Home Search Collections Journals About Contact us My IOPscience

Creating a spread-out Bragg peak in proton beams

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2011 Phys. Med. Biol. 56 N131

(http://iopscience.iop.org/0031-9155/56/11/N01)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 130.194.20.173
This content was downloaded on 25/09/2013 at 12:04

Please note that terms and conditions apply.

IOP PUBLISHING PHYSICS IN MEDICINE AND BIOLOGY

Phys. Med. Biol. 56 (2011) N131–N138 doi:10.1088/0031-9155/56/11/N01

NOTE

Creating a spread-out Bragg peak in proton beams

David Jette1,2 and Weimin Chen1

1 ICT Radiotherapy Services, Livingston, NJ 07039, USA
2 Department of Medical Physics, Rush University, Chicago, IL 60612, USA

E-mail: [email protected]

Received 8 February 2011, in final form 9 April 2011

Published 10 May 2011
Online at stacks.iop.org/PMB/56/N131

Abstract
The model of Bortfeld and Schlegel (1996 Phys. Med. Biol. 41 1331–9) for
determining the weights of proton beams required to create a spread-out Bragg
peak (SOBP) gives a significantly tilted SOBP. However, by arbitrarily varying
its parameter p, which relates the range of protons to their energy, we have
been able to create satisfactory SOBPs. MCNPX Monte Carlo calculations have
been carried out to determine p, demonstrating the success of this modification.
Optimal values of p are tabulated for various combinations of maximum beam
energy E0 (50, 100, 150, 200 and 250 MeV) and SOBP width χ (15%,
20%, 25%, 30%, 35% and 40%), as well as for a correction factor needed
to calculate the SOBP dose. An example shows the application of these results
to analyzing the dose deposited by deuterons and alpha particles in broad proton
beams.
(Some figures in this article are in colour only in the electronic version)

1. Introduction

It is well known that a strictly monoenergetic proton beam is unsuitable for cancer treatment
because of its longitudinally narrow Bragg peak. Rather, it is necessary to ‘spread out’ the
Bragg peak to provide uniform dose within the target volume, by providing a suitably weighted
energy distribution of the incident beam. The problem of producing a spread-out Bragg peak
(SOBP) through a weighted collection of monoenergetic proton beams has been studied by
various investigators, in particular by Bortfeld and Schlegel (1996), Kooy et al (2003), Pedroni
et al (2005), Hérault et al (2005), Kooy et al (2005), Lu and Kooy (2006), Hérault et al (2007),
Kang et al (2007), Polf et al (2007), Kang et al (2008) and Schell and Wilkens (2009).
Bortfeld and Schlegel (1996) have developed a simple way for determining the weights
of proton beams of various initial energies required to create an SOBP for a proton beams.

0031-9155/11/110131+08$33.00 © 2011 Institute of Physics and Engineering in Medicine Printed in the UK N131
N132 D Jette and W Chen

15

Proton dose (MeV-cm2/g)

10

0
0 5 10 15 20
Depth (cm)

Figure 1. Expected dose in water of a collection of normally incident unidirectional monoenergetic

broad beams weighted according to the method of Bortfeld and Schlegel. The maximum energy
of the proton beams is 150 MeV, and the intended SOBP width χ is one-quarter of the range of
the beam. The solid curve is for a proton depth–dose model without range straggling, while the
dashed curve includes range straggling.

Their work is based on a close power-law relationship between the range R of protons and
their energy E:

R = αE p0 , (1)

where α = 0.0022 and p0 = 1.77. The values of these two parameters are properly taken from
a subsequent study (Bortfeld 1997) which develops analytic incorporation of range straggling
into the Bragg curve; the present authors have used Bortfeld’s range-straggling work to model
the dose from a proton beam with an initial Gaussian angular distribution (Jette et al 2010).
The method of Bortfeld and Schlegel was extended by Kooy et al (2003) to take into account
finite beam source-to-axis distance (SAD), but in this work we consider only the original
model of Bortfeld and Schlegel for infinite SAD, i.e. for a broad parallel beam.
As an example of this method of producing an SOBP, the solid curve of figure 1 shows the
expected depth dose for a broad proton beam of maximum energy E0 = 150 MeV penetrating a
water phantom, with an intended SOBP width χ which is one-quarter of the full range R0 of the
beam. This is for a collection of monoenergetic unidirectional broad proton beams normally
incident upon the irradiated material, without range straggling, and the SOBP is seen to be
practically flat. However, when these same weights are used for Bortfeld’s model of proton
depth dose with range straggling, the predicted SOBP is tilted, as seen in the dashed curve of
figure 1. (The area under these two curves is proportional to the total energy deposited by the
component proton beams. It is less under the dashed curve than under the solid curve because
the range-straggling model for those component beams includes loss of primary protons.)
Unfortunately, as will be found, this tilting of the SOBP is considerably more marked
for Monte Carlo calculations, increasing for increasing E0 and for increasing χ . However, it
turns out that a simple modification of the beam weights proposed by Bortfeld and Schlegel,
dependent upon E0 and χ , suffices to produce a flat SOBP. This modification is the subject of
the present work.
Creating a spread-out Bragg peak in proton beams N133

2. Creating the SOBP

We start by rewriting (without changing) the weight formula of Bortfeld and Schlegel, at the
same time correcting an error concerning the weight for the lowest energy used. (This error
occurs because, for the lowest energy with peak at R = da in their notation, they obtain the
weight by integrating their equation (4) from da − /2 to da + /2 rather than from da to
da+/2, where da is lower limit of the SOBP and  is the width of the interval between
adjacent component Bragg peaks. However, they do correctly obtain the weight for the largest
energy used, at depth db, by integrating their equation (4) from db − /2 to db. Otherwise,
for component Bragg peaks at depths d they integrate from d − /2 to d + /2.) The range
of the proton beam for its maximum energy E0 is R0, and the SOBP is to extend from depth (1
− χ )·R0 to depth R0, so that the fraction of the full range taken by the SOBP is χ . The SOBP
is divided into n equal intervals, and monoenergetic beams of range equal to the depth of the
ends of these intervals are used, with weights given by wk for k = 0, 1, . . . , n. Thus the ranges
rk used are
   
k
rk = 1 − 1 − χ R0 , (2)
n
and the corresponding energies ek are found from equation (1):
 r  p1
k
ek = 0
. (3)
α
The (normalized) weights are then
⎧  
⎪
⎪ 1 1−1/p
⎪1 − 1 −
⎪ k=0
⎪
⎪ 2n
⎪
⎨  1−1/p   1−1/p
1 1 1 1
wk = 1− k− − 1− k+ k = 1, . . . , n − 1 (4)
⎪
⎪ n 2 n 2
⎪
⎪  1−1/p
⎪
⎪ 1
⎪
⎩ k = n.
2n
Evidently these weights are independent of E0 and χ . They depend upon the parameter p
which appears in the power law of equation (1). For figure 1 we have used p = p0 = 1.77, but
we shall be allowing p to vary in order to produce a flat SOBP.
Finally, for a broad beam of unit fluence, the expected constant dose in the region of the
flat SOBP is
π (p0 − 1)
DSOBP = , (5)
2 1/p0
ρp0 α sin (π/p0 ) (χ R0 )1−1/p0
where ρ is the density of the irradiated material and R0 is the range of the proton beam (i.e. of
protons of the maximum energy E0.) Since p0 = 1.77, for water we have
25.0 358
DSOBP = 0.435
= 0.435 0.77 (6)
(χ R0 ) χ E0
in units of MeV cm2 g−1 (with E0 in MeV). Thus if a constant dose D0 in the region of the flat
SOBP is desired, the broad-beam fluence should be D0/DSOBP in accord with equation (5) of
Bortfeld and Schlegel (1996).

3. Results

We have carried out broad-beam MCNPX Monte Carlo calculations (Pelowitz 2008, Hérault
et al 2005, 2007) for various combinations of E0, χ and p. The MCNPX transport used the
N134 D Jette and W Chen

40

Proton dose (MeV-cm2/g)

30

20

10

0
0 1 2 3
Depth (cm)

Figure 2. Depth dose for SOBP configurations of maximum energy E0 = 50 MeV. The SOBP
widths are χ = 15%, 25% and 40%. The solid curves have the value of p given in table 1, and the
corresponding dotted curves are for p = 1.77.

LA150 data library (created specifically for MCNPX) for proton energies up to 150 MeV,
and the Bertini/Dresner physics model (the MCNPX default) for higher proton energies.
Elastic (multiple scattering) and non-elastic interactions of the incident protons were included,
specifically for the production of secondary protons, deuterons, alpha particles, neutrons,
photons, tritons and 3He particles. For figures 1–4, which involve only proton transport, we
ran 106 histories, providing standard deviations in the depth dose values of around 0.2% or
less. Figures 5 and 6 include production and transport of deuterons and alpha particles, and
2 × 107 histories were run to obtain those results. The beam weights calculated through
equation (4) were for n = 10; thus, the SOBP Monte Carlo runs each used 11 (monoenergetic)
proton-beam components. One hundred longitudinal intervals were used for each beam
energy, but with finer resolution in the SOBP region; for example, for a 150 MeV proton
beam there were 66 intervals of 0.2 cm each (0–13.2 cm) and then 34 intervals of 0.1 cm each
(13.2–16.6 cm).
Five maximum energies E0 of the proton beam were used: 50, 100, 150, 200 and
250 MeV. And six SOBP widths χ were used: 15%, 20%, 25%, 30%, 35% and 40%. For
each of these 30 combinations of E0 and χ , we varied p in equation (4) until a reasonably flat
SOBP was obtained. Typically we made four or five such runs, varying p by 0.01 on each
side of the optimal value. This optimal value was often well defined within 0.01 variation of
p, and a difference of 0.02 in p was always noticeable. (On one side of the optimal value of p
the SOBP tilts one way, and the other way on the other side of p.) These optimal values of p
are given in table 1, and they are close enough to each other that interpolation is easily carried
out.
Figures 2–4 show the Monte Carlo depth dose curves for the optimal values of p, for
E0 = 50, 150 and 250 MeV, and for χ = 15%, 25% and 40%. Also shown in these figures
are the corresponding curves for p = 1.77. Clearly the modification of p in calculating the
weights through equation (4) is necessary, and the results are very good, although there is a
small dip in the SOBP for E0 = 250 MeV.
Creating a spread-out Bragg peak in proton beams N135

20

Proton dose (MeV-cm2/g)

15

10

0
0 5 10 15 20

Depth (cm)

Figure 3. Same as figure 2, but for E0 = 150 MeV.

10

8
Proton dose (MeV-cm2/g)

0
0 10 20 30 40
Depth (cm)

Figure 4. Same as figure 2, but for E0 = 250 MeV.

Table 1. Optimal values of the parameter p for various widths χ of the spread-out Bragg peak, for
various values of the maximum energy of the proton beam. χ is the fraction of the entire range of
the beam which is occupied by the spread-out Bragg peak.

Maximum energy
E0 (MeV) χ = 15% χ = 20% χ = 25% χ = 30% χ = 35% χ = 40%

50 1.48 1.45 1.43 1.43 1.42 1.41

100 1.46 1.43 1.42 1.41 1.40 1.38
150 1.43 1.40 1.39 1.37 1.36 1.35
200 1.40 1.37 1.34 1.33 1.32 1.30
250 1.34 1.32 1.29 1.27 1.26 1.24
N136 D Jette and W Chen

0.3

Broad-beam dose (MeV-cm2/g)

0.2

0.1

0
0 5 10 15 20
Depth (cm)

Figure 5. Dose deposited by various particles in a broad beam of 150 MeV protons. Solid curve:
protons (both primary and secondary), with the dose multiplied by 0.01. Dashed curve: deuterons.
Dotted curve: alpha particles.

0.15
Broad-beam dose (MeV-cm2/g)

0.1

0.05

0
0 5 10 15 20
Depth (cm)

Figure 6. Same as figure 5, but for an SOBP configuration (p = 1.39) of width χ = 25% and
maximum beam energy E0 = 150 MeV.

There remains the question of how accurately equation (6) provides the constant dose in
the region of the SOBP. In fact, it is somewhat high, and table 2 gives the correction factor
which must be multiplied by DSOBP to obtain the actual MCNPX result.
As pointed out by Bortfeld and Schlegel (1996), knowing the weights required to create
an SOBP can be quite helpful in practical applications, either directly or as a good initial guess
for iterative calculations. One use to which the present authors have put these results concerns
the production of deuterons in proton beams (Jette and Chen 2011). As seen in figure 5, the
deuteron dose in a broad beam of 150 MeV protons penetrating water rises to about 2% of the
incident-proton dose at the surface and then drops sharply to zero before the depth of the Bragg
peak of the protons. The question naturally arises as to whether this phenomenon also occurs
for an SOBP configuration, and figure 6 for an SOBP width χ = 25% demonstrates that indeed
Creating a spread-out Bragg peak in proton beams N137

Table 2. Correction factors for computing the SOBP dose from equation (6).

Maximum energy
E0 (MeV) χ = 15% χ = 20% χ = 25% χ = 30% χ = 35% χ = 40%

50 0.95 0.95 0.96 0.96 0.97 0.97

100 0.90 0.91 0.92 0.92 0.93 0.94
150 0.84 0.86 0.87 0.88 0.89 0.89
200 0.78 0.80 0.82 0.84 0.84 0.87
250 0.72 0.75 0.77 0.80 0.81 0.84

it does. Not surprisingly, the height and the shape of the deuteron curve stay pretty much the
same because the deuteron dose is near zero in the Bragg peak region, while the maximum
height of the proton curve is halved in going from a monoenergetic beam to an SOBP one.
(In both figures, the incident proton fluence is the same: one proton cm−2.) Thus, whereas
the peak deuteron dose relative to the peak proton dose is about 0.5% for a monoenergetic
proton beam, it is instead about 1% of the treatment dose when a realistic SOBP configuration
is used.

4. Conclusion

We have successfully modified the method of Bortfeld and Schlegel (1996) to determine the
weights of proton beams of various initial energies required to create a spread-out Bragg
peak (SOBP) for proton beams. The flatness of the resulting SOBPs is generally very good,
although for the highest beam energies they do exhibit a small dip. We also have tabulated
the correction factor needed to obtain the constant dose in the region of the flat SOBP. It is
perhaps remarkable that the method of Bortfeld and Schlegel does so well with only this simple
modification. These results can be quite helpful in practical applications, either directly or as
a good initial guess for iterative calculations.

References

Bortfeld T 1997 An analytical approximation of the Bragg curve for therapeutic proton beams Med. Phys. 24 2024–33
Bortfeld T and Schlegel W 1996 An analytic approximation of depth–dose distributions for therapeutic proton beams
Phys. Med. Biol. 41 1331–9
Hérault J, Iborra N, Serrano B and Chauvel P 2005 Monte Carlo simulation of a protontherapy platform devoted to
ocular melanoma Med. Phys. 32 910–19
Hérault J, Iborra N, Serrano B and Chauvel P 2007 Spread-out Bragg peak and monitor units calculation with the
Monte Carlo code MCNPX Med. Phys. 34 680–8
Jette D and Chen W 2011 Secondary particles in proton beams: deuterons, in preparation
Jette D, Yuan J and Chen W 2010 Oblique incidence for broad monoenergetic proton beams Med. Phys. 37 5683–90
Kang J H, Wilkens J J and Oelfke U 2007 Demonstration of scan path optimization in proton therapy Med.
Phys. 34 3457–64
Kang J H, Wilkens J J and Oelfke U 2008 Non-uniform depth scanning for proton therapy systems employing active
energy variation Phys. Med. Biol. 53 N149–55
Kooy H M, Rosenthal S J, Engelsman M, Mazal A, Slopsema R, Paganetti H and Flanz J B 2005 The prediction of
output factors for spread-out proton Bragg peak fields in clinical practice Phys. Med. Biol. 50 5847–56
Kooy H M, Schaefer M, Rosenthal S and Bortfeld T 2003 Monitor unit calculations for range-modulated spread-out
Bragg peak fields Phys. Med. Biol. 48 2797–808
Lu H-M and Kooy H 2006 Optimization of current modulation function for proton spread-out Bragg peak fields Med.
Phys. 33 1281–7
N138 D Jette and W Chen

Pedroni E, Scheib S, Böhringer T, Coray A, Grossmann M, Lin S and Lomax A 2005 Experimental characterization
and physical modelling of the dose distribution of scanned proton pencil beams Phys. Med. Biol. 50 541–61
Pelowitz D B 2008 MCNPX user’s manual, v. 2.6.0 Publication LA-CP-07-1473 (Los Alamos, CA: Los Alamos
National Laboratory) htpp://mcnpx.lanl.gov
Polf J C, Harvey M C and Smith A R 2007 Initial beam size study for passive scatter proton therapy. II. Changes in
delivered depth dose profiles Med. Phys. 34 4213–8
Schell S and Wilkens J J 2009 Modifying proton fluence spectra to generate spread-out Bragg peaks with laser
accelerated proton beams Phys. Med. Biol. 54 N459–66