photon and photoneutron spectra produced in radiotherapy ... · al., 2009, hernandez-adame et al.,...

XII International Symposium/XXII National Congress on Solid State Dosimetry

September 5th to 9th, 2011. Mexico city

237

Photon and photoneutron spectra produced in radiotherapy LINACs

H.R. Vega-Carrillo1, S.A. Martínez-Ovalle2, J.L. Benites-Rengifo3 and A.M. Lallena4

1Unidad Academica de Estudios Nucleares

Universidad Autónoma de Zacatecas C. Cipres 10, Fracc. La Peñuela 98068 Zacatecas, Zac. Mexico

2Grupo de Física Nuclear Aplicada y Simulación

Universidad Pedagógica y Tecnológica de Colombia Avenida Central del Norte Km.1. Vía Paipa

Tunja, Boyacá, Colombia

3Universidad Autonoma de Nayarit

Postgrado CBAP Xalisco, Nayarit, Mexico

4Dpto. de Física Atómica, Molecular y Nuclear

Universidad de Granada E-18071 Granada. SPAIN

e-mail: [email protected]

Abstract

A Monte Carlo calculation, using the MCNPX code, was carried out in order to estimate the photon and neutron spectra in two locations of two linacs operating at 15 and 18 MV. Detailed models of both linac heads were used in the calculations. Spectra were estimated below the flattening filter and at the isocenter. Neutron spectra show two components due to evaporation and knock-on neutrons. Lethargy spectra under the filter were compared to the spectra calculated from the function quoted by Tosi et al. that describes reasonably well neutron spectra beyond 1 MeV, though tends to underestimate the energy region between 10-6 and 1 MeV. Neutron and Bremsstrahlung spectra show the same features regardless of the linac voltage. The amount of photons and neutrons produced by the 15 MV linac is smaller than that found for the 18 MV linac. As expected, Bremsstrahlung spectra ends according to the voltage used to accelerate the electrons. Keywords: Neutron, photon, spectrum, LINAC, Monte Carlo



238

Introduction

Worldwide, the number of cancer cases is growing. Nowadays cancer kills more

people than tuberculosis, HIV, and malaria combined. Developing countries, where

resources to prevent, diagnose and treat cancer are limited or nonexistent, are

worst hit by the cancer crisis (IAEA, 2010).

Cancer patients are treated by radiation alone or combined with surgery and/or

systemic therapy including chemical, immunological and genetic treatments.

Radiation can be applied as systemic radiotherapy, teletherapy, or brachytherapy.

In teletherapy a radiation beam is addressed to the patient aiming at killing cancer

cells, unaffecting the normal tissues around the tumor. Thus, the classical radiation

therapy, CRT, was improved to 3D-CRT, evolving to Intensity-Modulated Radiation

Therapy, IMRT. Changes included the multileaf collimator system, MLC, and

advanced optimization algorithms (Xu, Bednarz and Paganetti, 2008).

Despite there are several options for cancer treatment, radiotherapy with linacs is

the technique most commonly used, either with electron or photon beams

(Mesbahi et al., 2010).

A lower skin dose, a higher depth dose, a reduced scattered dose to healthy

tissues outside the tissue-target volume, and less rounded isodose curves, are

some of the advantages provided by the use of high-energy x-rays instead of low

energy photons (Brusa et al., 2008, Hashemi et al., 2007). However using high-

energy x-ray machines is a radiation protection issue because besides the

therapeutic beam, linacs produce neutrons that induce reactions in materials inside

the bunker, such as activation (Polaczek et al., 2010) and prompt -ray reactions.



239

Neutrons deliver an undesirable dose to the patient, this dose can cause

secondary cancers.

Working above 8 MV, linacs produce neutrons regardless they operate with

electrons or x-rays. In case of electrons, neutrons are produced via virtual photon

reactions (e, e’n), while x-rays generate neutrons in (, n) nuclear reactions. Cross

sections for photoneutron production are roughly between 100 and 200 times

larger than for electro-neutron production, therefore studies have been focused in

photoneutrons (Martinez-Ovalle et al., 2011, Lin et al., 2001, McGinley 1998,

NCRP, 1984).

Photoneutrons are mainly induced in the high-Z materials on the linac head such

as the target, the flattening filter, the jaws, the MLC and the head shielding; also,

they are produced in the patient body and materials in the treatment hall whose

cross sections show the well known Giant Dipole Resonance (Hall et al., 1995,

IAEA, 2000)

Photoneutrons are produced when photons collide with a nucleus, its energy is

distributed among the nucleons and, eventually, a neutron near the nucleus

surface acquires enough energy to emerge as an evaporation neutron. These

neutrons are emitted isotropically. Another reaction occurring during photon-

nucleus reactions and that contributes also to the neutron productionis that in

which the photon gives out all its energy to a single neutron that is kicked-out from

the nucleus. These neutrons are mostly emitted in the direction of the incoming

photon and are known as knock-on neutrons. To describe the photoneutron

spectrum Tosi et al., (1991) derived the following expression:



240

nmax SE

0 n

max

n

max

2

dESE

Eln

SE

Eln

T

Eexp

T

EEn (1)

Here, is the fraction of evaporation neutrons, T is the target nuclear temperature,

is the fraction of knock-on neutrons, Emax is the maximum energy of accelerated

electrons and Sn is the separation energy of neutrons in the target nucleus. For W,

= 0.8929, = 0.1071, T = 0.5 MeV, and Sn = 7.34 MeV. For the Bremsstrahlung

Cu target, Sn = 10.56 MeV and T = 1 MeV (Agosteo and Foglio, 1994, Agosteo et

al., 1995).

The actual geometry and composition of the linac head are not always available.

For neutron transport studies around a linac using Monte Carlo methods, some

simplifications are usually assumed. Thus a point-like photoneutron source whose

energy distribution is given by Equation (1) is located at the centre of a 10 cm-thick

spherical shell, with 20 cm external radius, for W head, and 15 cm-thick spherical

shell for Pb head (Agosteo et al, 1995). According to the American Association of

Physics in Medicine (AAPM, 1986), the W head is preferred, because the Pb head

provides a dose enhancement of a factor 2.

This simple head model has been successfully used (Facure et al., 2005, Followill

et al., 2003) even with a thicker W sphere and different models for knock-on

photoneutrons (Barquero et al., 2005). In particular, it has been pointed out that the

results obtained using this simple model in Monte Carlo calculations show no

significant differences with those found for detailed geometries of the linac head

(Agosteo et al., 1995, Carinou, Kamenopoulou and Stamatelatos, 1999). On the

other hand, photoneutron features, such as thermal fluences, spectra, absorbed



241

dose, and ambient dose equivalents have been investigated either using

experimental procedures or Monte Carlo calculations (Barquero et al., 2005, Kry et

al., 2009, Hernandez-Adame et al., 2011, Vega-Carrillo and Baltazar-Raigoza,

2011).

The aim of this work is to calculate, using Monte Carlo methods, the photon and

neutron spectra of two linac heads, and to compare the neutron spectra with the

analytical function of Tosi et al., (1991).



242

Materials and methods

Spectra were calculated using the Monte Carlo code MCNPX code (Pelowitz,

2005). Detailed head models of the 15 MV Siemens Primus and the 18 MV Varian

Clinac 2100 C/D were used. Figures 1 and 2 show the geometry of the main

components of both linacs.

Figure 1. Target, compensator, flattening filter and jaws models for the 15 MV

Simens Primus linac

In the calculations an electron beam hits the 0.1 cm-diameter target. Electron

energies were sampled from a Gaussian distribution with a mean energy of 15.3

MeV and a standard deviation of 1.5 MeV. These values were fixed by comparing

the calculated percentage depth dose, PDD, with the experimentally measured in a

water phantom (Martinez-Ovalle et al., 2011). This procedure was also used for the

18 MV Varian Clinac 2100 C/D where a 18.3 MeV electron beam was used. Figure

3 shows the comparison between calculated and measured PDDs for both linacs.



243

The geometry of both linacs included the shielding cover that was built up using Pb

and Fe.

Figure 2. 18 Varian Clinac 2100 C/D geometry model

Neutron spectra were tallied using point-like detectors located below the flattening

filter and at the isocentre. In order to reduce the computational time, calculations

were carried out in two steps. First a unidirectional electron beam was made to

collide with the target and the Bremsstrahlung spectrum was calculated in both

detectors, using the mode e/. Second, the x-ray spectrum obtained in the detector

located under the flattening filter was used as source term, in mode e//n, and the

photoneutron spectrum was estimated in both detectors.

For all calculations the amount of histories was large enough to have uncertainties

below 1%. Spectra were calculated using two energy arrays with 47 (MB) and 26

(FB) energy groups.



244

In order to compare the Tosi’s function with the photoneutron spectra obtained

here, lethargy spectra, E E(E), were calculated. In the case of the Tosi’s function

the lethargy spectrum was obtained using equation 2.

Figure 3. Calculated and measured PDDs for the 15 MV Siemens Primus and the

18 MV Varian Clinac 2100 C/D linacs



245

E E(E) = E n(E) , (2)

Here, n(E) is given by equation (1). For the spectra here calculated, that are

discrete, we used.

l

u

EE

E

Eln

EEE

(3)

In equation (3), Eu and El are the upper and lower energy bounds of the class

interval for each energy group.



246

Results

As said above, photoneutron spectra in linacs include the contribution of

evaporation and knock-on neutrons, with a smaller contribution coming from (,nn)

and (,pn) reactions. In figure 4, the neutron spectra for the 15 MV LINAC, in the

detectors below the filter (in red) and at the isocenter (in blue) are shown.

Figure 4. 15 MV linac photoneutron spectra below the filter and at the isocenter

In the region between 0.1 and 10 MeV, two peaks with maxima around 0.3-0.5

MeV and 1.2-2 MeV are present. Neutrons whith energies between 1 eV and 1

MeV are the evaporation neutrons while knock-on neutrons appear above 1 MeV.

15 MV

Neutron energy [ MeV ]

10-6 10-5 10-4 10-3 10-2 10-1 100 101

E(E

)

[

cm

-2 -

MeV

-1

]

10-15

10-14

10-13

10-12

10-11

10-10

10-9

10-8

10-7

Filter

Isocenter



247

The upper energy in these spectra is 6 MeV. In figure 5 the corresponding spectra

for the 18 MV linac are plotted. They show the same features as those found for

the 15 MV linac, except that in this case the upper energy is 9 MeV. Also, a larger

amount of neutrons per photon are produced in this case.

18 MV

Energy [ MeV ]

10-6 10-5 10-4 10-3 10-2 10-1 100 101

E(E

)

[

cm

-2 -

MeV

-1

]

10-15

10-14

10-13

10-12

10-11

10-10

10-9

10-8

10-7

10-6

Filter

Isocenter

Figure 5. 18 MV linac photoneutron spectra below the filter and at the isocenter

For both linacs, the difference between the spectra below the flattening filter and at

the isocenter is mainly due to the distance between both detectors’ locations.

In figure 6, the lethargy spectrum calculated with the function of Tosi et al. is

compared to those obtained in our simulations. Both MB and FB spectra are

shown.



248

Figure 6. Photoneutron lethargy spectra below the flattening filter for the 15 MV

linac. The spectra obtained from the function of Tosi et al. is shown in

red. The MB and FB Monte Carlo results are shown in blue and black,

respectively

The Tosi et al spectrum shows evaporation neutrons from 1 eV up to 4 MeV, with a

maximum for 1 MeV; the small tail above 4 MeV corresponds to knock-on

neutrons. Both Monte Carlo spectra have the maximum shifted towards slightly

lower energies having a larger contribution for all energies below that of the

maximum.

15 MV

Energy [ MeV ]

10-6 10-5 10-4 10-3 10-2 10-1 100 101 102

E

E(E

)

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

Tosi et al., function

Monte Carlo MB

Monte Carlo FB



249

In figure 7, the results found for the 18 MV linac are shown. Similar features, as

those for 15 MV linac, are noticed; however differences are larger in this case in

agreement with the findings of Huang et al. (2005).

18 MV

Energy [ MeV ]

10-6 10-5 10-4 10-3 10-2 10-1 100 101 102

E

E(E

)

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

Tosi et al., function

Monte Carlo MB

Monte Carlo FB

Figure 7. Same as figure 6 but for the 18 MV linac

In figure 8 the photon spectra for 15 and 18 MV linacs calculated below the

flattening filter and at the isocentre are shown. For both linacs the Bremsstrahlung

spectra estimated under the flattening filter have a peak near 0.5 MeV and a

continuous energy spectrum up to the incident electron energy (Patil et al., 2011).

As the electrons colliding with the target are slowed down producing heat and



250

Bremsstrahlung x-rays that are continuous. Due to the distance between the target

and the isocenter, the number of photons that reach the isocenter is strongly

reduced and this effect is most visible in case of the lower energy photons.

Regardless of the linac voltage, the photon maximum energy at the isocenter is

approximately 1.5 MeV. Spectra at this position are into agreement with spectra

reported in literature (Patil et al., 2011, Sheikh-Bagheri and Rogers, 2002).

Energy [ MeV ]

0 5 10 15 20

Ph

oto

n f

lue

nc

e

[

c

m-2

]

10-9

10-8

10-7

10-6

10-5

10-4

10-3

18 MV by the filter

15 MV by the filter

18 MV at the IC

15 MV at the IC

Figure 8. Photon spectra below the filter and at the isocenter for both linacs



251

Conclusions

Detailed models of 15 MV Siemens Primus and 18 MV Varian Clinac 2100C/D

heads were used to estimate the photon and neutron spectra below the flattening

filter and at the isocenter. The spectra were calculated using the Monte Carlo code

MCNPX. The features of the incident electron beam were tuned by comparing the

PDDs calculated with the measured ones.

For both linacs, and for both locations, neutron spectra has the same features: they

present two peaks around 0.3-0.5 MeV and 1.2-2 MeV. Evaporation neutrons

appear between 1 eV and 1 MeV, whilst knock-on neutrons are beyond 1 MeV.

Spectra at the isocenter are reduced with respect to those below the filter mainly

due to the distance between both locations.

Neutron lethargy spectra were compared with those obtained from the function of

Tosi et al. that has its maximum at 1 MeV. In both linacs, the agreement is

reasonable above 1 MeV, though the spectra found for the function of Tosi et al.

underestimate the Monte Carlo spectra below this energy.

For both linacs, Bremsstrahlung spectra calculated below the flattening filter have a

peak near 0.5 MeV and a continuous spectra that, as expected, end at the

maximum electron energy. At the isocenter, Bremsstrahlung spectra have the

maximum between 1 and 2 MeV, regardless the accelerating voltage. The 15 MV-

linac continuous spectrum is smaller than the 18 MV spectrum.



252

References

AAPT. 1986. Neutron measurements around high energy x ray radiotherapy machines. American Association of Physicist in Medicine Report 19. New York.

Agosteo, S., Foglio Para, A., 1994. Energy and spatial dependence of neutron fluxes in radiotherapy romos for a simple dose estimate method. Nucl. Instr. Meth. Phys. Res. B, 93, 362-369.

Agosteo, S., Foglio Para, A., Maggioni, B., Sangiust, V., Terrani, S., Borasi, G., 1995. Radiation transport in a radiotherapy room. Health Phys., 68, 27-34.

Barquero, R., Edwards, T., Iñiguez, M.P., Vega-Carrillo, H.R., 2005. Monte Carlo simulation estimates of neutron doses to critical organs of a patient undergoing 18 MV X-ray LINAC-based radiotherapy. Med. Phys., 32, 3579-3588.

Barquero, R., Mendez, R., Vega-Carrillo, H.R., Iniguez, M.P., Edwards, T.M., 2005. Neutron spectra and dosimetric features around an 18 MV linac accelerator. Health Phys., 88, 48-58.

Brusa, A., Cesana, A., Stucchi, C., Terrani, M., Zanellati, F., 2008. Long term activation in a 15 MeV radiotherapy accelerator. Med. Phys., 35, 3049-3053.

Carinou, E., Kamenopoulou, V., Stamatelatos, I.E., 1999. Evaluation of neutron dose in the maze of medical electron accelerators. Med. Phys., 26, 2520-2525.

Facure, A., Falcao, R.C., Silva, A.X., Crispim, V.R., Vitorelli, J.C., 2005. A study of neutron spectra from medical linear accelerators. Appl. Radiat. Isot., 62, 69–72.

Followill, D.S., Stovall, M.S., Kry, S.F., Ibbott, G.S.J., 2003. Neutron source strength measurements for Varian, Siemens, Elekta, and General Electric linear accelerators. J. Appl. Clin. Med. Phys., 4, 189-194.

Hall, E.J., Martin, S.G., Almols, H., Hei, T.K., 1995. Photoneutrons from medical linac accelerators – Radiobiological measurements and risk estimates. Int. J. Radiat. Oncol. Biol. Phys., 33, 225-230.

Hashemi, S.M., Hashemi-Malayeri, B., Raisali, G., Shokrani, P., Sharafi, A.A., 2007. A study of the photoneutrons dose equivalent resulting from a Saturne 20 medical linac using Monte Carlo method. Nukleonika, 52, 39-43.

Hernandez-Adame, L., Contreras-Sandoval, H., Vega-Carrillo, H.R., Perez, L.H., 2011. Design of a treatment room for an 18 MV linac. Nucl. Tech., 175, 105-112.



253

Huang, W.L., Li, Q.F., Lin, Y.Z., 2005. Calculation of photoneutrons produced in the targets of electron linear accelerators for radiography and radiotherapy applications. Nucl. Instrum. Meth. Phys. Res. B, 229, 339-347.

IAEA, 2010. Fighting cancer in developing countries. International Atomic Energy Agency Bulletin 52-1. Vienna.

IAEA, 2000. Handbook on photonuclear data for applications. Cross sections and spectra. International Atomic Energy Agency IAEA-TECDOC-1178. Vienna.

Kry, S.F., Howell, R.M., Salehpour, M., Followill, D.S., 2009. Neutron spectra and dose equivalents calculated in tissue for high-energy radiation therapy. Med. Phys., 36,1244-1250.

Lin, J.P., Chu, T.C., Lin, S.Y., Liu, M.T., 2001. The measurement of photoneutrons in the vicinity of a Siemens Primus linear accelerator. Appl. Rad. Isot., 55, 315-321.

Martinez-Ovalle, S.A., Barquero, R., Gómez-Ros, J.M., Lallena, A.M., 2011. Neutron dose equivalent and neutron spectra in tissue for clinical linacs operating at 15, 18 and 20 MV. Radiat. Prot. Dosim., doi:10.1093/rpd/ncq501.

McGinley, P.H., 1998. Shielding techniques for radiation oncology facilities. Medical Physics Publication Co. Madison WI.

Mesbahi, A., Keshtkar, A., Mohammadi, E., Mohammadzadeh, M., 2010. Effect of wedge filter and field size on photoneutron dose equivalent for an 18 MV photon beam of a medical linear accelerator. Appl. Radiat. Isot., 68, 84-89.

NCRP, 1984. Neutron contamination from medical accelerators. National Council on Radiation Protection and Measurements Report 79. Washington, DC, USA.

Patil, B.J., Chavan, S.T., Pethe, S.N., Krishnan, R., Bhoraskar, V.N., Dhole, S.D., 2011. Estimation of neutron production from accelerator head assembly of 15 MV medical LINAC using FLUKA simulations. Nucl. Instrum. Meth. Phys. Res. B, doi:10.1016/j.nimb.2011.04.013.

Pelowitz, D.B. (ed). (2005). MCNPX user manual, version 2.5.0. Los Alamos National Laboratory report LA-UR-02-2607. New Mexico.

Polaczek-Grelik, K., Orlef, A., Dybek, M., Konefa, A., Zipper, W., 2010. Linear accelerator therapeutic dose-induced radioactivity dependence. Appl. Radiat. Isot., 68, 763-766.

Sheikh-Bagheri, D., Rogers, D.W.O., 2002. Monte Carlo calculation of nine megavoltage photon beam spectra using the BEAM code. Med. Phys., 29, 391-402.

Tosi, G., Torresin, A., Agosteo, S., Flogio Para, A.F., Sangiust, V., Zeni, L., 1991. Neutron measurements around medical electron accelerators by active and passive detection techniques. Med. Phys., 18, 54-60.



254

Vega-Carrillo, H.R., Baltazar-Raigosa, A., 2011. Photoneutron spectra around an 18 MV linac. J. Radioanal. Nucl. Chem., 287, 323-327.

Xu, X.G., Bednarz, B., Paganetti, H., 2008. A review of dosimetry studies on external-beam radiation treatment with respect to secondary cancer induction. Phys. Med. Biol., 53, R193-R241.

photon and photoneutron spectra produced in radiotherapy ... · al., 2009, hernandez-adame et al.,...

Documents