Dosimetric effects of multileaf collimator leaf width on intensity-modulated

radiotherapy for head and neck cancer

Chae-Seon Hong, Sang Gyu Ju, Minkyu Kim, Jung-in Kim, Jin Man Kim, Tae-Suk Suh, Youngyih Han, Yong
Chan Ahn, Doo Ho Choi, Heerim Nam, and Hee Chul Park

Citation: Medical Physics 41, 021712 (2014); doi: 10.1118/1.4860155

View online: http://dx.doi.org/10.1118/1.4860155
View Table of Contents: http://scitation.aip.org/content/aapm/journal/medphys/41/2?ver=pdfcov
Published by the American Association of Physicists in Medicine

Articles you may be interested in

Monte Carlo simulation based study of a proposed multileaf collimator for a telecobalt machine
Med. Phys. 40, 021705 (2013); 10.1118/1.4773308

Dosimetric investigation of breath-hold intensity-modulated radiotherapy for pancreatic cancer

Med. Phys. 39, 48 (2012); 10.1118/1.3668314

In vivo verification of superficial dose for head and neck treatments using intensity-modulated techniques
Med. Phys. 36, 59 (2009); 10.1118/1.3030951

Energy modulated electron therapy using a few leaf electron collimator in combination with IMRT and 3D-CRT:
Monte Carlo-based planning and dosimetric evaluation
Med. Phys. 32, 2976 (2005); 10.1118/1.2011089

The effects of intra-fraction organ motion on the delivery of intensity-modulated field with a multileaf collimator
Med. Phys. 30, 1736 (2003); 10.1118/1.1578771
Dosimetric effects of multileaf collimator leaf width on intensity-modulated
radiotherapy for head and neck cancer
Chae-Seon Hong, Sang Gyu Ju,a) and Minkyu Kim
Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine,
Seoul 135-710, South Korea

Jung-in Kim
Department of Radiation Oncology, Kangbuk Samsung Hospital, Sungkyunkwhan University School of
Medicine, Seoul 110-746, South Korea
Jin Man Kim
Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine,
Seoul 135-710, South Korea

Tae-Suk Suhb)
Department of Biomedical Engineering and Research Institute of Biomedical Engineering, The Catholic
University of Korea, Seoul 137-701, South Korea
Youngyih Han, Yong Chan Ahn, and Doo Ho Choi
Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine,
Seoul 135-710, South Korea

Heerim Nam
Department of Radiation Oncology, Kangbuk Samsung Hospital, Sungkyunkwhan University School of
Medicine, Seoul 110-746, South Korea
Hee Chul Park
Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine,
Seoul 135-710, South Korea

(Received 23 August 2013; revised 12 December 2013; accepted for publication 13 December 2013;
published 15 January 2014)
Purpose: The authors evaluated the effects of multileaf collimator (MLC) leaf width (2.5 vs. 5 mm)
on dosimetric parameters and delivery efficiencies of intensity-modulated radiation therapy (IMRT)
and volumetric-modulated arc therapy (VMAT) for head and neck (H&N) cancers.
Methods: The authors employed two types of mock phantoms: large-sized head and neck (LH&N)
and small-sized C-shape (C-shape) phantoms. Step-and-shoot IMRT (S&S_IMRT) and VMAT treat-
ment plans were designed with 2.5- and 5.0-mm MLC for both C-shape and LH&N phantoms. Their
dosimetric characteristics were compared in terms of the conformity index (CI) and homogeneity
index (HI) for the planning target volume (PTV), the dose to organs at risk (OARs), and the dose-
spillage volume. To analyze the effects of the field and arc numbers, 9-field IMRT (9F-IMRT) and
13-field IMRT (13F-IMRT) plans were established for S&S_IMRT. For VMAT, single arc (VMAT1 )
and double arc (VMAT2 ) plans were established. For all plans, dosimetric verification was performed
using the phantom to examine the relationship between dosimetric errors and the two leaf widths.
Delivery efficiency of the two MLCs was compared in terms of beam delivery times, monitor units
(MUs) per fraction, and the number of segments for each plan.
Results: 2.5-mm MLC showed better dosimetric characteristics in S&S_IMRT and VMAT for C-
shape, providing better CI for PTV and lower spinal cord dose and high and intermediate dose-
spillage volume as compared with the 5-mm MLC (p < 0.05). However, no significant dosimetric
benefits were provided by the 2.5-mm MLC for LH&N (p > 0.05). Further, beam delivery efficiency
was not observed to be significantly associated with leaf width for either C-shape or LH&N. However,
MUs per fraction were significantly reduced for the 2.5-mm MLC for the LH&N. In dosimetric
error analysis, absolute dose evaluations had errors of less than 3%, while the Gamma passing rate
was greater than 95% according to the 3%/3 mm criteria. There were no significant differences in
dosimetric error between the 2.5- and 5-mm MLCs.
Conclusions: As compared with MLC of 5-mm leaf widths, MLC with finer leaf width (2.5-mm)
can provide better dosimetric outcomes in IMRT for C-shape. However, the MLC leaf width may
only have minor effects on dosimetric characteristics in IMRT for LH&N. The results of the present

021712-1 Med. Phys. 41 (2), February 2014 0094-2405/2014/41(2)/021712/10/$30.00 © 2014 Am. Assoc. Phys. Med. 021712-1
021712-2 Hong et al.: Dosimetric effects of MLC leaf width on IMRT
study will serve as a useful assessment standard when assigning or introducing equipment
for the treatment of H&N cancers. © 2014 American Association of Physicists in Medicine.
[http://dx.doi.org/10.1118/1.4860155]
Key words: multileaf collimator, leaf width, IMRT, VMAT, head and neck cancer
1. INTRODUCTION
Because of their anatomical location, head and neck cancers
are often in close proximity to especially radiation-sensitive
normal organs such as the spinal cord, parotid gland, and eye-
balls. Consequently, it is difficult to minimize the dose to
these organs while delivering a curative dose to the tumor,
using three-dimensional conformal radiotherapy alone.1–5 In
this regard, intensity-modulated radiation therapy (IMRT) is
an especially effective treatment technique, which makes it
easier to improve the conformity of the dose to the tumor
through beam intensity modulation, thereby reducing the dose
to the surrounding normal organs. IMRT plays a very impor-
tant role in radiation therapy for head and neck cancers, and
various IMRT techniques have been investigated and used in
clinical practice.6–9
The multileaf collimator (MLC) has two main purposes in
IMRT. First, it is used to shield normal organs from unnec-
essary radiation. Second, it is used to modulate the beam in-
tensity for IMRT. Specifically, MLCs are used to create small
segments (beamlets) with different beam intensities, which,
when overlapped, yield the desired intensity map. Therefore,
the width of MLC leaves is not only important for shielding
resolution at the target margin but also for the intensity map
resolution within the target volume. Following the invention
of the MLC system, a variety of different forms of MLCs have
also been developed and used in clinical practice.10–12
Typically, MLCs with fine (≤3 mm) widths permit more
precise beam shaping (allowing a better match of the beam
aperture to the target projection) and high-resolution beam
optimization. Despite these advantages, MLCs with fine leaf
widths also have disadvantages that limit their clinical ap-
plication. The MLC apparatus is composed of several dozen
leaves as well as motors and controls that move each of the
leaves. MLC systems with fine leaf widths naturally require
more leaves and motors. However, due to space constraints
within the gantry head, the possible numbers of leaves and
motors is limited. Consequently, the maximum field size that
the MLC can cover simultaneously is typically restricted to
22 cm (HDMLC, Varian) in the longitudinal direction, de-
pending on the specific type of machine. During the treatment
of head and neck cancer, the 22-cm field length is often ex-
ceeded when the lower neck region is involved. Therefore,
the limitations on field size are important points of consider-
ation when selecting equipment for treatment or introducing
new equipment for clinical use.
For the treatment of a tumor of a given size, the use of
MLC with fine leaf width leads to a relatively greater num-
ber of MLC leaf junctions included within the field than the
use of MLC with thick leaf width. This increases the possi-
bility of increased dose uncertainty due to greater interleaf
Medical Physics, Vol. 41, No. 2, February 2014
021712-2
leakage and tongue-and-groove effects that occur at the MLC
leaf junctions.13–15
From this perspective, quantitative assessment of the dosi-
metric effects of MLC leaf width in IMRT is essential. In-
deed, several researchers have previously published studies
that are related to this topic. MLC with finer leaf width was
found to enhance target dose conformity and reduce the dose
to the surrounding normal tissue in IMRT for a small-volume
target in the brain.16–20 However, MLC with finer leaf width
was not always effective in IMRT for head and neck cancers
when the target sizes were relatively large.21, 22 Because the
exact findings varied across these previous studies, it has been
difficult to clearly identify the effects.23–25 Moreover, there
has been little research on the effects of MLC leaf width on
volumetric-modulated arc therapy (VMAT), an increasingly
common approach to treatment.26 In addition, prior reports
have not evaluated large-volume targets, such as the head and
neck. Finally, the present study is the first to attempt a dose-
uncertainty analysis according to the MLC leaf width using a
phantom. Therefore, this research may provide useful guide-
lines for the introduction of new equipment into practice as
well as for the selection of equipment in current clinical prac-
tice. More specifically, the present study investigated step-
and-shoot IMRT (S&S_IMRT) and VMAT using a head and
neck phantom to analyze the dependence of dosimetric effects
and treatment efficiencies on MLC leaf width.
2. MATERIALS AND METHODS
To evaluate the dosimetric effects of MLC leaf width
on IMRT for head and neck cancers, two experiments each
were conducted for S&S_IMRT and VMAT using a 2.5-mm
MLC (HD120MLC, Novalis Tx, Varian) and a 5-mm MLC
(M120MLC, iX, Varian). The 2.5-mm MLC consisted of
60 pairs of leaves, comprising 32 inner pairs of 2.5 mm, 26
outer pairs of 5 mm, and 2 outermost pairs of 7 mm, at the
isocenter.15 First, treatment plans were designed for the re-
spective conditions, and their dosimetric characteristics were
compared. Second, the treatment plans were implemented, us-
ing the phantom to measure the absolute point dose and sagit-
tal dose distributions. Dosimetric error was assessed by com-
paring measurements with calculated data from the treatment
planning system (TPS). Additionally, the beam delivery ef-
ficiencies of both MLCs were compared in terms of deliv-
ery time, number of segments, and monitor units (MUs) per
fraction.
2.A. Phantom design
We employed a custom-made cylindrical type phantom,
which represents a head and neck cancer patient, with
021712-3 Hong et al.: Dosimetric effects of MLC leaf width on IMRT
F IG . 1. Axial slices of both phantoms with the PTV and OARs: (a) C-shape
phantom with the PTV, spinal cord, and body and (b) LH&N phantom with
the PTV, parotid gland, spinal cord, and body.
dimensions of 265 mm (length) by 180 mm (diameter). The
phantom was made of homogeneous water-equivalent acryl
and included seven internal ion chamber (Pinpoint, PTW)
holes to allow absolute point dose measurements. Further,
it was designed to allow the insertion of film in the sagittal
plane for film dosimetry. After computed tomography (CT)
scanning of the phantom, we imported two types of struc-
ture sets that were developed by the American Association of
Physicists in Medicine, TG-119: the mock head and neck set
(LH&N), which represents a normal-sized H&N tumor and
the C-shaped set (C-shape), which represents a small-sized
tumor.27 Two virtual patients were created using these struc-
ture sets. The PTV for the LH&N phantom was 534 cm3 ,
which included the entire anterior volume from the base of
the skull to the upper neck and the posterior neck node. The
volumes of both parotids were approximately 30 cm3 , while
the spinal cord was defined as a cylindrical shape with a diam-
eter of 1.5 cm. The smallest gap between the spinal cord and
the posterior margin of the PTV was approximately 1.0 cm.
The PTV of the C-shape phantom was 120 cm3 . The spinal
cord substitute was a cylinder with a diameter of 1.5 cm, set
at a minimum 0.75-cm gap from the PTV (Fig. 1). The body
contour of the phantom was delineated to evaluate the integral
dose within the phantom in the DVH analysis.
2.B. Treatment planning
For the virtual C-shape and LH&N phantoms,
HD120MLC (2.5 mm) and M120MLC (5 mm) were
each employed to generate S&S_IMRT and VMAT plans
(Pinnacle3 , version 9.2, Philips Medical System). In the case
of the HD120MLC, 2.5-mm MLCs are only available for a
length of 8 cm in the central region. Therefore, the isocenter
of the beam was directed at the center of the PTV to cover
the entire PTV and OARs with the 2.5-mm MLC. For TPS
beam modeling for both linear accelerators, we applied 1%
of the interleaf leakage at the intersection of two adjacent
MLC leaves.28 To obtain optimal dose distributions in each
condition with a 6-MV photon beam, we used the same
beam parameters: a maximum of 130 segments, a minimum
segment area of 4 cm2 , a minimum segment MU of 3 MU,
and dose constraints for inverse planning (Table I).27 To
ensure comparability, the same number of iterations (250)
Medical Physics, Vol. 41, No. 2, February 2014
021712-3
TABLE I. Dose constraints for inverse planning.
Case Structure Constraint 1 (Gy) Constraint 2 (Gy)
C-Shape PTV D95 ≥ 50 D10 < 55
Spinal cord D10 < 25
LH&N PTV D99 > 46.5 D20 < 55
D90 ≥ 50
Spinal cord Dmax < 40
Parotids D50 < 20
Note: Dxx = dose received by xx% of the volume; Dmax = maximum dose.
was provided during dose optimization for each plan. During
inverse planning, priority was given to the PTV. When the
PTV dose reached the constraint early, optimization was
continued to reduce the OAR dose until the iteration limit,
while maintaining the PTV dose. All plans were calculated
using the collapsed-cone convolution algorithm on a 2-mm
dose grid for a 50-Gy prescription dose and 25 fractions to
the PTV.
For S&S_IMRT, we generated two types of plans: with
nine (9F-IMRT) and thirteen (13F-IMRT) equally spaced
beams. The direct machine parameter optimization module,
which was provided by the TPS manufacturer, was used for
dose optimization. Couch and collimator rotation were not
applied. For VMAT, single arc (VMAT1 ) and double arc
(VMAT2 ) plans were generated for both the C-shape and
LH&N using the SmartArc module that was supplied along
with the TPS. VMAT1 began at a gantry angle of 181◦ and
completed a full 360◦ clockwise rotation, while VMAT2 com-
pleted one full rotation in the same manner, followed by a sec-
ond rotation in the reverse direction. To minimize overlap of
interleaf leakage and the tongue-and-groove effect, the colli-
mator angle was set at 45◦ for VMAT1 . For the same reasons,
the angles for VMAT2 were set at 45◦ in the clockwise direc-
tion and 315◦ in the counterclockwise direction, so that they
would not coincide.29 Angular control point (CP) spacing was
set at 4◦ ,30 and a total of 91 control points were created for
each arc.
2.C. Plan comparison
In order to analyze the dosimetric effects of MLC leaf
width on the plan quality, the conformity (CI) (Refs. 31 and
32) and homogeneity index (HI) (Ref. 25) were calculated for
the PTV of each plan. The CI is defined as follows:
PTV2PIV
CI = , (1)
PTV × PIV
where PTVPIV is the PTV encompassed within the PIV, which
is the volume covered by the prescription isodose surface. HI
is an indication of dose uniformity within the PTV; a value of
1 indicates a uniform PTV dose distribution and is the ideal
value. The HI is defined as follows:
D5
HI = , (2)
D95
where D5 and D95 are the minimum doses received by 5%
and 95% of the PTV, respectively. To evaluate doses to OARs,
021712-4 Hong et al.: Dosimetric effects of MLC leaf width on IMRT
F IG . 2. Sagittal dose distributions for the C-shape phantom. VMAT (bottom row) showed generally better dose distributions than S&S_IMRT (upper row). The
2.5-mm MLC provided slightly a better PTV coverage and a lower spinal cord dose than the 5-mm MLC.
the maximum dose (Dmax ) was measured for the spinal cord,
while the mean dose was measured for the parotid glands.
Additionally, we computed the dose-spillage volume33, 34 to
evaluate the normal tissue sparing effect in IMRT according
to MLC leaf width. The dose-spillage volume is defined as
follows:
VX% − PTV
Dose − spillage volume = , (3)
PTV
where VX% is the volume covered by the X% isodose surface.
Results were categorized as high dose-spillage volume (Vx%
≥ 90% of the prescription dose, VHS ), intermediate dose-
spillage volume (Vx% ≥ 50% of the prescription dose, VIS ),
and low dose-spillage volume (Vx% ≥25% of the prescription
dose, VLS ).
Finally, the delivery efficiency of the MLCs was compared
to evaluate the treatment efficiencies of the different MLC leaf
widths. Delivery efficiency was evaluated by measuring the
MUs per fraction, total number of segments, and the actual
beam delivery time for each plan.
2.D. Verification of dosimetric error
To investigate the effects of MLC leaf width on dose error
in IMRT, the absolute point dose and sagittal dose distribu-
tions were measured for all plans. Subsequently, the distribu-
tion values were compared with the values calculated by the
TPS. Absolute doses at the PTV center of the virtual patient
phantom and at the center of the spinal cord were measured
using PinPoint chambers (TN 31015, PTW, Germany). Differ-
ences between the two MLCs were compared based on TPS
data.
Medical Physics, Vol. 41, No. 2, February 2014
021712-4
Two-dimensional dose distributions were obtained in the
mid-sagittal plane using Radiochromic film (EBT3, ISP,
Wayne). Approximately 24 h after irradiation, all exposed
films were digitized in the landscape format (Dosimetry Pro
advantage Red LED, VIDAR R
), according to the manufac-
35, 36
ture instructions. Optical densities on the film were con-
verted into relative doses by applying our film calibration
protocol.37 Finally, the results were converted into absolute
doses by applying the absolute point dose that had been mea-
sured using the ionizing chamber in the PTV. The measured
dose distribution was compared with the calculated dose dis-
tribution from the TPS by applying Gamma analysis with a
3% dose difference and a 3-mm distance to agreement thresh-
olds (RIT113 Version 6.0, Radiological Imaging Technol-
ogy). We used paired t-tests (PASW statistics ver. 18, IBM)
to assess dosimetric differences between the 2.5- and 5-mm
MLCs. Values of p < 0.05 were considered statistically sig-
nificant.
3. RESULTS
3.A. Plan comparison
3.A.1. C-shape
As compared with S&S_IMRT, VMAT showed generally
better dose distributions in the sagittal plane, which can pro-
vide better PTV coverage (Fig. 2). Further, the 2.5-mm MLC
provided slightly better PTV coverage than the 5-mm MLC
as well as a lower spinal cord dose (Figs. 2 and 3). Table II
presents S&S_IMRT and VMAT plan quality and delivery
efficiency results for the C-shape, according to MLC leaf
width. The CI was the best for VMAT2 , followed by VMAT1 ,
021712-5 Hong et al.: Dosimetric effects of MLC leaf width on IMRT
F IG . 3. Comparison of dose-volume histograms for the 2.5- and 5.0-mm MLCs for the C-shape phantom: (a) 9F_IMRT, (b) 13F_IMRT, (c) VMAT1 , and (d)
VMAT2 .
13F–IMRT, and 9F–IMRT in descending order. The MLC
with 2.5-mm leaf width provided a significantly better CI than
did the MLC with 5-mm leaf width (p = 0.047). The differ-
ence in CI according to leaf width was larger for VMAT than
for S&S_IMRT. However, use of the 2.5-mm MLC was not
associated with a significant improvement in HI (p = 0.252).
Further, the HI was slightly worse for S&S_IMRT than for
VMAT.
The MLC with 2.5-mm leaf width was associated with a
significantly lower spinal cord dose than the MLC with 5-mm
leaf width (p = 0.002), with maximum reduction of 5.07% in
the case of VMAT1 . The lowest maximum spinal cord dose
was observed for VMAT2 , followed by VMAT1 , 13F–IMRT,
and 9F–IMRT in ascending order. Dose-spillage volume anal-
ysis indicated that VHS and VIS were significantly lower
for the 2.5-mm MLC than for the 5-mm MLC (p < 0.05),
although we observed no significant differences in VLS
(p = 0.09). In addition, we found that the 2.5-mm MLC was
associated with a greater reduction in dose-spillage volumes
for VMAT than for S&S_IMRT.
Analysis of delivery efficacy revealed that the number of
segments increased for the 2.5-mm MLC for S&S_IMRT, but
no differences were observed for VMAT because the same
angular CP spacing was applied for optimization. In contrast,
MUs per fraction and delivery time were increased with the
5-mm MLC (with the exception of MUs per fraction for 13F–
IMRT). However, the increases were not statistically signifi-
Medical Physics, Vol. 41, No. 2, February 2014
021712-5
cant (p > 0.05). Delivery times were clearly smaller for both
VMAT plans than for both S&S_IMRT plans.
3.A.2. LH&N
As compared with S&S_IMRT, VMAT showed generally
better dose distributions in the sagittal plane, which could
provide better PTV coverage and minimize peripheral doses
(Fig. 4). VMAT was also observed to provide substantially
better dose homogeneity in the target. However, no significant
differences in DVH were observed between the 2.5- and 5-
mm MLCs (Fig. 5). Table III presents S&S_IMRT and VMAT
plan quality and delivery efficiency results for the LH&N, ac-
cording to MLC leaf width. We did not observe any signifi-
cant differences in the CI or HI of the PTV according to leaf
width (p > 0.05). Although S&S_IMRT tended to be asso-
ciated with better CI and HI than VMAT, the difference was
not very large for LH&N. Unlike the C-shape analysis, the
LH&N analysis did not reveal any significant differences in
maximum spinal cord dose, mean parotid dose, VHS , VIS , and
VLS according to leaf width (p > 0.05).
The analysis of delivery efficacy revealed that the number
of segments increased with the 2.5-mm MLC for S&S_IMRT,
but no differences were observed for VMAT because the
same angular CP spacing was applied for optimization. MUs
per fraction were significantly elevated for the 5-mm MLC
(p < 0.05). We found no significant differences in delivery
 021712-6 Hong et al.: Dosimetric effects of MLC leaf width on IMRT 021712-6

time according to leaf width (p = 0.218). MUs per fraction

P value (2.5-vs

Note: VHS = high-dose spillage volume (≥90% of the prescription dose); VIS = intermediate-dose spillage volume (≥50% of the prescription dose); VLS = low-dose spillage volume (≥25% of the prescription dose);
5-mm MLC)
and delivery times were clearly smaller for the VMAT plans

0.047
0.252
0.002
0.017

0.018
0.090
0.190
0.217
0.115
than for the S&S_IMRT plans.

3.B. Verification of dosimetric error

% diff. (%)

4.67
− 1.63
− 4.60
− 14.43

− 10.62
− 4.22

− 6.93
− 2.15
Table IV presents the result of our dosimetric error anal-

0
ysis, which was based on TPS data. For both C-shape and
LH&N, we have included the percentage difference in abso-
lute point dose measurements and the percentage pass rate
11.847
5-mm

0.813
1.043

0.388

2.986
2488

of Gamma analysis. No significant differences were observed

3.25
182
750
VMAT2

between the 2.5- and 5-mm MLCs. In C-shape analysis, the

percentage differences in absolute point dose measurements
2.5-mm

11.347

at the PTV and spinal cord were smaller for S&S_IMRT

0.851
1.026

0.332

2.669
2374

3.18
182
698

(<0.7%) than for VMAT (<−2.8%). However, statistically

significant differences were not observed in the LH&N analy-
% diff. (%)

sis. In both C-shape and LH&N analyses, high Gamma pass-

6.52
− 1.32
− 5.07
− 16.83

− 11.19
− 4.66

− 16.34
− 4.26

ing rates (>95%) were observed for the 3%/3 mm criteria for
0

S&S_IMRT and VMAT.

12.061
5-mm

0.798
1.057

0.416

3.252
2577

1.41

4. DISCUSSION
667
91
VMAT1

In the present study, we aimed to quantitatively analyze

2.5-mm

the dosimetric effects of MLC leaf widths in IMRT for head

11.499
0.850
1.043

0.346

2.888
2446

1.35
558

and neck tumors of different sizes. To achieve this, we em-

91

ployed virtual phantom patients, which were created from

mock structure sets. Avoiding the use of actual patients al-
% diff. (%)

3.65
0.19
− 2.91
− 8.71

− 11.21
− 1.04
6.09
4.59
− 5.88

lowed us to disregard patient-related variability and made it

possible to measure radiation doses using ionization cham-
bers and films. With these measurements, we were able to
compare the doses that were delivered with doses calculated
12.093
5-mm

0.739
1.054

0.574

3.979

using the TPS, thereby quantifying dosimetric errors in IMRT

2886

7.31
115
653
13F_IMRT

for different MLC leaf widths.

TABLE II. Comparisons of dosimetric characteristics and delivery efficiency for C-shape.

The MLC is a key invention in the history of radiation

2.5-mm

therapy.36 Indeed, a MLC does not only serve as a shielding

11.967
0.766
1.056

0.524

3.533
2802

6.88
122
683

device that improves target dose conformity—it also plays a

crucial role in generating the intensity-modulated beam for
IMRT. The leaf width of MLCs is important for both the tar-
% diff. (%)

0.95
0.09
− 3.14
− 3.44

− 2.92
− 0.55
6.33
− 7.13
− 3.62

get shaping resolution and fine intensity modulation. A thin

leaf width is expected to provide superior dosimetric advan-
tages in IMRT planning because it generally results in higher
resolutions. Many previous studies have investigated the dosi-
11.991
5-mm

0.733
1.058

0.610

4.459
3183

metric benefits of thin-leaf MLCs in IMRT and stereotactic

5.53
659
79
9F_IMRT

radiosurgery for small-volume targets.22, 38 The results of the

present study are consistent with the results of these previous
2.5-mm

11.925

studies. In particular, we have confirmed that 2.5-mm MLC

0.740
1.059

0.589

4.329
3083

5.33
612
84

can provide better a CI for PTV when treating a C-shape le-

sion with S&S_IMRT or VMAT. We have further confirmed
Dmax (cGy)
VHS (ratio)

that 2.5-mm MLC can reduce maximum spinal cord doses.

VLS (ratio)
VIS (ratio)

Indeed, according to the results of the dose-spillage volume

analysis, 2.5-mm MLC can very effectively minimize high
Dmax = maximum dose.
HI
CI

Delivery time (min)

(VHS , 90% of the prescribe dose) and intermediate (VIS , 50%

of the prescribe dose) dose volumes near the target, but can-
Dose spillage

not reduce low dose volumes (VLS , 25% of prescribe dose).

Spinal cord
Parameter

Segments

These advantages tend to be more evident in VMAT than in

MUs/fx.
volume

IMRT. However, leaf width was not observed to significantly

PTV

improve the HI for the PTV.

Medical Physics, Vol. 41, No. 2, February 2014

021712-7 Hong et al.: Dosimetric effects of MLC leaf width on IMRT 021712-7

F IG . 4. Sagittal dose distributions for the LH&N phantom. VMAT (bottom row) showed generally better dose distributions than S&S_IMRT (upper row). No
significant differences were observed between the 2.5- and 5-mm MLCs.

Relatively few studies have investigated the dosimetric and 5-mm MLCs in VMAT for head and neck cancers,
effects of MLC leaf width in IMRT for head and neck can- although VMAT is becoming an increasingly common
cers with comparatively large tumors. In particular, no pre- option for head and neck cancers. In a previous study of
vious studies have compared the dosimetric effects of 2.5- 2.5-, 5-, and 10-mm MLCs in IMRT of the head and neck, no

F IG . 5. Comparison of dose-volume histograms for the 2.5- and 5.0-mm MLCs for the LH&N phantom: (a) 9F_IMRT, (b) 13F_IMRT, (c) VMAT1 , and (d)
VMAT2 .

Medical Physics, Vol. 41, No. 2, February 2014

 021712-8 Hong et al.: Dosimetric effects of MLC leaf width on IMRT 021712-8

significant differences were observed for the PTV or OARs.21

P value (2.5- vs. 5-mm MLC)

Note: VHS = high-dose spillage volume (≥90% of the prescription dose); VIS = intermediate-dose spillage volume (≥50% of the prescription dose); VLS = low-dose spillage volume (≥25% of the prescription dose);
In a comparison of 4- and 10-mm MLCs in IMRT for na-
sopharyngeal carcinoma, researchers found that the 4-mm
MLC offered slightly more optimal target coverage, with-
0.176

0.652

0.297
0.141
0.346

0.182
0.344
0.071

0.017
0.218
out improving the doses delivered to OARs.25 In a study
that compared 5- and 10-mm MLCs in both S&S_IMRT
and sliding-window IMRT of the head and neck, the authors
concluded that leaf width did not substantially affect target
conformity or normal-tissue sparing.22 In the present study,
we observed similar results; dosimetric parameters did not
% diff. (%)

2.76

2.40

0.00
− 0.29
− 0.16

− 15.84
− 0.73
− 1.43

− 3.47
− 3.08
differ significantly between 2.5- and 5-mm MLCs in either
S&S_IMRT or VMAT for LH&N. Consequently, MLC leaf
width could have fairly minimal effects on the dosimetric
characteristics of S&S_IMRT and VMAT for LH&N, even
5-mm

0.832
1.038

1.923
0.202

3.365
3417
1814

3.25
182
490

though substantial differences were observed for C-shape

VMAT2

targets.
2.5-mm

According to our hospital data, approximately 21.3% of

0.855
1.035

0.170
1.909
3.317
3412
1857

3.15
182
473

all head and neck cancer patients had PTV length greater than
22 cm (35% for nasopharyngeal cancer, 6.25% for oropha-
% diff. (%)

ryngeal cancer), which exceeds the maximum field size for

3.63
0.00
− 0.36
3.45
− 15.71
1.92
− 0.94
0.00
− 5.44
− 3.68

Varian Novalis Tx with a 2.5-mm MLC. Currently, this prob-

lem is addressed by using a two-isocenter IMRT technique or
by rotating the collimator angle to expand the field size.39, 40
5-mm

0.210
0.826
1.046

1.982
3.393
3531
1819

However, these methods can cause hot spots or cold spots at

1.36
478
91
VMAT1

the junction of the field and can also increase the number of
subfields, which may itself increase the total MUs and the
2.5-mm

0.856

0.177
1.046

2.020
3.361
3519
1882

treatment time, respectively. In contrast, the Varian iX ma-

1.31
452
91

chine is not subject to a maximum field size limit during the

treatment of head and neck cancers, because it provides a
% diff. (%)

0.47
− 0.09
1.12
− 1.22

2.48
− 0.39
− 0.78
− 2.57
− 10.41

maximum field length of 40 cm with a 5-mm MLC. In addi-

0

tion, our study indicated that there is no significant dosimetric

penalty for treating patients with cancers of the head and neck
using a machine with 5-mm leaves.
5-mm

0.824
1.053

0.213
2.013
3.363
3476
1881

8.93
129
622
13F_IMRT

Therefore, our study can provide guidance for allocating

TABLE III. Comparisons of dosimetric characteristics and delivery efficiency for LH&N.

treatment machines for IMRT of the head and neck cancer as

2.5-mm

well as for selecting new systems for a radiation therapy cen-

0.824
1.052

0.214
2.063
3.350
3515
1858

128
606
8.0

ter. Delivery efficiency is an important factor for patient safety

and has economic benefits. In the present study, we found
% diff. (%)

that more segments were generated by the 2.5-mm MLC in

0.12
− 0.09
1.34
− 1.89
− 0.48
− 0.19
− 0.24
− 4.55
− 1.45
− 2.67

S&S_IMRT, but the MUs per fraction and delivery time were
both reduced in most cases. However, the number of segments
was reduced for the 2.5-mm MLC for the LH&N, while the
5-mm

0.826
1.065

0.209
2.070
3.373
3545
1859

same reductions in MU per fraction and delivery time were

110
619
7.5
9F_IMRT

observed. The fine-leaf MLC appears to make effective use of

resources during dose optimization.
2.5-mm

0.827
1.064

0.208
2.066
3.365
3593
1824

Our study employed treatment machines from the same

105
610
7.3

vendor to focus on the dosimetric effects of MLC leaf width.

Particularly, machines were selected from the same ven-
Dmean (cGy)
Dmax (cGy)

VHS (ratio)

VLS (ratio)
VIS (ratio)

dor to minimize any unknown dosimetric component that

might result from the different physical and geometrical
HI
CI

characteristics of different machines. However, the Varian

HD120MLC differs slightly from the previous models of Var-
Dmax = maximum dose.
Dose spillage volume

Delivery time (min)

ian M120MLC, mainly with respect to its leaf-end design and

leaf height.15 In addition, although both of the MLCs were
Parotid gland

made of tungsten alloy, they have slightly different tungsten

Spinal cord
Parameter

Segments

contents. Fortunately, these differences were not large enough

MUs/fx.

to affect the dosimetric characteristics. Furthermore, the

PTV

majority of TPSs consider the physical and geometric

Medical Physics, Vol. 41, No. 2, February 2014

021712-9 Hong et al.: Dosimetric effects of MLC leaf width on IMRT 021712-9

TABLE IV. Results of dosimetric error for C-shape and LH&N.

9F_IMRT 13F_IMRT VMAT1 VMAT2

2.5-mm 5-mm 2.5-mm 5-mm 2.5-mm 5-mm 2.5-mm 5-mm P value (2.5- vs.
MLC MLC MLC MLC MLC MLC MLC MLC 5-mm MLC)

C-shape % diff. (PTV,%) 0.697 − 0.668 0.460 − 0.525 − 0.754 0.860 − 1.021 − 1.325 0.721
% diff. (Cord,%) 0.372 0.312 0.208 − 0.105 − 2.185 − 2.037 − 2.834 − 2.680 0.345
γ (3%, 3 mm) % 98.61 98.93 98.35 99.44 98.89 98.03 96.71 98.50 0.378
pass rate (%)
LH&N % diff. (PTV,%) − 0.356 − 0.212 − 0.115 0.662 − 0.321 − 0.008 0.372 − 0.184 0.437
% diff. (Cord,%) 0.977 0.779 0.619 0.657 0.637 − 0.131 − 0.660 − 0.342 0.556
γ (3%, 3 mm) % 96.66 98.85 98.13 98.74 98.86 98.03 98.03 97.58 0.613
pass rate (%)

characteristics of the MLC as a component of their dose ACKNOWLEDGMENTS

calculations.
This work was supported by the Technology Innovation
For treatment of the same target volumes, the use of
Program, 10040362, Development of an integrated manage-
MLC with narrower leaf width results in a higher number of
ment solution for radiation therapy funded by the Ministry of
MLC leaf junctions within the radiation field. Because of in-
Knowledge Economy (MKE, Korea) and by Samsung Medi-
creases in interleaf leakage and the tongue-and-groove effect,
cal Center grant [GFO1130081]. The authors report no con-
a greater number of leaf junctions increase the possibility of
flicts of interest in conducting the research.
dose uncertainty. However, we observed no significant differ-
ences in dosimetric error between the 2.5- and 5-mm MLCs.
This may result from the specific TPS that was used in the
present study, which accounted for interleaf leakage and the a) Author to whom correspondence should be addressed. Electronic mail:
tongue-and-groove effect during beam modeling. These were sg.ju@samsung.com; Telephone: 82-2-3410-2617; Fax: 82-2-3410-2619.
b) Author to whom correspondence should be addressed. Electronic mail:
also reflected in the radiation dose calculations that were used
suhsanta@catholic.ac.kr; Telephone: 82-2-2258-7232; Fax: 82-2-2258-
to optimize the IMRT plan. Therefore, they may have had no
7506.
influence on error assessment. 1 B. Longobardi, E. De Martin, C. Fiorino, I. Dell’oca, S. Broggi, G. M. Cat-

taneo, and R. Calandrino, “Comparing 3DCRT and inversely optimized

5. CONCLUSION
We evaluated the dosimetric effect of MLC leaf width in
IMRT for head and neck cancer. MLC with 2.5-mm leaves
IMRT planning for head and neck cancer: Equivalence between step-and-
shoot and sliding window techniques,” Radiother. Oncol. 77, 148–156
(2005).
2 C. A. Kristensen, F. Kjaer-Kristoffersen, W. Sapru, A. K. Berthelsen,
A. Loft, and L. Specht, “Nasopharyngeal carcinoma. Treatment planning
with IMRT and 3D conformal radiotherapy,” Acta Oncol. 46, 214–220
(2007).
3 T. Gupta, J. Agarwal, S. Jain, R. Phurailatpam, S. Kannan, S. Ghosh-

showed better dosimetric characteristics, providing better
dose conformity to the target and reducing spinal cord and
peripheral doses in our examination of a C-shape. These ben-
efits were greater for VMAT than for S&S_IMRT. In con-
trast, the use of 2.5-mm MLC in IMRT for the LH&N, rep-
resenting a large target volume, was not observed to provide
dosimetric benefits. In addition, we found no significant dif-
ferences in dosimetric error according to MLC leaf width.
Consequently, MLC with fine leaves may provide only a small
benefit in terms of dosimetric characteristics in S&S_IMRT
and VMAT for LH&N targets compared with C-shape
targets.
Laskar, V. Murthy, A. Budrukkar, K. Dinshaw, K. Prabhash, P. Chaturvedi,
and A. D’Cruz, “Three-dimensional conformal radiotherapy (3D-CRT)
versus intensity modulated radiation therapy (IMRT) in squamous cell car-
cinoma of the head and neck: A randomized controlled trial,” Radiother.
Oncol. 104, 343–348 (2012).
4 A. Al-Mamgani, P. Van Rooij, L. Tans, D. N. Teguh, and P. C. Lev-
endag, “Toxicity and outcome of intensity-modulated radiotherapy ver-
sus 3-dimensional conformal radiotherapy for oropharyngeal cancer:
A matched-pair analysis,” Technol. Cancer Res. Treat. 12, 123–130
(2013).
5 M. Lambrecht, D. Nevens, and S. Nuyts, “Intensity-modulated radiother-
apy vs. parotid-sparing 3D conformal radiotherapy: Effect on outcome and
toxicity in locally advanced head and neck cancer,” Strahlenther. Onkol.
189, 223–229 (2013).
6 M. Johnston, S. Clifford, R. Bromley, M. Back, L. Oliver, and T. Eade,

No significant differences in delivery efficiency were ob-
served between the two MLCs, except for MUs per fraction
for the LH&N phantom. The results of this study will serve
“Volumetric-modulated arc therapy in head and neck radiotherapy: A plan-
ning comparison using simultaneous integrated boost for nasopharynx and
oropharynx carcinoma,” Clin. Oncol. (R. Coll. Radiol. 23, 503–511 (2011).
7 M. Voordeckers, A. Farrag, H. Everaert, K. Tournel, G. Storme,

as useful decision-making criteria when assigning equipment
for IMRT in clinical practice, such as for head and neck can-
cers. The results will also help guide the introduction of new
equipment into clinical practice.
D. Verellen, and M. De Ridder, “Parotid gland sparing with helical
tomotherapy in head-and-neck cancer,” Int. J. Radiat. Oncol., Biol., Phys.
84, 443–448 (2012).
8 M. Kong, S. E. Hong, J. Choi, and Y. Kim, “Comparison of sur-
vival rates between patients treated with conventional radiotherapy and

Medical Physics, Vol. 41, No. 2, February 2014

021712-10 Hong et al.: Dosimetric effects of MLC leaf width on IMRT
helical tomotherapy for head and neck cancer,” J. Radiat. Oncol. 31, 1–11
(2013).
9 M. T. Studenski, V. Bar-Ad, J. Siglin, D. Cognetti, J. Curry, M. Tuluc, and
A. S. Harrison, “Clinical experience transitioning from IMRT to VMAT for
head and neck cancer,” Med. Dosim. 38, 171–175 (2013).
10 R. Mohan, K. Jayesh, R. C. Joshi, P. Narayanamurthy, S. K. Majumdar,
and M. Al-Idirisi, “Comparison of dosimetric characteristics of 120-leaf
and 80-leaf multi-leaf collimators in a Varian linear accelerator for a 6-MV
photon beam,” Radiol. Phys. Technol. 1, 223–228 (2008).
11 S. Lissner, K. Schubert, S. Kluter, D. Oetzel, and J. Debus, “A method for
testing the performance and the accuracy of the binary MLC used in helical
tomotherapy,” Z. Med. Phys. 23, 153–161 (2013).
12 M. L. Taylor, L. N. McDermott, P. N. Johnston, M. Haynes, T. Ackerly,
T. Kron, and R. D. Franich, “Stereotactic fields shaped with a micro-
multileaf collimator: Systematic characterization of peripheral dose,” Phys.
Med. Biol. 55, 873–881 (2010).
13 J. Deng, T. Pawlicki, Y. Chen, J. Li, S. B. Jiang, and C. M. Ma, “The MLC
tongue-and-groove effect on IMRT dose distributions,” Phys. Med. Biol.
46, 1039–1060 (2001).
14 J. S. Li, T. Lin, L. Chen, R. A. Price, Jr., and C. M. Ma, “Uncertainties in
IMRT dosimetry,” Med. Phys. 37, 2491–2500 (2010).
15 D. S. Sharma, P. M. Dongre, V. Mhatre, and M. Heigrujam, “Physi-
cal and dosimetric characteristic of high-definition multileaf collimator
(HDMLC) for SRS and IMRT,” J. Appl. Clin. Med. Phys. 12, 142–160
(2011).
16 A. Dhabaan, E. Elder, E. Schreibmann, I. Crocker, W. J. Curran, N. M. Oye-
siku, H. K. Shu, and T. Fox, “Dosimetric performance of the new high-
definition multileaf collimator for intracranial stereotactic radiosurgery,” J.
Appl. Clin. Med. Phys. 11, 197–211 (2010).
17 K. Ohtakara, S. Hayashi, H. Tanaka, and H. Hoshi, “Dosimetric compar-
ison of 2.5 mm vs. 3.0 mm leaf width micro-multileaf collimator-based
treatment systems for intracranial stereotactic radiosurgery using dynamic
conformal arcs: Implications for treatment planning,” Jpn. J. Radiol. 29,
630–638 (2011).
18 J. A. Tanyi, C. M. Kato, Y. Chen, Z. Chen, and M. Fuss, “Impact of the
high-definition multileaf collimator on linear accelerator-based intracranial
stereotactic radiosurgery,” Br. J. Radiol. 84, 629–638 (2011).
19 J. Saitoh, Y. Saito, T. Kazumoto, S. Kudo, A. Ichikawa, N. Hayase,
K. Kazumoto, H. Sakai, and K. Shibuya, “Therapeutic effect of linac-based
stereotactic radiotherapy with a micro-multileaf collimator for the treat-
ment of patients with brain metastases from lung cancer,” Jpn. J. Clin. On-
col. 40, 119–124 (2010).
20 J. Y. Kim, S. Y. Park, D. H. Lee, S. H. Lee, T. H. Kim, and K. H. Cho,
“Comparison of treatment plans with multileaf collimators of different leaf
width,” Korean J. Med. Phys. 15, 173–178 (2004).
21 V. Jacob, W. Bayer, S. T. Astner, R. Busch, and P. Kneschaurek, “A plan-
ning comparison of dynamic IMRT for different collimator leaf thicknesses
with helical tomotherapy and RapidArc for prostate and head and neck tu-
mors,” Strahlenther. Onkol. 186, 502–510 (2010).
22 S. A. Yoganathan, K. R. Mani, K. J. Das, A. Agarwal, and S. Kumar,
“Dosimetric effect of multileaf collimator leaf width in intensity-modulated
radiotherapy delivery techniques for small- and large-volume targets,” J.
Med. Phys. 36, 72–77 (2011).
23 J. B. Fiveash, H. Murshed, J. Duan, M. Hyatt, J. Caranto, J. A. Bonner, and
R. A. Popple, “Effect of multileaf collimator leaf width on physical dose
distributions in the treatment of CNS and head and neck neoplasms with
intensity modulated radiation therapy,” Med. Phys. 29, 1116–1119 (2002).
24 F. Zwicker, H. Hauswald, S. Nill, B. Rhein, C. Thieke, F. Roeder, C. Timke,
A. Zabel-du Bois, J. Debus, and P. E. Huber, “New multileaf collimator
with a leaf width of 5 mm improves plan quality compared to 10 mm in
Medical Physics, Vol. 41, No. 2, February 2014
021712-10
step-and-shoot IMRT of HNC using integrated boost procedure,” Strahlen-
ther. Onkol. 186, 334–343 (2010).
25 S. Wang, Y. Gong, Q. Xu, S. Bai, Y. Lu, Q. Jiang, and N. Chen, “Impacts of
multileaf collimators leaf width on intensity-modulated radiotherapy plan-
ning for nasopharyngeal carcinoma: Analysis of two commercial elekta de-
vices,” Med. Dosim. 36, 153–159 (2011).
26 Z. van Kesteren, T. M. Janssen, E. Damen, and C. van Vliet-Vroegindeweij,
“The dosimetric impact of leaf interdigitation and leaf width on VMAT
treatment planning in Pinnacle: Comparing Pareto fronts,” Phys. Med. Biol.
57, 2943–2952 (2012).
27 G. A. Ezzell, J. W. Burmeister, N. Dogan, T. J. LoSasso, J. G. Mechalakos,
D. Mihailidis, A. Molineu, J. R. Palta, C. R. Ramsey, B. J. Salter, J. Shi,
P. Xia, N. J. Yue, and Y. Xiao, “IMRT commissioning: Multiple institution
planning and dosimetry comparisons, a report from AAPM Task Group
119,” Med. Phys. 36, 5359–5373 (2009).
28 T. LoSasso, C. S. Chui, and C. C. Ling, “Physical and dosimetric aspects
of a multileaf collimation system used in the dynamic mode for imple-
menting intensity modulated radiotherapy,” Med. Phys. 25, 1919–1927
(1998).
29 M. Oliver, I. Gagne, C. Popescu, W. Ansbacher, and W. A. Beckham,
“Analysis of RapidArc optimization strategies using objective function val-
ues and dose-volume histograms,” J. Appl. Clin. Med. Phys. 11, 10–23
(2010).
30 K. Bzdusek, H. Friberger, K. Eriksson, B. Hardemark, D. Robinson, and
M. Kaus, “Development and evaluation of an efficient approach to volu-
metric arc therapy planning,” Med. Phys. 36, 2328–2339 (2009).
31 A. van’t Riet, A. C. Mak, M. A. Moerland, L. H. Elders, and W. van der
Zee, “A conformation number to quantify the degree of conformality in
brachytherapy and external beam irradiation: Application to the prostate,”
Int. J. Radiat. Oncol., Biol., Phys. 37, 731–736 (1997).
32 L. Feuvret, G. Noel, J. J. Mazeron, and P. Bey, “Conformity index: A re-
view,” Int. J. Radiat. Oncol., Biol., Phys. 64, 333–342 (2006).
33 J. A. Tanyi, P. A. Summers, C. L. McCracken, Y. Chen, L. C. Ku, and
M. Fuss, “Implications of a high-definition multileaf collimator (HD-MLC)
on treatment planning techniques for stereotactic body radiation therapy
(SBRT): A planning study,” Radiat. Oncol. 4, 22–28 (2009).
34 R. D. Timmerman, B. D. Kavanagh, L. C. Cho, L. Papiez, and L. Xing,
“Stereotactic body radiation therapy in multiple organ sites,” J. Clin. Oncol.
25, 947–952 (2007).
35 D. F. Lewis, “Performance of the Vidar R
Red LED Dosimetry Pro
AdvantageTM : A scanner optimized for use with GAFCHROMIC R
EBT Dosimetry Film (available URL: http://www.vidar.com/film/images/
stories/PDFs/products/dosimetrypro/pdf/vidarredled.pdf).
36 V. T. DeVita, Jr., and S. A. Rosenberg, “Two hundred years of cancer re-
search,” N. Engl. J. Med. 366, 2207–2214 (2012).
37 S. G. Ju, Y. Han, O. Kum, K. H. Cheong, E. H. Shin, J. S. Shin, J. S. Kim,
and Y. C. Ahn, “Comparison of film dosimetry techniques used for quality
assurance of intensity modulated radiation therapy,” Med. Phys. 37, 2925–
2933 (2010).
38 Q. J. Wu, Z. Wang, J. P. Kirkpatrick, Z. Chang, J. J. Meyer, M. Lu,
C. Huntzinger, and F. F. Yin, “Impact of collimator leaf width and treatment
technique on stereotactic radiosurgery and radiotherapy plans for intra- and
extracranial lesions,” Radiat.Oncol. 4, 3–12 (2009).
39 C. C. Lee, A. Wu, M. Garg, S. Mutyala, S. Kalnicki, G. Sayed, and D. Mah,
“A new approach to reduce number of split fields in large field IMRT,”
Med. Dosim. 36, 1–5 (2011).
40 S. P. Srivastava, I. J. Das, A. Kumar, and P. A. Johnstone, “Dosi-
metric comparison of split field and fixed jaw techniques for large
IMRT target volumes in the head and neck,” Med. Dosim. 36, 6–9
(2011).