Volumetric modulated arc therapy: IMRT in a single gantry arc

Karl Ottoa兲
Vancouver Cancer Centre, BC Cancer Agency, Vancouver, British Columbia V5Z 4E6, Canada
共Received 25 June 2007; revised 21 September 2007; accepted for publication 5 November 2007;
published 26 December 2007兲
In this work a novel plan optimization platform is presented where treatment is delivered efficiently
and accurately in a single dynamically modulated arc. Improvements in patient care achieved
through image-guided positioning and plan adaptation have resulted in an increase in overall treat-
ment times. Intensity-modulated radiation therapy 共IMRT兲 has also increased treatment time by
requiring a larger number of beam directions, increased monitor units 共MU兲, and, in the case of
tomotherapy, a slice-by-slice delivery. In order to maintain a similar level of patient throughput it
will be necessary to increase the efficiency of treatment delivery. The solution proposed here is a
novel aperture-based algorithm for treatment plan optimization where dose is delivered during a
single gantry arc of up to 360 deg. The technique is similar to tomotherapy in that a full 360 deg of
beam directions are available for optimization but is fundamentally different in that the entire dose
volume is delivered in a single source rotation. The new technique is referred to as volumetric
modulated arc therapy 共VMAT兲. Multileaf collimator 共MLC兲 leaf motion and number of MU per
degree of gantry rotation is restricted during the optimization so that gantry rotation speed, leaf
translation speed, and dose rate maxima do not excessively limit the delivery efficiency. During
planning, investigators model continuous gantry motion by a coarse sampling of static gantry
positions and fluence maps or MLC aperture shapes. The technique presented here is unique in that
gantry and MLC position sampling is progressively increased throughout the optimization. Using
the full gantry range will theoretically provide increased flexibility in generating highly conformal
treatment plans. In practice, the additional flexibility is somewhat negated by the additional con-
straints placed on the amount of MLC leaf motion between gantry samples. A series of studies are
performed that characterize the relationship between gantry and MLC sampling, dose modeling
accuracy, and optimization time. Results show that gantry angle and MLC sample spacing as low as
1 deg and 0.5 cm, respectively, is desirable for accurate dose modeling. It is also shown that
reducing the sample spacing dramatically reduces the ability of the optimization to arrive at a
solution. The competing benefits of having small and large sample spacing are mutually realized
using the progressive sampling technique described here. Preliminary results show that plans gen-
erated with VMAT optimization exhibit dose distributions equivalent or superior to static gantry
IMRT. Timing studies have shown that the VMAT technique is well suited for on-line verification
and adaptation with delivery times that are reduced to ⬃1.5– 3 min for a 200 cGy
fraction. © 2008 American Association of Physicists in Medicine.
关DOI: 10.1118/1.2818738兴

Key words: VMAT, IMAT, arc therapy, IMRT, tomotherapy

I. INTRODUCTION ery time by requiring a larger numbers of beam directions,

The goal of this work is to develop a treatment plan optimi-
zation and delivery platform that is 共1兲 time efficient; 共2兲
capable of producing highly conformal dose distributions
which are 共3兲 delivered with superior dosimetric accuracy.
The solution proposed here is a novel algorithm for treatment
plan optimization where the dose volume is delivered during
a single 360 deg gantry arc. The technique is referred to as
volumetric modulated arc therapy 共VMAT兲.
Recently, on-line imaging techniques have provided clini-
cians with additional tools for verifying patient position and
adapting treatment plans.1–5 Employing these technologies
will require additional time for image acquisition, evaluation,
patient repositioning, and adaptation.6 Intensity-modulated
radiation therapy 共IMRT兲 has also increased treatment deliv-
increased monitor units 共MU兲, and, in the case of
tomotherapy,7,8 a slice-by-slice delivery.
Intensity modulated arc therapy9 共IMAT兲 was proposed by
Yu in 1995 but has not yet attained widespread adoption.
IMAT has the theoretical advantage of increased flexibility in
deriving highly conformal plans by using a large number of
beam directions. Unfortunately, this increased flexibility is
somewhat negated by the additional constraints placed on
MLC leaf motions. In order for the gantry to move continu-
ously throughout delivery, the MLC leaf position changes
between consecutive gantry positions must be restricted.
Most investigators overcome this restriction by using several
superimposed arcs,9–11 which increases treatment time. Also,
optimization of the dynamic gantry motion is typically per-
formed using a coarse sampling of the gantry directions and

310 Med. Phys. 35 „1…, January 2008 0094-2405/2008/35„1…/310/8/$23.00 © 2008 Am. Assoc. Phys. Med. 310
311 Karl Otto: Single arc radiation therapy 311

MLC leaf positions. Each sample is effectively a “snapshot” weights are physically impossible and are therefore rejected
of the gantry and MLC at a time point during the delivery. by the optimization. Constraints are also placed on MLC leaf
Continuous gantry and MLC motion is approximated by in- motion and MU variation to preserve continuous delivery.
dividual beams that are optimized at 5–10 deg gantry angle These are referred to as “efficiency constraints.” Defined in
increments with MLC leaf positions varying by up to 2–4 cm terms of gantry rotation angle, they are
between gantry angle samples.9–14 Poor sampling can sub-
stantially degrade plan accuracy, thereby reducing the quality
of the delivered plan and potentially resulting in an unaccept-
⌬x
⌬␪
ⱕ 冉 冊
dx
d␪ max
,

able dosimetric error.10 Increasing the gantry and MLC sam-
pling frequency will improve accuracy but it will also further
restrict the ability of the optimization and leaf sequencing
algorithms to derive an acceptable plan. VMAT provides a
novel solution to several limitations encountered with IMAT.
In particular, the VMAT technique is designed so that opti-
mized plans may be delivered:
共1兲 Efficiently, in a single gantry arc
共2兲 With high dose conformality, using a full 360 deg range
of gantry directions
共3兲 Accurately, with high-resolution sampling of beam di-
rections during planning.
冉 冊
⌬MU dMU
ⱕ , 共1兲
⌬␪ d␪ max
where x, MU, and ␪ are MLC leaf position, MU weight, and
gantry angle, respectively. 共dx / d␪兲max and 共dMU/ d␪兲max are
selected so that there is no particular component of the de-
livery system compromising a continuous and efficient deliv-
ery. For example, at maximum speed the gantry rotates
through a 360 deg arc in 60 s 共共d␪ / dt兲max = 6 deg/ s兲. The
maximum leaf motion speed of the Varian Millennium 120
leaf MLC is 共dx / dt兲max = 3.0 cm/ s. By constraining VMAT
delivery to a maximum leaf displacement of 0.5 cm per de-
gree of gantry rotation,

II. METHODS 冉 冊 冉 冊 冒冉 冊
dx
d␪ max
=
dx
dt max
d␪
dt max
= 0.5 cm/deg, 共2兲
II.A. Optimization
the maximum total time for MLC motion over a 360 deg arc
The classical approach to IMRT and IMAT planning in- is 60 s, which matches the 60 s gantry rotation period.
volves optimization of fluence maps followed by a MLC leaf Similar constraints defined for 共dMU/ d␪兲max are based on
sequencing step. An alternative approach is to predefine a the gantry rotation speed and maximum dose rate, where

series of beam’s eye view 共BEV兲 aperture shapes that are
Boolean operations of target and healthy tissue structures15
共e.g., BEV of prostate excluding rectum兲. The number of MU
共radiation output兲 for each BEV aperture is then optimized
based on a dose-volume cost function. Both of these ap-
proaches have been applied to IMAT.9–11,14,16–19 Recently,
several investigators have proposed directly optimizing leaf
positions and segment weights for both static gantry IMRT
共Refs. 20–22兲 as well as IMAT.11 Generally, this optimization
technique involves selecting a MLC aperture, then changing
either the MU weight or a MLC leaf position for that sample.
VMAT dose optimization employs a similar aperture-based
method that incorporates MLC leaf positions and MU
weights as optimization parameters. The cost function is
based on dose-volume constraints and is defined using the
formulation proposed by Bortfeld et al.23 Minimum and
maximum dose constraints are specified as a function of vol-
ume 共e.g., 67 Gy to 95% volume兲 for target and healthy
tissue structures individually. For each constraint a relative
priority or importance value is also assigned. The cost is
calculated for each constraint using a standard quadratic dose
difference function multiplied by the priority value. The total
cost is the sum total of all individual constraint cost values.
II.B. Optimization constraints
During the optimization MLC leaf positions or MU
weights are constrained such that the aperture shapes and
MU values are physically achievable in practice. For ex-
ample, overlapping of opposing leaves or negative MU
冉 冊 冉 冊 冒冉 冊
dMU dMU d␪
= . 共3兲
d␪ max dt max dt max
In practice, a MU weight that causes the maximum dose rate
setting 共dMU/ dt兲max to be exceeded can be delivered by re-
ducing gantry rotation speed 共d␪ / dt兲. Deceleration of the
gantry is undesirable because it will result in increased treat-
ment time as well as potentially less accurate delivery due to
the substantial angular momentum of the linac apparatus.
The maximum MU weight is therefore also constrained
throughout the optimization to ensure that the maximum
dose rate is rarely exceeded.
Each iteration of the optimization involves randomly se-
lecting an available gantry sample, then changing either the
MU weight or a MLC leaf position for that sample. If a
proposed change does not violate a mechanical or efficiency
constraint the dose distribution and cost function are calcu-
lated. If the cost is reduced the change is accepted, otherwise
it is rejected.
II.C. Progressive sampling of gantry and MLC
positions
Of central importance to the VMAT algorithm is the
mechanism used to sample the dynamic source motion by a
finite number of static beams. All techniques that involve
source motion 共e.g., IMAT and tomotherapy兲 model the mov-
ing linac source as a series of static source position samples.
The instantaneous MLC configuration is defined at each

Medical Physics, Vol. 35, No. 1, January 2008

312 Karl Otto: Single arc radiation therapy
FIG. 1. Continuous source and MLC motion is initially modeled as a series of static source positions in 共a兲. As the optimization progresses new samples are
introduced with the first new sample placed between the first and second existing samples in 共b兲. Further samples are added in 共c兲 to span the full range of
gantry motion. The next sample is added at the beginning of the gantry range in 共d兲. Samples are continuously added in this way until a desired sampling
frequency is reached.
sample, and the MU setting for the sampling interval is as-
signed to that static MLC configuration. For example, a gan-
try rotation from 0 to 90 deg can be modeled as a series of
nine evenly spaced samples each covering a 10 deg range.
The source position and instantaneous MLC configuration
are defined at 5, 15, 25, 35, 45, 55, 65, 75, and 85 deg of
gantry rotation. MU values set at these source positions are
for the ranges 0–10, 10–20, 20–30, 30–40, 40–50, 50–60,
60–70, 70–80, and 80–90 deg, respectively. For the actual
delivery, the MLC leaves and gantry move linearly in be-
tween each sample. Clearly, the accuracy of the delivered
plan will depend on the amount MLC leaf positions change
and how far the gantry rotates in between each sample.10,24
At the start of VMAT optimization a relatively coarse
sampling of the gantry positions is used to model the gantry
rotation range. Samples are included at the beginning and
end of the range with evenly distributed samples in between
as depicted in Fig. 1共a兲. After a number of iterations 共MLC
and/or MU weight changes兲 an additional sample is added to
the pool of optimizable gantry positions. The new sample is
added midway between two existing samples. MLC posi-
tions for the new sample are linearly interpolated from the
MLC positions of the adjacent samples. The MU weight of
the new sample is a function of the MU weight of the adja-
cent samples. More precisely, when the new sample is intro-
duced the sampling interval covered by the two adjacent
samples is reduced to accommodate the new sample,
MUold共S − 1兲 MUold共S + 1兲
MUnew共S兲 = + , 共4兲
3 4
2 MUold共S − 1兲
MUnew共S − 1兲 = , 共5兲
3
3 MUold共S + 1兲
MUnew共S + 1兲 = , 共6兲
4
where MUnew, MUold, and S are the new MU weight, the old
MU weight 共prior to the new sample兲, and the sample index
of the new sample, respectively. Nonuniform sampling inter-
vals arise during the introduction of new samples 关see Figs.
1共b兲 and 1共d兲兴 and result in a nonuniform weighting of adja-
cent samples where new samples are being added. This effect
is accounted for by nonuniformly redistributing MU weights
Medical Physics, Vol. 35, No. 1, January 2008
312
for sample S + 1 and S − 1 as shown in Eqs. 共4兲–共6兲. Each
time a new sample is added, the VMAT algorithm continues
to optimize both the previous beam samples as well as the
newly added sample. Samples are further introduced using
the schedule depicted in Fig. 1共c兲. After the full gantry range
has been resampled 共effectively doubling the sampling fre-
quency兲, the process continues by returning to the start of the
gantry range as shown in Fig. 1共d兲. In this way the number of
samples is continually expanded until a desired sampling fre-
quency is attained.
Using a computer with 3.0 GHz Xeon processor each it-
eration 共dose and cost calculation兲 takes approximately 1 ms
while each sample addition takes approximately 1 s. There is
also a preprocessing step that takes approximately 5 min. All
calculation times depend on the size of the target volume共s兲
as well as the complexity of the case. Total optimization
times vary from 10 min to 1 h. Algorithms are currently
implemented in the MATLAB software prototyping environ-
ment. There are several noteworthy characteristics to this
approach. The first is with regard to the relationship between
the sampling interval and the efficiency constraints. The ef-
ficiency constraints defined in Eq. 共1兲 can be rewritten as
冉 冊
dx
⌬x ⱕ ⌬␪ ,
d␪ max
冉 冊
dMU
⌬MU ⱕ ⌬␪ . 共7兲
d␪
max
The MLC leaf position change 共⌬x兲 and MU weight 共⌬MU兲
permitted between samples is directly proportional to the
gantry rotation range 共⌬␪兲 between samples. When a coarse
sampling of gantry positions is used, there is a larger gantry
range between samples 共i.e., larger ⌬␪兲. Therefore, from Eq.
共7兲, a coarse sampling of gantry positions will allow for a
larger maximum MLC leaf displacement and MU weight
change at any iteration, thereby providing more flexibility in
deriving an optimal plan. Conversely, as new samples are
introduced ⌬␪ will decrease and reduce the flexibility of the
optimization. The overall effect is that VMAT optimization is
effectively not restricted by efficiency constraints at the start
but gradually becomes more restricted as new samples are
introduced. A similar characteristic related to the plan accu-
racy is also observed. Due to the large gantry range between
313 Karl Otto: Single arc radiation therapy 313

samples at the beginning of optimization the accuracy of the cost value. The degree to which optimization flexibility is
optimized plan with respect to the plan that would be deliv- reduced because of higher sampling was characterized by
ered will be relatively poor. As new samples are introduced determining the relationship between the gantry/MLC sam-
the sampling frequency increases, thereby improving the pling frequency and CPU time.
modeling accuracy of dynamic MLC and gantry motion. The
final number of optimized samples is chosen to ensure a II.F. Nasopharynx example
sampling frequency that provides adequate dosimetric accu-
The VMAT technique was applied to a nasopharynx car-
racy.
cinoma patient. During planning, minimum and maximum
Samples are introduced throughout the optimization using
dose-volume constraints were placed on three separate PTVs
a schedule that is a function of the optimization progress.
and multiple critical structures 共brainstem, spinal cord, pa-
The interval 共e.g., CPU time or number of successful itera-
rotids, brain, optical apparatus兲 and are identical to those
tions兲 between the addition of each new sample follows the
used for the patient’s fixed gantry clinical treatment plan
form of a decreasing exponential. When there are fewer
共Eclipse v6.5, Varian Medical Systems Inc.兲. A VMAT plan
samples, i.e., at the beginning of the optimization, the num-
was generated with the optimization time fixed at 60 min
ber of iterations between each sample addition is relatively
共3.2 GHz Xeon processor, 2 GB RAM兲. For comparison, 360
large. As the number of samples increases there are exponen-
deg arc plans using fixed sampling 共constant number of
tially fewer iterations between each new sample. The optimi-
samples兲 were also generated with a range of different sam-
zation terminates once all samples have been introduced.
pling frequencies. The final cost values of these plans were
Each time a sample is added the cost is recalculated with the
then compared to the final cost of the VMAT plan.
new beam configuration. The addition of a new sample will
Finally, plan quality and delivery time were evaluated by
generally perturb the optimization progress, causing a tem-
comparing the VMAT plan to a static gantry 7 beam IMRT
porary increase in the cost. The magnitude of this effect is
plan 共Eclipse––sliding window25兲. Identical constraints were
larger when there are fewer samples but becomes increasing
used in generating the VMAT and static gantry plans. VMAT
less significant as more samples are added. The schedule
plans are delivered on a Varian Cl21EX linac using research
described above is specifically designed to compensate for
control software provided by Varian Medical Systems Inc.
these perturbations by having exponentially more iterations
MLC leaf positions, gantry angles, and fractional MU values
between sample additions when there are fewer samples.
are transferred and monitored by the software throughout the
delivery. A 200 cGy fraction was delivered using 600 MU/
II.D. Dosimetric accuracy and sampling
min maximum dose rate setting.
To maximize dosimetric accuracy when approximating
dynamic source and collimation by multiple static beams, it III. RESULTS
is necessary to determine the minimum sampling frequency
III.A. Dosimetric accuracy and sampling
for gantry and MLC leaf positions. For this purpose a series
of VMAT plans was generated with varying sampling reso- The sampling requirements for VMAT planning were de-
lution. These plans were then compared to the same plan but termined using VMAT plans generated with a range of gantry
resampled at high frequency, which more closely approxi- and maximum MLC leaf sampling. For these plans the gan-
mates true dynamic motion. From these results an estimate try positions were interpolated to high resolution with 0.5
of the required sampling frequency is determined. deg of gantry rotation in between samples. Correspondingly,
the MLC leaves move a maximum of 0.25 cm between
II.E. Optimization time and fixed sampling samples as can be calculated from Eq. 共2兲. Dose calculation
error resulting from different sample spacing was determined
A study was performed to evaluate the effect of different
by comparing each plan with its corresponding highly
sampling frequencies on optimization time. Increasing the
sampled “reference” plan. The results of this study for the
sampling frequency will increase the number of optimizable
nasopharynx carcinoma patient are shown in Fig. 2. Other
gantry positions and restrict the maximum MLC leaf position
treatment sites showed similar results to those presented
changes in between consecutive samples 关Eq. 共6兲兴. The im-
here. From this study it was determined that a sampling of 1
pact of sampling on optimization time was assessed using a
deg and a maximum of 0.5 cm is required for gantry rotation
series of arc plans generated with fixed sampling. These
and MLC leaf motions, respectively, in order to ensure that
plans were created without the progressive sampling method
less than 5% of the volume has a dose error ⬎3%. This
共i.e., not VMAT兲 so that a fixed number of samples are used
threshold was chosen to coincide with typical IMRT quality
throughout the entire optimization. Gantry rotation is re-
assurance criteria. Note that the results of this study are in-
stricted to a single 360 deg arc in order to match the beam
dependent of any errors associated with the delivery device,
delivery geometry of VMAT. Optimization of leaf position
including modeling of MLC characteristics.
and MU weights was performed using a simulated
annealing-based method. For these plans, simulated anneal-
III.B. Optimization time and fixed sampling
ing is required so that the optimization does not become
trapped in a local minimum. The CPU time was recorded for Based on the results from the previous section, VMAT
each plan once the optimization had reached a predetermined plans are generated with gantry angles and MLC positions

Medical Physics, Vol. 35, No. 1, January 2008

314 Karl Otto: Single arc radiation therapy 314

FIG. 3. Optimization CPU time is plotted as a function of single arc gantry

FIG. 2. The percentage volume 共target and critical structures兲 exceeding angle sample spacing and maximum MLC sample spacing for desired final
10%, 5%, and 3% dose error is shown as a function of the gantry and cost values of 400 and 1200.
maximum MLC leaf sample spacing.

III.C. Nasopharynx example

Plotted in Fig. 4 is the final cost value for 60 min optimi-
sampled every 1 deg and a maximum 0.5 cm, respectively. zations using a single optimized arc with the conventional
There are therefore ⬃360 separate beam apertures used to fixed sampling technique as well as single arc optimization
optimize a VMAT plan. Theoretically, the large number of using the VMAT progressive sampling technique. The final
beams should provide increased flexibility in deriving a cost value for the VMAT plan is an order of magnitude lower
high-quality plan. Unfortunately, increased sampling reduces than the best result when fixed sampling 共8 deg gantry and a
the range of leaf motion between consecutive samples, and maximum 4 cm MLC leaf sample spacing兲 is used. Regard-
reduces the ability of the optimization algorithm to fully ex-
ploit the benefits of a large number of beams. Plotted in Fig.
3 is the CPU time vs gantry sample spacing for plans with
fixed sampling 共i.e. plans were generated without the VMAT
progressive sampling technique兲. As the sample spacing is
reduced the CPU time increases exponentially. Note that the
cost values used here are more than an order of magnitude
greater than an acceptable clinical plan. From these results it
can be extrapolated that it would take too long 共⬎1 month兲
to attain the cost value of an acceptable plan 共if achievable兲
with the required 1 deg gantry and a maximum 0.5 cm MLC
sampling.
Results of the studies described above characterize the
relationship between sampling, dose modeling accuracy, and
optimization time. In particular:
共1兲 Adequate dosimetric accuracy requires a high sampling
frequency of gantry and MLC positions.
共2兲 Greater sampling of MLC leaf positions increases opti-
mization CPU time exponentially.

Introducing samples throughout the optimization permits the

VMAT algorithm to rapidly approach a solution without
compromising dosimetric accuracy. Furthermore, a large
number of samples are 共eventually兲 available for deriving a
higher quality plan.
FIG. 4. Final cost values are plotted as a function of gantry angle sample
spacing and maximum MLC sample spacing for plans optimized with fixed
sampling. A plan with the VMAT progressive sampling technique 共final
gantry and MLC maximum sample spacing of 1 deg and 0.5 cm, respec-
tively兲 generated for the same patient is shown for comparison.

Medical Physics, Vol. 35, No. 1, January 2008

315 Karl Otto: Single arc radiation therapy 315

FIG. 5. VMAT and Eclipse isodoses 共95% of PTV70, PTV56, and 31.5 Gy兲
are shown on a representative CT slice for the nasopharynx example. Dose
distributions have substantially different shapes due to the use of 360 deg of
gantry rotation with VMAT and seven fixed beam directions with Eclipse.

less, the plan with 8 deg fixed sampling would result in an

unacceptable dosimetric error, as was shown in Fig. 2. It
should be noted that all of the fixed sampling plans grossly
failed to meet the requirements of an acceptable plan. The
VMAT plan met or exceeded the clinical objectives for all
targets and critical structures. Finally, it should be noted that
the 16 deg and maximum 8 cm MLC translation sampled
plan 共23 samples兲 fails to achieve the cost value of the 8 deg
and maximum 4 cm sampling plan 共45 samples兲. This shows
that there is a benefit to increasing the number of samples,
even though the MLC motions become more restrictive. Re-
ducing the sample spacing further to 4 deg and maximum 2
cm proves too restrictive and results in an inferior plan. The
dramatic difference in the final cost indicates that VMAT
optimization with progressive sampling is able to exploit the
benefits of a large number of samples while avoiding sam-
pling restrictions.
Representative CT slice dose distributions are shown for
VMAT and fixed gantry IMRT plans in Fig. 5. DVH com-
parisons for the nasopharynx case 共Fig. 6兲 show overlapping
DVHs 共indicating equivalent coverage兲 for the three target
volumes. The parotid glands exhibit a substantial reduction
at high and low doses for VMAT optimization. The same is
seen for the spinal cord with a somewhat smaller improve-
ment for the brainstem. Shown in Fig. 7 is the posterior
beam’s eye view of the VMAT instantaneous MLC aperture
shape visualized with target and healthy tissue structures. In
Fig. 7共a兲 the MLC is shown transparent to the targets, spinal
cord, brainstem, and parotid glands. In Fig. 7共b兲 the MLC
leaves are opaque, showing that the critical structures are
effectively blocked while the targets are receiving dose from
the open beam. Note that the collimator is rotated to 45 deg FIG. 6. DVHs for a nasopharynx cancer test patient 共VMAT and Eclipse兲.
in this case. Generally, 45 deg has been found to be prefer- Overlapping PTV DVHs show that almost identical target coverage is at-
able in most cases tested. Thus far, VMAT optimization has tained. VMAT critical structure DVHs show a substantial improvement over
fixed gantry IMRT.
shown similar dose distributions for other patients and treat-
ment sites; although a comprehensive planning study has not 共600 MU/min兲. Similar reductions in treatment times have
been performed yet. been seen for other patients and treatment sites with delivery
The delivery time for the VMAT treatment plan 共defined times ranging between ⬃1.5 and 3 min.
as the time between the start of the first beam to the end of
the last beam兲 shown in Figs. 5 and 6 was 107 s 共1.8 min兲. IV. DISCUSSION
For comparison, the patient’s fixed gantry clinical IMRT plan Radiation delivery with continuous gantry and MLC leaf
was delivered in 426 s 共7.1 min兲 at the same dose rate setting motion is modeled using a finite sampling of gantry and

Medical Physics, Vol. 35, No. 1, January 2008

316 Karl Otto: Single arc radiation therapy
FIG. 7. The posterior beam’s eye view of the VMAT instantaneous MLC
aperture shape visualized with target and healthy tissue structures at that
beam direction. In 共a兲 the MLC is shown transparent to the targets 共light
gray or green兲, spinal cord, brainstem, and parotid glands 共dark gray or red兲.
In 共b兲 the MLC leaves are opaque, showing that the critical structures are
effectively blocked while the targets are receiving dose from the open beam.
Note that the collimator is rotated to 45 deg in this case.
MLC leaf positions. The competing benefits of using a small
and large number of samples during plan optimization can be
mutually realized using the VMAT optimization technique.
By starting with a small number of samples and gradually
introducing new samples, a high-quality plan can be
achieved in a relatively short period of time. Also, a suffi-
cient number of samples ensures the preservation of dose
modeling accuracy.
One of the goals of VMAT is to deliver treatment plans
quickly and, as such, the optimization algorithms have been
structured to consider time efficiency. In considering time
efficiency it is desirable to have the radiation beam active for
the entire treatment session. More specifically, the amount of
time that the radiation output is reduced or interrupted due to
mechanical motion limits of gantry, MLC, and MLC carriage
should be at a minimum. Of course, this constraint must be
eased somewhat in order to produce a complex dose distri-
bution. The VMAT technique provides an effective compro-
mise between optimization flexibility and time efficiency by
restricting the delivery to a single 360 deg arc. Additional
restrictions are also considered for MLC leaf motions, gantry
rotation speed, and dose rate fluctuations so that no linac
component becomes a limiting factor in reducing the treat-
ment time. In the past, advances in dose distribution quality
have come at the cost of longer treatment times. This has
been the case for fixed gantry IMRT, IMAT, and tomo-
therapy. A remarkable feature of VMAT optimization is that
high-quality dose distributions can be delivered in substan-
tially less time than IMRT techniques and even some 3D
conformal methods.
VMAT is similar to tomotherapy in that the linac source
rotates through a 360 deg arc around the patient. The MLC
leaf width of 5 mm used here 共Varian Millennium 120 leaf
MLC兲 is also similar to the tomotherapy machine which has
a 6.25 mm leaf width 共at isocenter兲. Both techniques are
capable of modulating the dose rate for each beam direction,
although the mechanism used to perform the modulation is
different. The Varian linac varies the dose rate by directly
reducing or increasing radiation output, while tomotherapy
accomplishes the same effect by varying the MLC leaf open-
Medical Physics, Vol. 35, No. 1, January 2008
316
ing time at each beam direction. The most important differ-
ences between the two methods arise from the slice-by-slice
delivery of tomotherapy vs the single arc volumetric delivery
of VMAT. A slice-by-slice delivery is inherently less efficient
with regard to both radiation output requirements 共MU兲 and
treatment time. Because only a single two-dimensional dose
plane is generated for each 360 deg rotation, a three-
dimensional dose distribution must be created by superpos-
ing multiple overlapping two-dimensional dose distributions.
VMAT is considerably more efficient because the entire
three-dimensional dose distribution is created in a single ro-
tation. Furthermore, with tomotherapy the slice thickness
used to generate each dose slice is 1 cm or larger, which can
limit the spatial resolution of the dose distribution in the
superior-inferior direction. There is no slice thickness effect
with VMAT due to the volumetric nature of the delivery. A
final important difference is the binary positioning of MLC
leaves with tomotherapy vs continuous MLC leaf positioning
with VMAT. Although the tomotherapy binary MLC has the
advantage that MLC leaves can move quickly in and out of
the radiation beam, VMAT continuous MLC leaf positioning
provides more precision in beam aperture shaping. Further
investigation including treatment planning studies will be re-
quired to evaluate whether binary or continuous MLC mo-
tion has any significant advantage.
A feature of aperture-based techniques recognized by pre-
vious investigators is the substantial reduction in MU over-
fluence based optimization methods.21,22,26,27 Reducing the
number of MU has the dual benefit of reduced treatment time
and less total body scatter dose. This advantage is an addi-
tional factor that contributes to the efficiency of the VMAT
technique.
Previous aperture-based approaches to IMAT planning
have employed simulated annealing as the optimization
algorithm.11 This algorithm attempts to avoid local minima
by accepting some MLC leaf and MU weight changes that
result in a cost increase. The VMAT algorithm only accepts
changes that improve the quality of the plan. When a new
sample is introduced there will generally be a perturbation of
the dose distribution resulting in a cost increase. These per-
turbations may in fact be beneficial in a similar way to the
simulated annealing approach, where the likelihood of be-
coming trapped in a local cost function minimum is reduced
by allowing some changes that increase the cost during the
optimization. Although the exponential sample expansion
schedule described here was shown to be effective, other
schedules might result in further improvements in plan qual-
ity and/or time efficiency. For example, the rate at which the
samples are added could be a function of the rate of change
of cost. That is, each new sample would be introduced only
when the cost decrease per iteration becomes smaller than
some predefined threshold. A further example would be to
introduce new samples in random order instead of sequen-
tially throughout the gantry range, as was shown in Fig. 1.
VMAT currently uses a pencil beam-based dose calcula-
tion algorithm28,29 to model deposition. Other algorithms,
which model electronic disequilibrium more accurately,
could also be employed. Monte Carlo techniques are particu-
317 Karl Otto: Single arc radiation therapy
larly well suited to modeling dynamic source and MLC mo-
tion due to the particle-by-particle calculation framework.
The calculation time for a Monte Carlo simulation to achieve
a desired level of uncertainty is proportional to the number
of simulated particles and not the number of beams, making
it potentially more time efficient as well as more accurate
than existing algorithms for modeling dynamic source and
MLC motion.
Finally, it should be noted that the VMAT technique is not
restricted to a 360 deg arc. Reduced arc ranges or non-
coplanar arcs may be desirable in some cases 共e.g., for avoid-
ing beam directions where radiation is transmitted through
immobilization devices兲. Furthermore, there is no inherent
restriction on the use of overlapping arcs, although prelimi-
nary results indicate that doing so provides only minimal
improvements in plan quality.
Treatment site-specific planning studies must be per-
formed before any formal conclusions can be made regarding
the dose conformality capabilities of VMAT with respect to
other planning/delivery methods. This is the primary focus of
ongoing development. Thus far, results have shown that
treatment plans generated with VMAT optimization have
dose distributions that are equivalent to or superior to a com-
mercial static gantry IMRT system.
V. CONCLUSION
VMAT is a treatment planning and delivery platform us-
ing a single 360 deg gantry arc. Novel optimization tech-
niques using progressive gantry and MLC sampling are ca-
pable of producing accurate highly conformal dose
distributions with superior delivery efficiency. The introduc-
tion of new methods for on-line imaging and adaptation will
inevitably increase radiation therapy treatment times, with
accurate motion compensation somewhat limited by the
beam delivery time. Time studies have shown that VMAT
optimization is well suited for on-line verification and adap-
tation with delivery times that are reduced to ⬃1.5– 3 min
for a 200 cGy fraction.
ACKNOWLEDGMENT
This work was supported in part by Varian Medical Sys-
tems.
a兲
Electronic mail: kotto@bccancer.bc.ca
1
D. Yan et al., “Adaptive radiation therapy,” Phys. Med. Biol. 42, 123–132
共1997兲.
2 27
P. J. Keall et al., “Motion adaptive x-ray therapy: A feasibility study,”
Phys. Med. Biol. 46, 1–10 共2001兲.
3 28
H. Rehbinder, C. Forsgren, and J. Lof, “Adaptive radiation therapy for
compensation of errors in patient setup and treatment delivery,” Med.
Phys. 31, 3363–3371 共2004兲.
4 29
L. E. Court et al., “An automatic CT-guided adaptive radiation therapy
technique by online modification of multileaf collimator leaf positions for
prostate cancer,” Int. J. Radiat. Oncol. Biol. Phys. 62, 154–163 共2005兲.
Medical Physics, Vol. 35, No. 1, January 2008
317
5
B. C. Davis et al., “Automatic segmentation of intra-treatment CT images
for adaptive radiation therapy of the prostate,” in International Confer-
ence on Medical Image Computing and Computer Assisted Intervention
共2005兲, Vol. 8, pp. 442–450.
6
A. Mestrovic et al., “Direct aperture optimization for online adaptive
radiation therapy,” Med. Phys. 34, 1631–1646 共2007兲.
7
T. R. Mackie et al., “Tomotherapy: A new concept for the delivery of
dynamic conformal radiotherapy,” Med. Phys. 20, 1709–1719 共1993兲.
8
M. Carol et al., “Initial clinical experience with the Peacock intensity
modulation of a 3-D conformal radiation therapy system,” Stereotact.
Funct. Neurosurg. 66, 30–34 共1996兲.
9
C. X. Yu, “Intensity-modulated arc therapy with dynamic multileaf colli-
mation: An alternative to tomotherapy,” Phys. Med. Biol. 40, 1435–1449
共1995兲.
10
D. M. Shepard et al., “An arc-sequencing algorithm for intensity modu-
lated arc therapy,” Med. Phys. 34, 464–470 共2007兲.
11
M. A. Earl et al., “Inverse planning for intensity-modulated arc therapy
using direct aperture optimization,” Phys. Med. Biol. 48, 1075–1089
共2003兲.
12
A. Gladwish et al., “Segmentation and leaf sequencing for intensity
modulated arc therapy,” Med. Phys. 34, 1779–1788 共2007兲.
13
C. X. Yu et al., “Clinical implementation of intensity-modulated arc
therapy,” Int. J. Radiat. Oncol. Biol. Phys. 53, 453–463 共2002兲.
14
E. Wong, J. Z. Chen, and J. Greenland, “Intensity-modulated arc therapy
simplified,” Int. J. Radiat. Oncol. Biol. Phys. 53, 222–235 共2002兲.
15
W. De Gersem et al., “An anatomy-based beam segmentation tool for
intensity-modulated radiation therapy and its application to head-and-
neck cancer,” Int. J. Radiat. Oncol. Biol. Phys. 51, 849–859 共2001兲.
16
K. Bratengeier, “2-Step IMAT and 2-Step IMRT in three dimensions,”
Med. Phys. 32, 3849–3861 共2005兲.
17
C. Cameron, “Sweeping-window arc therapy: An implementation of ro-
tational IMRT with automatic beam-weight calculation,” Phys. Med. Biol.
50, 4317–4336 共2005兲.
18
S. M. Crooks et al., “Aperture modulated arc therapy,” Phys. Med. Biol.
48, 1333–1344 共2003兲.
19
M. A. MacKenzie and D. M. Robinson, “Intensity modulated arc deliv-
eries approximated by a large number of fixed gantry position sliding
window dynamic multileaf collimator fields,” Med. Phys. 29, 2359–2365
共2002兲.
20
W. De Gersem et al., “Leaf position optimization for step-and-shoot
IMRT,” Int. J. Radiat. Oncol. Biol. Phys. 51, 1371–1388 共2001兲.
21
D. M. Shepard et al., “Direct aperture optimization: A turnkey solution
for step-and-shoot IMRT,” Med. Phys. 29, 1007–1018 共2002兲.
22
M.-P. Milette and K. Otto, “Maximizing the potential of direct aperture
optimization through collimator rotation,” Med. Phys. 34, 1431–1438
共2007兲.
23
T. Bortfeld, J. Stein, and K. Preiser, “Clinically relevant intensity modu-
lation optimization using physical criteria,” in XII International Confer-
ence on the Use of Computers in Radiation Therapy, Salt Lake City, UT,
1997, pp. 1–4.
24
M. W. Kissick, T. R. Mackie, and R. Jeraj, “A delivery transfer function
共DTF兲 analysis for helical tomotherapy,” Phys. Med. Biol. 52, 2355–2365
共2007兲.
25
S. V. Spirou and C. S. Chui, “Generation of arbitrary intensity profiles by
dynamic jaws or multileaf collimators,” Med. Phys. 21, 1031–1041
共1994兲.
26
J. L. Bedford and S. Webb, “Constrained segment shapes in direct-
aperture optimization for step-and-shoot IMRT,” Med. Phys. 33, 944–958
共2006兲.
C. Cotrutz and L. Xing, “Segment-based dose optimization using a ge-
netic algorithm,” Phys. Med. Biol. 48, 2987–2998 共2003兲.
P. R. Storchi, L. J. van Battum, and E. Woudstra, “Calculation of a pencil
beam kernel from measured photon beam data,” Phys. Med. Biol. 44,
2917–2928 共1999兲.
P. Storchi and E. Woudstra, “Calculation of the absorbed dose distribution
due to irregularly shaped photon beams using pencil beam kernels derived
from basic beam data,” Phys. Med. Biol. 41, 637–656 共1996兲.