- Email: [email protected]
Neutron production from flattening filter free high energy medical linac: A Monte Carlo study
Radiation Physics and Chemistry 116 (2015) 176–180
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Neutron production from flattening filter free high energy medical linac: A Monte Carlo study M.A. Najem a,n, F.A. Abolaban b, Z. Podolyák a, N.M. Spyrou a a b
Department of Physics, University of Surrey, Guildford, Surrey GU2 7XH, UK Nuclear Engineering Department, Faculty of Engineering, King Abdulaziz University, P.O. Box 80204, Jeddah 21589, Kingdom of Saudi Arabia
H I G H L I G H T S
Neutron fluence decreased inside the treatment room by the flattening filter removal. The reduction varies with energy due to the dissimilarity in flattening filter compositions. Neutron doses equivalent of the 15FFF SnS IMRT for prostate decreased by 71.3%. This reduction reduces the risk of inducing a secondary cancer from 1.2% to 0.3%.
art ic l e i nf o
a b s t r a c t
Article history: Received 7 July 2014 Received in revised form 27 January 2015 Accepted 30 January 2015 Available online 31 January 2015
One of the problems arising from using a conventional linac at high energy ( 48 MV) is the production of neutrons. One way to reduce neutron production is to remove the flattening filter (FF). The main purpose of this work was to study the effect of FF removal on neutron fluence and neutron dose equivalent inside the treatment room at different photon beam energies. Several simulations based on Monte Carlo techniques were carried out in order to calculate the neutron fluence at different locations in the treatment room from different linac energies with and without a FF. In addition, a step-and-shoot intensity modulated radiotherapy (SnS IMRT) for prostate cancer was modelled using the 15 MV photon beam with and without a FF on a water phantom to calculate the neutron dose received in a full treatment. The results obtained show a significant drop-off in neutrons fluence and dose equivalent when the FF was removed. For example, the neutron fluence was decreased by 54%, 76% and 75% for 10, 15 and 18 MV, respectively. This can decrease the neutron dose to the patient as well as reduce the shielding cost of the treatment room. The neutron dose equivalent of the SnS IMRT for prostate cancer was reduced significantly by 71.3% when the FF was removed. It can be concluded that the flattening filter removal from the head of the linac could reduce the risk of causing secondary cancers and the shielding cost of radiotherapy treatment rooms. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Flattening filter free Photoneutron production Monte Carlo technique Secondary induced cancer
1. Introduction Photon beams of linear accelerators (linacs) operating above 8 MV are contaminated with neutrons which cause more damage to tissues than photons and electrons. Flattening filters increase the head scatter that increases the photoneutron production from various head components (i.e. primary collimator, jaws and MLC) and can produce noticeable photoneutron contributions (in the case when high Z materials are used) (Georg et al., 2011). Several studies have investigated the effect of removing the flattening filter, flattening filter free mode (FFF), on photoneutron n
Corresponding author. E-mail address: [email protected] (M.A. Najem).
http://dx.doi.org/10.1016/j.radphyschem.2015.01.040 0969-806X/& 2015 Elsevier Ltd. All rights reserved.
production (Kry et al., 2008, 2007, Mesbahi, 2009, Mesbahi and Nejad, 2008). Kry et al. (2007) used gold activation foils inside polyethylene cylinders with cadmium cover to measure the neutron fluence and neutron dose equivalent for the unflattened 18 MV beam (18FFF) of a Varian 21EX linac in different positions and field sizes. In their measurement the dose per MU was 1 and 3.65 cGy MU 1 for FF and FFF beams, respectively. They found that the neutron dose equivalent reduced by (20.0 7 1.2)% per MU which is equivalent to a (76 75)% reduction per Gy at the dmax. The total reduction in neutron dose for IMRT treatment was about (69.0 74.1)%. Kry et al. (2008) used gold foils with a set of six Bonner spheres ranging from 5 to 30 cm to measure the neutron fluence and neutron dose equivalent for the 18FFF beam. Both MLC and Jaws
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
were closed. They found that the neutron strength was reduced by (81 75)% and the neutron dose equivalent at surface reduced from 2.31 70.14 (with a FF) to 0.45 70.03 mSv Gy 1 (without a FF). They also used the MCNPX MC code to calculate the ratio of neutrons produced from each linac head component for different field sizes. The contribution of FF on neutron production ranged from (12.80 70.13)% to (13.50 7 0.14)% for 0 0 to 10 10 cm2 field sizes, respectively. Mesbahi (2009) carried out a MC simulation to model the 18FFF beam of an Elekta linac and calculate the electron and neutron contaminations. He found that the electron contamination increased by 1.6 and 1.45 times for 10 10 and 30 30 cm2 field sizes, respectively, while the neutron level reduced by (61 70.6)% and (47 70.5)% for 10 10 and 30 30 cm2 field sizes, respectively, when the FF was taken out. The neutron reduction results in a decrease in the number of photons created in the target required to produce the same dose at the isocentre as well as a decrease in the head scatter. To our knowledge, no one has studied the effect of removing the FF on neutron production from Varian and Elekta linacs at photon beam energy other than 18 MV except a previous work to the authors which was carried out on the 15 MV photon beam (Najem et al., 2014). Therefore, more studies need to be carried out to gain a full understanding of neutron production from these linacs in case they are used at higher energies as manufacturers (e.g. Elekta) start to increase the nominal electron energy of FFF linacs in order to match the percentage depth dose of the flattened beam (Foster et al., 2014). The purpose of this work is to calculate the neutron fluence at different locations inside a treatment room housing a FFF linac using Monte Carlo technique. This was carried out, first, in order to estimate the total neutron fluence and neutron dose equivalent produced by these types of linacs and compare them with what were produced by conventional linac. Second, to get a clear idea about the effect of the nominal energy of the FFF beam on photoneutron production. In addition, a step-and-shoot (SnS) intensity modulated radiotherapy (IMRT) for a prostate cancer case was modelled on a water phantom using the 15 MV beam with and without a flattening filter to calculate the neutron reduction in a full treatment.
2. Materials and methods
Fig. 1. Cross-sectional view of a plan crossing the isocentre of the modelled concrete treatment room including locations where neutron calculations were performed.
Facilities Division, 2006). Fig. 1 shows a cross sectional view of the modelled treatment room at a plane crossing the isocentre. The number of histories was set to 6 109 in each setup (X and FFF modes) to reduce the statistical uncertainty to an acceptable level. 2.2. Calculated parameters Four parameters were calculated. The neutron dose equivalent (H0) was calculated within the first centimetre of the water phantom on the central axis. The neutron fluence in air, neutron strength (Qn) and neutron spectra were calculated at three different locations: the isocentre, inside the treatment vault at a distance of 1 m from the isocentre and inside the maze at a distance of 6.2 m from the isocentre (see Fig. 1). The Qn inside the room and in the maze can be calculated from the following Eqs. (1) and (2), respectively (NCRP, 2005; Kry et al., 2008):
Q n, room =
2.1. MC simulation A treatment room housing a Varian Clinac linac with different photon beam energies (10, 15 and 18 MV) with (X) and without (FFF) a FF was modelled using FLUKA (Ferrari et al., 2005). The actual geometry of the Varian Clinac head used in the simulation was obtained from the Varian manufacturer. The energy of incident electron beam that hits the target was set to 10, 15 and 18 MeV and its diameter was set to 3 mm. The “PHOTONUC” card was turned on in FLUKA to activate the photoneutron reaction. The electron and photon cut-off energies were set to 700 and 50 keV, respectively. The interaction length for the inelastic interaction of photon–nucleus was set to 0.02 using the “LAM-BIAS” card to enhance the efficiency of photoneutron production. A 40 40 40 cm3 water phantom was simulated at 100 cm source to surface distance. The jaws and MLC were fixed to form a 10 10 cm2 field-size at the phantom surface. The treatment vault design was modelled with the same dimension of the treatment room layout in the “Appendix 1” of Department of Health: Estates and Facilities Division (2006). In addition, a concrete with density 2.35 g cm 3 was used as the material of treatment room wall as stated in the same reference (Department of Health: Estates and
177
Q n, maze =
−1 φtot ⎛ 1 6.6 ⎞ ⎜⎜ ⎟⎟ + a ⎝ 4 πd12 S ⎠
2 π φtot a
−1 ⎛ 1 6.6 ⎞ ⎜⎜ ⎟⎟ + Sr ⎠ ⎝ 2 d22
(1)
(2)
Where; φtot is the total neutron fluence, a is the neutron transmission factor penetrating the head shielding, d1 is the distance between the target and the point of calculation, S is the treatment room surface area excluding the maze, d2 is the distance between the isocentre and the point of calculation and Sr is the treatment room surface area including the maze. 2.3. MC validation These simulations were validated by comparing the D10 (dose at depth 10 cm) and d80 (depth at relative dose 80%) of 10X, 15X, and 18X photon beams with those published in BJR supplement 25 (Jordan, 1996). Table 1 lists the D10 and d80 calculated using MC technique and published in BJR supplement 25 of the 10X, 15X and 18X beams for a 10 10 cm2 field size. As can be seen in Table 1, the variation in BJR supplement 25 and MC results were less than 2%.
178
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Table 1 D10 and d80 calculated using MC technique and published in BJR supplement 25 of the 10X, 15X and 18X photon beams for a 10 10 cm2 field size in a water phantom.
10X 15X 18X
D10 (%) [MC]
D10 (%) [BJR]
d80 (cm) [MC]
d80 (cm) [BJR]
72.6 7 1.5 77.6 7 1.6 79.5 7 1.6
73.0 77.0 79.0
7.9 70.1 9.2 70.1 9.8 70.1
8.0 9.1 9.7
2.4. IMRT plan In order to calculate the reduction in neutron dose in a full treatment by removing the FF, a SnS IMRT plan of an arbitrary prostate cancer case using 15X beam was created by the treatment planning system (TPS) at the Royal Surrey County Hospital (RSCH), Guildford, UK. This plan was produced using the Varian Eclipse TPS version 10.0.42 (Varian Medical Systems, Palo Alto, CA, USA). This IMRT plan was performed with five beams (posterior, left posterior, left interior, right interior and right posterior photon beams) to deliver 74 Gy in 34 fractions. The total MUs to deliver each 2 Gy was 471 MU. The same plan settings then were modelled with and without a FF using the MC code FLUKA. Since FLUKA cannot model a full treatment in one run, 66 simulations were carried out to model each segmentation (MLC position) of the 15X and 15FFF SnS IMRT plans, then combining all simulations together to obtain the final results. In addition, the SnS IMRT plan of prostate was modelled in a 40 40 30 cm3 water phantom instead of patient CT images for simplicity and to reduce the total time required to finish all simulations. The voxel size was set to 0.5 0.5 0.5 cm3. The electron and photon cut-off energies were set both to 5 MeV to reduce the time of each simulation because the lowest photoneutron production threshold of linac materials is for tungsten 183W at 6.19 MeV (Chadwick et al., 2000). The statistical uncertainty in neutron calculations was less than 5% for all voxels.
3. Results and discussion 3.1. Neutron dose equivalent The calculated neutron dose equivalents of flattened and unflattened beams at the surface of the phantom on the central axis for different energies are listed in Table 2. The calculated dose equivalent for conventional beams in this work agreed well with published data. For example, Zanini et al. (2004) found the neutron dose equivalent at 3 cm from the isocentre for a 10 10 cm2 field size to be 3.3 7 0.3 mSv Gy 1. The percentage reduction in neutron dose equivalent when the FF was removed is the same as in fluence. The decrease in neutron dose equivalent will have a significant impact on reducing the possibility of secondary induced cancers. 3.2. Neutron fluence in air The neutron fluence for 10, 15 and 18 MV flattened and unflattened beams at the isocentre and the other two locations are listed in Table 3. It can be seen from this table that the neutron
fluence at the isocentre for the 15X beam is larger than what was obtained in a previous work (Najem at al. 2014). This results from the increase in neutron scatter dose due to the treatment room as well as the water phantom. To ensure this hypothesis is correct, a MC simulation of the 15X beam was carried out in the same treatment room but without a water phantom. The result obtained showed that neutron fluence decreased to 1.1 107 n cm 2. It can be noticed that the neutron fluence decreased with all beams when the FF was removed. The reduction was 54.4%, 76% and 75% for 10, 15 and 18 MV beams, respectively. The maximum reduction was obtained for the 15 MV beam due to FF compositions as it is made totally from tungsten (the (γ,n) reaction threshold for 183W is 6.19 MeV). The FF was made from copper (for the 10 MV photon beam) and from steel and tantalum (for the 18 MV beam). The (γ, n) reaction thresholds for 63C u, 56Fe and 181Ta are 10.9, 11.2 and 7.6 MeV, respectively (Chadwick et al., 2000). 3.3. Neutron strength The Qn at these locations for all beams can be calculated using Eqs. (1) and (2). The neutron strengths for all beams at these locations are shown in Table 4. It can be seen that the calculated Qn for flattened beams at different locations are approximately equal, especially for 15 and 18 MV beams, as the uncertainty increased for the 10 MV beam due to the lower photoneutron production. It can be seen for unflattened beams that the Qn is approximately constant regardless of the location, but it has a slightly higher value at the isocentre than the other two locations as the energy increases compared with flattened beams. These results are consistent with the production mechanism of photoneutrons theory which indicates that direct neutrons, the dominant at the high energy region, are produced with the same direction of photons generated them (Zabihzadeh et al., 2009). Consequently, a larger amount of fast neutrons are emitted towards the isocentre because the FF removal reduces the photon scatter in the linac head and increases the number of photons that travel straight forward. However, Eqs. (1) and (2) can still be valid for unflattened beams as an approximation. The reduction in Qn for the 18 MV photon beam by removing the FF, which was found to be (78.4 71.4)% in this work, agrees with what was obtained by Kry et al. (2007) as they found the reduction in neutron strength by the FF removal with closed jaws was (81 7 5)%. 3.4. Neutron spectra The neutron spectra for 18X and 18FFF beams at these three locations are shown in Fig. 2. It can be seen that, despite the reduction in the number of neutrons produced, both beams have thermal peaks and a slight shift of the fast peaks towards the higher energy region. Moreover, the reduction in fast neutron peak by moving from the isocentre to other locations is higher than that in the thermal neutron region, which agrees with the literature and is due to the neutron scatter with concrete walls of the treatment room (Pena et al., 2005). With reference to the data listed above and the results obtained by Kry et al. (2009a) and Jank et al. (2014), the reduction in photon scatter and neutron fluence can reduce the amount of neutrons leakage from treatment room and therefore reduce the annual
Table 2 Neutron dose equivalent at different locations in the treatment room for 10, 15 and 18 MV photon beams with and without a FF. 10 MV H0 (mSv Gy
1
)
X 0.030 7 0.005
15 MV FFF 0.020 7 0.003
X 2.20 70.09
18 MV FFF 0.54 70.02
X 3.10 70.07
FFF 0.80 7 0.01
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Table 3 Neutron fluences ( 106 n cm
2
) at isocentre (Iso), location 1 (L1) and location 2 (L2) in the treatment room for 10, 15 and 18 MV photon beams with and without a FF.
10 MV
15 MV
X Iso L1 L2
179
FFF
0.317 0.01 0.167 0.01 (1.80 7 0.01) 10
2
0.147 0.01 (5.5 7 0.3) 10 (6.3 7 0.8) 10
2 3
18 MV
X
FFF
X
FFF
14.0 7 0.1 7.8 7 0.1 0.60 7 0.02
3.40 70.03 1.40 70.02 0.13 70.01
18.5 70.1 10.4 70.7 0.80 70.03
4.60 7 0.02 1.90 7 0.01 0.20 7 0.01
Table 4 Neutron strengths ( 1012) at different locations in the treatment room for 10, 15 and 18 MV photon beams with and without a FF. 10 MV
Iso L1 L2 Qaverage
15 MV
18 MV
X
FFF
X
FFF
X
FFF
0.034 7 0.001 0.030 7 0.002 0.0447 0.007 0.036 7 0.001
0.0147 0.001 0.0107 0.001 0.0157 0.002 0.0147 0.001
1.60 7 0.02 1.40 7 0.02 1.50 7 0.07 1.50 7 0.01
0.340 70.003 0.250 70.003 0.30 70.01 0.31 70.01
2.05 7 0.01 1.80 7 0.01 1.90 7 0.06 2.05 7 0.03
0.508 7 0.003 0.340 7 0.002 0.38 7 0.01 0.417 0.01
Fig. 2. The neutron spectra inside the treatment room for (a) the 18X beam at different locations (b) 18FFF beam at different locations (c) 18X and 18FFF at isocentre and (d) 18X and 18FFF at the other two locations.
dose to staff if this type of linac is housed in existing treatment room or can reduce the total amount of treatment room shielding required, which is found to be decreased by 8–20% (Kry et al., 2009a; Jank et al., 2014), thus reducing the installation cost. However, this conclusion needs further investigation by calculating the average workload for each beam of energy per week. The calculation is required due to first, the increase in dose rate by the FF removal, which could decrease the treatment time and therefore, increasing the number of patients treated per week. Second,
the increase in total MUs used per treatment course, when the FFF beam was used, which was reported by several studies (Cashmore, 2008; Zwahlen et al., 2012). 3.5. IMRT plan The calculated neutron dose equivalent for 15X and 15FFF MC SnS IMRT prostate plan carried out on a water phantom are shown in Fig. 3. As can be seen in this figure, neutrons deposit their
180
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Fig. 3. The neutron dose equivalent for SnS IMRT prostate plan modelled on a water phantom using (a) 15X and (b) 15FFF.
energy within the first few centimetres from the phantom or patient surface, which agrees with the literature (Kry et al., 2009b; Martinez-Ovalle et al., 2011). The maximum neutron dose equivalent for the 15X IMRT plan was found to be 9.1 70.3 m Sv MU 1. This is equivalent to 2.1 70.1 mSv Gy 1 as the total number of MUs in this plan was 471 MU to deliver 2 Gy to isocentre for each radiotherapy fractionation. This result agreed with what was obtained by Hussein et al. (2012) as they found the neutron dose equivalent of IMRT for prostate cancer using Polyallyl-diglycol-carbonate detector to be 2.0 7 0.3 mSv Gy 1. The neutron dose equivalent decreases sharply with the FF removal. The maximum dose equivalent for the 15FFF plan was found to be 2.4 70.1 mSv MU 1 (0.56 70.02 mSv Gy 1), which is equivalent to a 71.3% reduction per MU. Since most prostate cancer patients receive in general a total dose of about 75 Gy in 30–40 fractions, the total neutron dose equivalent of the 15FFF plan is 42 mSv rather than 158 mSv in the case of 15X plan. From these results, the lifetime risk of neutron to induce secondary cancers in the general population is found to be 1.2% and 0.3% for 15X and 15FFF IMRT plans, respectively. These values were calculated according to what was reported in the NCRP report 116 supposing the lifetime risk of radiation to induce secondary cancers per Sv in general population is 7.3% (NCRP 1993).
4. Conclusion The neutron level inside the treatment room decreased with the FF removal at all energies. The reduction amount varied with energy due to the dissimilarity in FF compositions. The decline in neutron dose equivalent and neutron using the FFF beam, which were up to 76%, has a significant impact on the outcome of radiotherapy treatment. In addition, the reduction in neutron fluence found in this work, and radiation scatter inside the treatment room found in other studies, could decrease the annual dose received by staff for existing treatment rooms or reduce the shielding cost when new treatment rooms are built. Neutron doses equivalent to the 15FFF SnS IMRT for prostate decrease by 71.3%. This reduction reduced the estimated risk of inducing a secondary cancer from 1.2% to 0.3%. It can be concluded that the FFF beam can enhance the outcomes of existing radiotherapy techniques by reducing the neutron dose to normal tissues.
Acknowledgement The authors would like to thank the Radiotherapy Physics Section at RSCH for providing us with the IMRT clinical plans.
References Cashmore, J., 2008. The characterization of unflattened photon beams from a 6 MV linear accelerator. Phys. Med. Biol. 53, 1933. Chadwick, M., Oblozinsky, P., Blokhin, A., Fukahori, T., Han, Y., Lee, Y., Martins, M., Mughabghab, S., Varlamov, V., Yu, B., 2000. Handbook of Photonuclear Data for Applications: Cross Sections and Spectra. IAEA, IAEA TECH-DOC 1178, Geneva, Switzerland. Department of Health: Estates and Facilities Division, 2006. Facilities for Cancer Services, Health building note no. 54, Stationery Office. Ferrari, A., Fasso, A., Ranft, J., Sala, P.R., 2005. FLUKA: A Multi-Particle Transport Code. CERN, Geneva, Switzerland. Foster, R.D., Speiser, M.P., Solberg, T.D., 2014. Commissioning and verification of the collapsed cone convolution superposition algorithm for SBRT delivery using flattening filter-free beams. J. Appl. Clin. Med. Phys. 15 (2), 39–49. Georg, D., Knöös, T., McClean, B., 2011. Current status and future perspective of flattening filter free photon beams. Med. Phys. 38, 1280–1293. Hussein, M., Aldridge, S., Urbano, T.G., Nisbet, A., 2012. The effect of 6 and 15 MV on intensity-modulated radiation therapy prostate cancer treatment: plan evaluation, tumour control probability and normal tissue complication probability analysis, and the theoretical risk of secondary induced malignancies. Br. J. Radiol. 85, 423–432. Jank, J., Kragl, G., Georg, D., 2014. Impact of a flattening filter free linear accelerator on structural shielding design. Z. Med. Phys. 24, 38–48. Jordan, T.J., 1996. Megavoltage x-ray beams: 2–50 MV. Br. J. Radiol. Suppl. 25, 62–109. Kry, S.F., Howell, R.M., Polf, J., Mohan, R., Vassiliev, O.N., 2009a. Treatment vault shielding for a flattening filter-free medical linear accelerator. Phys. Med. Biol. 54, 1265–1273. Kry, S.F., Howell, R.M., Salehpour, M., Followill, D.S., 2009b. Neutron spectra and dose equivalents calculated in tissue for high-energy radiation therapy. Med. Phys. 36, 1244–1250. Kry, S.F., Howell, R.M., Titt, U., Salehpour, M., Mohan, R., Vassiliev, O.N., 2008. Energy spectra, sources, and shielding considerations for neutrons generated by a flattening filter-free Clinac. Med. Phys. 35, 1906–1911. Kry, S.F., Titt, U., Pönisch, F., Vassiliev, O.N., Salehpour, M., Gillin, M., Mohan, R., 2007. Reduced neutron production through use of a flattening-filter-free accelerator. Int. J. Radiat. Oncol. Biol. Phys. 68, 1260–1264. Martinez-Ovalle, S., Barquero, R., Gómez-Ros, J., Lallena, A., 2011. Neutron dose equivalent and neutron spectra in tissue for clinical linacs operating at 15, 18 and 20 MV. Radiat. Prot. Dosim. 147, 498–511. Mesbahi, A., 2009. A Monte Carlo study on neutron and electron contamination of an unflattened 18 MV photon beam. Appl. Radiat. Isot. 67, 55–60. Mesbahi, A., Nejad, F.S., 2008. Monte Carlo study on a flattening filter-free 18 MV photon beam of a medical linear accelerator. Radiat. Med. 26, 331–336. Najem, M.A., Spyrou, N.M., Podolyák, Z., Abolaban, F.A., 2014. The physical characteristics of the 15 MV Varian Clinac 2100C unflattened beam. Radiat. Phys. Chem. 95, 205–209. NCRP, 1993. Limitation of Exposure to Ionizing Radiation. National Council on Radiation Protection and Measurements, Bethesda, Report 116. NCRP, 2005. Structural Shielding Design and Evaluation for Megavoltage X-ray and Gamma-ray Radiotherapy Facilities. National Council on Radiation Protection and Measurements, Bethesda, Report 151. Pena, J., Franco, L., Gomez, F., Iglesias, A., Pardo, J., Pombar, M., 2005. Monte Carlo study of Siemens PRIMUS photoneutron production. Phys. Med. Biol. 50, 5921–5933. Zabihzadeh, M., Ay, M.R., Allahverdi, M., Mesbahi, A., Mahdavi, S.R., Shahriari, M., 2009. Monte Carlo estimation of photoneutrons contamination from high-energy X-ray medical accelerators in treatment room and maze: a simplified model. Radiat. Prot. Dosim. 135 (1), 21–32. Zanini, A., Durisi, E., Fasolo, F., Ongaro, C., Visca, L., Nastasi, U., Burn, K., Scielzo, G., Adler, J., Annand, J., et al., 2004. Monte Carlo simulation of the photoneutron field in linac radiotherapy treatments with different collimation systems. Phys. Med. Biol. 49 (4), 571–582. Zwahlen, D.R., Lang, S., Hrbacek, J., Glanzmann, C., Kloeck, S., Najafi, Y., Streller, T., Studer, G., Zaugg, K., Luetolf, U.M., 2012. The use of photon beams of a flattening filter-free linear accelerator for hypofractionated volumetric modulated arc therapy in localized prostate cancer. Int. J. Radiat. Oncol. Biol. Phys. 83, 1655–1660.
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Neutron production from flattening filter free high energy medical linac: A Monte Carlo study M.A. Najem a,n, F.A. Abolaban b, Z. Podolyák a, N.M. Spyrou a a b
Department of Physics, University of Surrey, Guildford, Surrey GU2 7XH, UK Nuclear Engineering Department, Faculty of Engineering, King Abdulaziz University, P.O. Box 80204, Jeddah 21589, Kingdom of Saudi Arabia
H I G H L I G H T S
Neutron fluence decreased inside the treatment room by the flattening filter removal. The reduction varies with energy due to the dissimilarity in flattening filter compositions. Neutron doses equivalent of the 15FFF SnS IMRT for prostate decreased by 71.3%. This reduction reduces the risk of inducing a secondary cancer from 1.2% to 0.3%.
art ic l e i nf o
a b s t r a c t
Article history: Received 7 July 2014 Received in revised form 27 January 2015 Accepted 30 January 2015 Available online 31 January 2015
One of the problems arising from using a conventional linac at high energy ( 48 MV) is the production of neutrons. One way to reduce neutron production is to remove the flattening filter (FF). The main purpose of this work was to study the effect of FF removal on neutron fluence and neutron dose equivalent inside the treatment room at different photon beam energies. Several simulations based on Monte Carlo techniques were carried out in order to calculate the neutron fluence at different locations in the treatment room from different linac energies with and without a FF. In addition, a step-and-shoot intensity modulated radiotherapy (SnS IMRT) for prostate cancer was modelled using the 15 MV photon beam with and without a FF on a water phantom to calculate the neutron dose received in a full treatment. The results obtained show a significant drop-off in neutrons fluence and dose equivalent when the FF was removed. For example, the neutron fluence was decreased by 54%, 76% and 75% for 10, 15 and 18 MV, respectively. This can decrease the neutron dose to the patient as well as reduce the shielding cost of the treatment room. The neutron dose equivalent of the SnS IMRT for prostate cancer was reduced significantly by 71.3% when the FF was removed. It can be concluded that the flattening filter removal from the head of the linac could reduce the risk of causing secondary cancers and the shielding cost of radiotherapy treatment rooms. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Flattening filter free Photoneutron production Monte Carlo technique Secondary induced cancer
1. Introduction Photon beams of linear accelerators (linacs) operating above 8 MV are contaminated with neutrons which cause more damage to tissues than photons and electrons. Flattening filters increase the head scatter that increases the photoneutron production from various head components (i.e. primary collimator, jaws and MLC) and can produce noticeable photoneutron contributions (in the case when high Z materials are used) (Georg et al., 2011). Several studies have investigated the effect of removing the flattening filter, flattening filter free mode (FFF), on photoneutron n
Corresponding author. E-mail address: [email protected] (M.A. Najem).
http://dx.doi.org/10.1016/j.radphyschem.2015.01.040 0969-806X/& 2015 Elsevier Ltd. All rights reserved.
production (Kry et al., 2008, 2007, Mesbahi, 2009, Mesbahi and Nejad, 2008). Kry et al. (2007) used gold activation foils inside polyethylene cylinders with cadmium cover to measure the neutron fluence and neutron dose equivalent for the unflattened 18 MV beam (18FFF) of a Varian 21EX linac in different positions and field sizes. In their measurement the dose per MU was 1 and 3.65 cGy MU 1 for FF and FFF beams, respectively. They found that the neutron dose equivalent reduced by (20.0 7 1.2)% per MU which is equivalent to a (76 75)% reduction per Gy at the dmax. The total reduction in neutron dose for IMRT treatment was about (69.0 74.1)%. Kry et al. (2008) used gold foils with a set of six Bonner spheres ranging from 5 to 30 cm to measure the neutron fluence and neutron dose equivalent for the 18FFF beam. Both MLC and Jaws
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
were closed. They found that the neutron strength was reduced by (81 75)% and the neutron dose equivalent at surface reduced from 2.31 70.14 (with a FF) to 0.45 70.03 mSv Gy 1 (without a FF). They also used the MCNPX MC code to calculate the ratio of neutrons produced from each linac head component for different field sizes. The contribution of FF on neutron production ranged from (12.80 70.13)% to (13.50 7 0.14)% for 0 0 to 10 10 cm2 field sizes, respectively. Mesbahi (2009) carried out a MC simulation to model the 18FFF beam of an Elekta linac and calculate the electron and neutron contaminations. He found that the electron contamination increased by 1.6 and 1.45 times for 10 10 and 30 30 cm2 field sizes, respectively, while the neutron level reduced by (61 70.6)% and (47 70.5)% for 10 10 and 30 30 cm2 field sizes, respectively, when the FF was taken out. The neutron reduction results in a decrease in the number of photons created in the target required to produce the same dose at the isocentre as well as a decrease in the head scatter. To our knowledge, no one has studied the effect of removing the FF on neutron production from Varian and Elekta linacs at photon beam energy other than 18 MV except a previous work to the authors which was carried out on the 15 MV photon beam (Najem et al., 2014). Therefore, more studies need to be carried out to gain a full understanding of neutron production from these linacs in case they are used at higher energies as manufacturers (e.g. Elekta) start to increase the nominal electron energy of FFF linacs in order to match the percentage depth dose of the flattened beam (Foster et al., 2014). The purpose of this work is to calculate the neutron fluence at different locations inside a treatment room housing a FFF linac using Monte Carlo technique. This was carried out, first, in order to estimate the total neutron fluence and neutron dose equivalent produced by these types of linacs and compare them with what were produced by conventional linac. Second, to get a clear idea about the effect of the nominal energy of the FFF beam on photoneutron production. In addition, a step-and-shoot (SnS) intensity modulated radiotherapy (IMRT) for a prostate cancer case was modelled on a water phantom using the 15 MV beam with and without a flattening filter to calculate the neutron reduction in a full treatment.
2. Materials and methods
Fig. 1. Cross-sectional view of a plan crossing the isocentre of the modelled concrete treatment room including locations where neutron calculations were performed.
Facilities Division, 2006). Fig. 1 shows a cross sectional view of the modelled treatment room at a plane crossing the isocentre. The number of histories was set to 6 109 in each setup (X and FFF modes) to reduce the statistical uncertainty to an acceptable level. 2.2. Calculated parameters Four parameters were calculated. The neutron dose equivalent (H0) was calculated within the first centimetre of the water phantom on the central axis. The neutron fluence in air, neutron strength (Qn) and neutron spectra were calculated at three different locations: the isocentre, inside the treatment vault at a distance of 1 m from the isocentre and inside the maze at a distance of 6.2 m from the isocentre (see Fig. 1). The Qn inside the room and in the maze can be calculated from the following Eqs. (1) and (2), respectively (NCRP, 2005; Kry et al., 2008):
Q n, room =
2.1. MC simulation A treatment room housing a Varian Clinac linac with different photon beam energies (10, 15 and 18 MV) with (X) and without (FFF) a FF was modelled using FLUKA (Ferrari et al., 2005). The actual geometry of the Varian Clinac head used in the simulation was obtained from the Varian manufacturer. The energy of incident electron beam that hits the target was set to 10, 15 and 18 MeV and its diameter was set to 3 mm. The “PHOTONUC” card was turned on in FLUKA to activate the photoneutron reaction. The electron and photon cut-off energies were set to 700 and 50 keV, respectively. The interaction length for the inelastic interaction of photon–nucleus was set to 0.02 using the “LAM-BIAS” card to enhance the efficiency of photoneutron production. A 40 40 40 cm3 water phantom was simulated at 100 cm source to surface distance. The jaws and MLC were fixed to form a 10 10 cm2 field-size at the phantom surface. The treatment vault design was modelled with the same dimension of the treatment room layout in the “Appendix 1” of Department of Health: Estates and Facilities Division (2006). In addition, a concrete with density 2.35 g cm 3 was used as the material of treatment room wall as stated in the same reference (Department of Health: Estates and
177
Q n, maze =
−1 φtot ⎛ 1 6.6 ⎞ ⎜⎜ ⎟⎟ + a ⎝ 4 πd12 S ⎠
2 π φtot a
−1 ⎛ 1 6.6 ⎞ ⎜⎜ ⎟⎟ + Sr ⎠ ⎝ 2 d22
(1)
(2)
Where; φtot is the total neutron fluence, a is the neutron transmission factor penetrating the head shielding, d1 is the distance between the target and the point of calculation, S is the treatment room surface area excluding the maze, d2 is the distance between the isocentre and the point of calculation and Sr is the treatment room surface area including the maze. 2.3. MC validation These simulations were validated by comparing the D10 (dose at depth 10 cm) and d80 (depth at relative dose 80%) of 10X, 15X, and 18X photon beams with those published in BJR supplement 25 (Jordan, 1996). Table 1 lists the D10 and d80 calculated using MC technique and published in BJR supplement 25 of the 10X, 15X and 18X beams for a 10 10 cm2 field size. As can be seen in Table 1, the variation in BJR supplement 25 and MC results were less than 2%.
178
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Table 1 D10 and d80 calculated using MC technique and published in BJR supplement 25 of the 10X, 15X and 18X photon beams for a 10 10 cm2 field size in a water phantom.
10X 15X 18X
D10 (%) [MC]
D10 (%) [BJR]
d80 (cm) [MC]
d80 (cm) [BJR]
72.6 7 1.5 77.6 7 1.6 79.5 7 1.6
73.0 77.0 79.0
7.9 70.1 9.2 70.1 9.8 70.1
8.0 9.1 9.7
2.4. IMRT plan In order to calculate the reduction in neutron dose in a full treatment by removing the FF, a SnS IMRT plan of an arbitrary prostate cancer case using 15X beam was created by the treatment planning system (TPS) at the Royal Surrey County Hospital (RSCH), Guildford, UK. This plan was produced using the Varian Eclipse TPS version 10.0.42 (Varian Medical Systems, Palo Alto, CA, USA). This IMRT plan was performed with five beams (posterior, left posterior, left interior, right interior and right posterior photon beams) to deliver 74 Gy in 34 fractions. The total MUs to deliver each 2 Gy was 471 MU. The same plan settings then were modelled with and without a FF using the MC code FLUKA. Since FLUKA cannot model a full treatment in one run, 66 simulations were carried out to model each segmentation (MLC position) of the 15X and 15FFF SnS IMRT plans, then combining all simulations together to obtain the final results. In addition, the SnS IMRT plan of prostate was modelled in a 40 40 30 cm3 water phantom instead of patient CT images for simplicity and to reduce the total time required to finish all simulations. The voxel size was set to 0.5 0.5 0.5 cm3. The electron and photon cut-off energies were set both to 5 MeV to reduce the time of each simulation because the lowest photoneutron production threshold of linac materials is for tungsten 183W at 6.19 MeV (Chadwick et al., 2000). The statistical uncertainty in neutron calculations was less than 5% for all voxels.
3. Results and discussion 3.1. Neutron dose equivalent The calculated neutron dose equivalents of flattened and unflattened beams at the surface of the phantom on the central axis for different energies are listed in Table 2. The calculated dose equivalent for conventional beams in this work agreed well with published data. For example, Zanini et al. (2004) found the neutron dose equivalent at 3 cm from the isocentre for a 10 10 cm2 field size to be 3.3 7 0.3 mSv Gy 1. The percentage reduction in neutron dose equivalent when the FF was removed is the same as in fluence. The decrease in neutron dose equivalent will have a significant impact on reducing the possibility of secondary induced cancers. 3.2. Neutron fluence in air The neutron fluence for 10, 15 and 18 MV flattened and unflattened beams at the isocentre and the other two locations are listed in Table 3. It can be seen from this table that the neutron
fluence at the isocentre for the 15X beam is larger than what was obtained in a previous work (Najem at al. 2014). This results from the increase in neutron scatter dose due to the treatment room as well as the water phantom. To ensure this hypothesis is correct, a MC simulation of the 15X beam was carried out in the same treatment room but without a water phantom. The result obtained showed that neutron fluence decreased to 1.1 107 n cm 2. It can be noticed that the neutron fluence decreased with all beams when the FF was removed. The reduction was 54.4%, 76% and 75% for 10, 15 and 18 MV beams, respectively. The maximum reduction was obtained for the 15 MV beam due to FF compositions as it is made totally from tungsten (the (γ,n) reaction threshold for 183W is 6.19 MeV). The FF was made from copper (for the 10 MV photon beam) and from steel and tantalum (for the 18 MV beam). The (γ, n) reaction thresholds for 63C u, 56Fe and 181Ta are 10.9, 11.2 and 7.6 MeV, respectively (Chadwick et al., 2000). 3.3. Neutron strength The Qn at these locations for all beams can be calculated using Eqs. (1) and (2). The neutron strengths for all beams at these locations are shown in Table 4. It can be seen that the calculated Qn for flattened beams at different locations are approximately equal, especially for 15 and 18 MV beams, as the uncertainty increased for the 10 MV beam due to the lower photoneutron production. It can be seen for unflattened beams that the Qn is approximately constant regardless of the location, but it has a slightly higher value at the isocentre than the other two locations as the energy increases compared with flattened beams. These results are consistent with the production mechanism of photoneutrons theory which indicates that direct neutrons, the dominant at the high energy region, are produced with the same direction of photons generated them (Zabihzadeh et al., 2009). Consequently, a larger amount of fast neutrons are emitted towards the isocentre because the FF removal reduces the photon scatter in the linac head and increases the number of photons that travel straight forward. However, Eqs. (1) and (2) can still be valid for unflattened beams as an approximation. The reduction in Qn for the 18 MV photon beam by removing the FF, which was found to be (78.4 71.4)% in this work, agrees with what was obtained by Kry et al. (2007) as they found the reduction in neutron strength by the FF removal with closed jaws was (81 7 5)%. 3.4. Neutron spectra The neutron spectra for 18X and 18FFF beams at these three locations are shown in Fig. 2. It can be seen that, despite the reduction in the number of neutrons produced, both beams have thermal peaks and a slight shift of the fast peaks towards the higher energy region. Moreover, the reduction in fast neutron peak by moving from the isocentre to other locations is higher than that in the thermal neutron region, which agrees with the literature and is due to the neutron scatter with concrete walls of the treatment room (Pena et al., 2005). With reference to the data listed above and the results obtained by Kry et al. (2009a) and Jank et al. (2014), the reduction in photon scatter and neutron fluence can reduce the amount of neutrons leakage from treatment room and therefore reduce the annual
Table 2 Neutron dose equivalent at different locations in the treatment room for 10, 15 and 18 MV photon beams with and without a FF. 10 MV H0 (mSv Gy
1
)
X 0.030 7 0.005
15 MV FFF 0.020 7 0.003
X 2.20 70.09
18 MV FFF 0.54 70.02
X 3.10 70.07
FFF 0.80 7 0.01
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Table 3 Neutron fluences ( 106 n cm
2
) at isocentre (Iso), location 1 (L1) and location 2 (L2) in the treatment room for 10, 15 and 18 MV photon beams with and without a FF.
10 MV
15 MV
X Iso L1 L2
179
FFF
0.317 0.01 0.167 0.01 (1.80 7 0.01) 10
2
0.147 0.01 (5.5 7 0.3) 10 (6.3 7 0.8) 10
2 3
18 MV
X
FFF
X
FFF
14.0 7 0.1 7.8 7 0.1 0.60 7 0.02
3.40 70.03 1.40 70.02 0.13 70.01
18.5 70.1 10.4 70.7 0.80 70.03
4.60 7 0.02 1.90 7 0.01 0.20 7 0.01
Table 4 Neutron strengths ( 1012) at different locations in the treatment room for 10, 15 and 18 MV photon beams with and without a FF. 10 MV
Iso L1 L2 Qaverage
15 MV
18 MV
X
FFF
X
FFF
X
FFF
0.034 7 0.001 0.030 7 0.002 0.0447 0.007 0.036 7 0.001
0.0147 0.001 0.0107 0.001 0.0157 0.002 0.0147 0.001
1.60 7 0.02 1.40 7 0.02 1.50 7 0.07 1.50 7 0.01
0.340 70.003 0.250 70.003 0.30 70.01 0.31 70.01
2.05 7 0.01 1.80 7 0.01 1.90 7 0.06 2.05 7 0.03
0.508 7 0.003 0.340 7 0.002 0.38 7 0.01 0.417 0.01
Fig. 2. The neutron spectra inside the treatment room for (a) the 18X beam at different locations (b) 18FFF beam at different locations (c) 18X and 18FFF at isocentre and (d) 18X and 18FFF at the other two locations.
dose to staff if this type of linac is housed in existing treatment room or can reduce the total amount of treatment room shielding required, which is found to be decreased by 8–20% (Kry et al., 2009a; Jank et al., 2014), thus reducing the installation cost. However, this conclusion needs further investigation by calculating the average workload for each beam of energy per week. The calculation is required due to first, the increase in dose rate by the FF removal, which could decrease the treatment time and therefore, increasing the number of patients treated per week. Second,
the increase in total MUs used per treatment course, when the FFF beam was used, which was reported by several studies (Cashmore, 2008; Zwahlen et al., 2012). 3.5. IMRT plan The calculated neutron dose equivalent for 15X and 15FFF MC SnS IMRT prostate plan carried out on a water phantom are shown in Fig. 3. As can be seen in this figure, neutrons deposit their
180
M.A. Najem et al. / Radiation Physics and Chemistry 116 (2015) 176–180
Fig. 3. The neutron dose equivalent for SnS IMRT prostate plan modelled on a water phantom using (a) 15X and (b) 15FFF.
energy within the first few centimetres from the phantom or patient surface, which agrees with the literature (Kry et al., 2009b; Martinez-Ovalle et al., 2011). The maximum neutron dose equivalent for the 15X IMRT plan was found to be 9.1 70.3 m Sv MU 1. This is equivalent to 2.1 70.1 mSv Gy 1 as the total number of MUs in this plan was 471 MU to deliver 2 Gy to isocentre for each radiotherapy fractionation. This result agreed with what was obtained by Hussein et al. (2012) as they found the neutron dose equivalent of IMRT for prostate cancer using Polyallyl-diglycol-carbonate detector to be 2.0 7 0.3 mSv Gy 1. The neutron dose equivalent decreases sharply with the FF removal. The maximum dose equivalent for the 15FFF plan was found to be 2.4 70.1 mSv MU 1 (0.56 70.02 mSv Gy 1), which is equivalent to a 71.3% reduction per MU. Since most prostate cancer patients receive in general a total dose of about 75 Gy in 30–40 fractions, the total neutron dose equivalent of the 15FFF plan is 42 mSv rather than 158 mSv in the case of 15X plan. From these results, the lifetime risk of neutron to induce secondary cancers in the general population is found to be 1.2% and 0.3% for 15X and 15FFF IMRT plans, respectively. These values were calculated according to what was reported in the NCRP report 116 supposing the lifetime risk of radiation to induce secondary cancers per Sv in general population is 7.3% (NCRP 1993).
4. Conclusion The neutron level inside the treatment room decreased with the FF removal at all energies. The reduction amount varied with energy due to the dissimilarity in FF compositions. The decline in neutron dose equivalent and neutron using the FFF beam, which were up to 76%, has a significant impact on the outcome of radiotherapy treatment. In addition, the reduction in neutron fluence found in this work, and radiation scatter inside the treatment room found in other studies, could decrease the annual dose received by staff for existing treatment rooms or reduce the shielding cost when new treatment rooms are built. Neutron doses equivalent to the 15FFF SnS IMRT for prostate decrease by 71.3%. This reduction reduced the estimated risk of inducing a secondary cancer from 1.2% to 0.3%. It can be concluded that the FFF beam can enhance the outcomes of existing radiotherapy techniques by reducing the neutron dose to normal tissues.
Acknowledgement The authors would like to thank the Radiotherapy Physics Section at RSCH for providing us with the IMRT clinical plans.
References Cashmore, J., 2008. The characterization of unflattened photon beams from a 6 MV linear accelerator. Phys. Med. Biol. 53, 1933. Chadwick, M., Oblozinsky, P., Blokhin, A., Fukahori, T., Han, Y., Lee, Y., Martins, M., Mughabghab, S., Varlamov, V., Yu, B., 2000. Handbook of Photonuclear Data for Applications: Cross Sections and Spectra. IAEA, IAEA TECH-DOC 1178, Geneva, Switzerland. Department of Health: Estates and Facilities Division, 2006. Facilities for Cancer Services, Health building note no. 54, Stationery Office. Ferrari, A., Fasso, A., Ranft, J., Sala, P.R., 2005. FLUKA: A Multi-Particle Transport Code. CERN, Geneva, Switzerland. Foster, R.D., Speiser, M.P., Solberg, T.D., 2014. Commissioning and verification of the collapsed cone convolution superposition algorithm for SBRT delivery using flattening filter-free beams. J. Appl. Clin. Med. Phys. 15 (2), 39–49. Georg, D., Knöös, T., McClean, B., 2011. Current status and future perspective of flattening filter free photon beams. Med. Phys. 38, 1280–1293. Hussein, M., Aldridge, S., Urbano, T.G., Nisbet, A., 2012. The effect of 6 and 15 MV on intensity-modulated radiation therapy prostate cancer treatment: plan evaluation, tumour control probability and normal tissue complication probability analysis, and the theoretical risk of secondary induced malignancies. Br. J. Radiol. 85, 423–432. Jank, J., Kragl, G., Georg, D., 2014. Impact of a flattening filter free linear accelerator on structural shielding design. Z. Med. Phys. 24, 38–48. Jordan, T.J., 1996. Megavoltage x-ray beams: 2–50 MV. Br. J. Radiol. Suppl. 25, 62–109. Kry, S.F., Howell, R.M., Polf, J., Mohan, R., Vassiliev, O.N., 2009a. Treatment vault shielding for a flattening filter-free medical linear accelerator. Phys. Med. Biol. 54, 1265–1273. Kry, S.F., Howell, R.M., Salehpour, M., Followill, D.S., 2009b. Neutron spectra and dose equivalents calculated in tissue for high-energy radiation therapy. Med. Phys. 36, 1244–1250. Kry, S.F., Howell, R.M., Titt, U., Salehpour, M., Mohan, R., Vassiliev, O.N., 2008. Energy spectra, sources, and shielding considerations for neutrons generated by a flattening filter-free Clinac. Med. Phys. 35, 1906–1911. Kry, S.F., Titt, U., Pönisch, F., Vassiliev, O.N., Salehpour, M., Gillin, M., Mohan, R., 2007. Reduced neutron production through use of a flattening-filter-free accelerator. Int. J. Radiat. Oncol. Biol. Phys. 68, 1260–1264. Martinez-Ovalle, S., Barquero, R., Gómez-Ros, J., Lallena, A., 2011. Neutron dose equivalent and neutron spectra in tissue for clinical linacs operating at 15, 18 and 20 MV. Radiat. Prot. Dosim. 147, 498–511. Mesbahi, A., 2009. A Monte Carlo study on neutron and electron contamination of an unflattened 18 MV photon beam. Appl. Radiat. Isot. 67, 55–60. Mesbahi, A., Nejad, F.S., 2008. Monte Carlo study on a flattening filter-free 18 MV photon beam of a medical linear accelerator. Radiat. Med. 26, 331–336. Najem, M.A., Spyrou, N.M., Podolyák, Z., Abolaban, F.A., 2014. The physical characteristics of the 15 MV Varian Clinac 2100C unflattened beam. Radiat. Phys. Chem. 95, 205–209. NCRP, 1993. Limitation of Exposure to Ionizing Radiation. National Council on Radiation Protection and Measurements, Bethesda, Report 116. NCRP, 2005. Structural Shielding Design and Evaluation for Megavoltage X-ray and Gamma-ray Radiotherapy Facilities. National Council on Radiation Protection and Measurements, Bethesda, Report 151. Pena, J., Franco, L., Gomez, F., Iglesias, A., Pardo, J., Pombar, M., 2005. Monte Carlo study of Siemens PRIMUS photoneutron production. Phys. Med. Biol. 50, 5921–5933. Zabihzadeh, M., Ay, M.R., Allahverdi, M., Mesbahi, A., Mahdavi, S.R., Shahriari, M., 2009. Monte Carlo estimation of photoneutrons contamination from high-energy X-ray medical accelerators in treatment room and maze: a simplified model. Radiat. Prot. Dosim. 135 (1), 21–32. Zanini, A., Durisi, E., Fasolo, F., Ongaro, C., Visca, L., Nastasi, U., Burn, K., Scielzo, G., Adler, J., Annand, J., et al., 2004. Monte Carlo simulation of the photoneutron field in linac radiotherapy treatments with different collimation systems. Phys. Med. Biol. 49 (4), 571–582. Zwahlen, D.R., Lang, S., Hrbacek, J., Glanzmann, C., Kloeck, S., Najafi, Y., Streller, T., Studer, G., Zaugg, K., Luetolf, U.M., 2012. The use of photon beams of a flattening filter-free linear accelerator for hypofractionated volumetric modulated arc therapy in localized prostate cancer. Int. J. Radiat. Oncol. Biol. Phys. 83, 1655–1660.