patients: the relative importance of patient
positioning and CT scan planning. Br. J. Radiol. 72(862):98693; 1999.
22. Pignol Jean, Philippe; Olivotto, Ivo; Rakovitch, Eileen; et al. A multicenter
randomized trial of breast intensity modulated radiation therapy to reduce
acute radiation dermatitis. J. Clin. Oncol. 26(13):208592; 2008.
23. Abo-Madyan, Yasser; Polednik, Martin; Angelika, Rahn; et al. Improving dose
homogeneity in large breasts by IMRT efcacy and dosimetric accuracy of
different techniques. Strahlenther. Onkol. 2008(184):8692; 2008.
24. Bhatnagar, A.K.; Brander, E.; Sonnik, D.; et al. Intensity modulated radiation
therapy (IMRT) reduces the dose to the contralateral breast when compared to
conventional tangential elds for primary breast irradiation. Breast Cancer Res.
Treat. 96(41):416; 2006.
25. Harsolia, Asif; Kestin, Larry; Grills, Inga; et al. Intensity-modulated radiotherapy
results in signicant decrease in clinical toxicities compared with conventional
wedge-based breast radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 68(5):137580;
2007.
26. Ludwig, Veronika; Schwab, Franz; Guckenberger, Matthias; et al. Comparison of
wedge versus segmented techniques in whole breast irradiation. Strahlenther.
Onkol. 184(6):30712; 2008.
27. Charles, Mayo S.; Marcia, Urie M.; Thomas, Fitzgerald J. Hybrid IMRT plans
concurrently treating conventional and IMRT beams for improved breast
irradiation and reduced planning time. Int. J. Radiat. Oncol. Biol. Phys. 61
(3):92232; 2005.
28. Wendy, Smith; Geetha, Menon; Nathan, Wolfe; et al. IMRT for the breast: a
comparison of tangential planning techniques. Phys. Med. Biol. 55:123141;
2010.