Radiation Protection Dosimetry Advance Access published November 11, 2013 Radiation Protection Dosimetry (2013), pp. 1–4
doi:10.1093/rpd/nct284
AN INDIRECT IN VIVO DOSIMETRY SYSTEM FOR OCULAR PROTON THERAPY A. Carnicer1, V. Letellier2, G. Rucka3, G. Angellier1, W. Sauerwein4 and J. He´rault1, * 1 Cyclotron Biome´dical, Centre Antoine Lacassagne, 227 Avenue de la Lanterne, 06200 Nice, France 2 Centre de protonthe´rapie, Institut Curie, Campus Universitaire d’Orsay, baˆtiment 101, 91898 Orsay Cedex, France 3 Centre de radiothe´rapie St Louis, Hoˆpital de la Croix Rouge, rue Andre´ Blondel, 83100 Toulon France 4 Universita¨tsklinikum Essen, Strahlenklink, 45122 Essen, Germany
Secondary radiation, particularly neutron radiation, is a cause of concern in proton therapy. However, one can take advantage of its presence by using it to retrieve useful information on the primary proton beam. At the Centre Antoine Lacassagne the secondary radiation in the treatment room has been studied in function of the beam modulation. A strong correlation was found between the secondary ambient dose equivalent per proton dose H*(10)/D and proton dose rate D/MU. A large volume ionisation chamber fixed on the wall at 2.5 m from the nozzle was used with an in-house computer interface to retrieve the value of D/MU derived from the measurement of photon H*(10) integrated over treatment time, using the correlation curve. This system enables the verification of D and D/MU to be made independently of the monitoring of the primary beam and represents a first step towards an alternative in vivo dosimetry in proton therapy.
INTRODUCTION Proton therapy facilities are polluted by secondary radiation. The role of neutrons in secondary cancer induction is a matter of concern and many data in the literature report on the levels of the neutron dose equivalent or ambient dose equivalent per proton dose H/D or H*(10)/D across the treatment room(1 – 9). Variations in the energy beam and beam delivery type lead to a wide range of reported values. But even for a similar type of facility and energy beam significant variations can be found depending on the configuration of the nozzle. Indeed, secondary particles are produced by proton interaction with the elements encountered all along the beam axis, especially those with a high-Z material. Therefore, each facility is characterised by a particular secondary radiation field. Usually, brass collimators that limit the section of the beam are the main source of production of secondaries. In passive scattering delivery systems another important source are the elements inserted into the beam to spread it laterally and in depth. The scattering foils made of high-Z materials that are used to enlarge and homogenise the beam represent a large contribution. The range shifter (RS) and the range modulating wheel (RMW), employed for beam depth arrangement, also produce a significant contribution even when they are composed of a low-Z material. To limit the supply of secondaries the scattering system at the Centre Antoine Lacassagne (CAL) in Nice is placed outside the treatment room and the RS and RMW are entirely made of Plexiglas. An interesting point is that for a particular treatment, the individual elements used to conform the beam to
the patient’s tumour size and shape will contribute to the secondary radiation field around the nozzle. Therefore, secondary radiation could in theory provide useful information on the patient-specific beam components, if the relationship between both is well known. This subject has been indirectly addressed in some few works(3, 7, 9), but no practical information for a possible quality control system is provided. An ongoing line of research in proton and ion beam therapy is the development of a system for in vivo dosimetry, in particular for range and path verification. The only method currently used in routine clinics is PET imaging, but this technique is limited because annihilation photons do not completely correlate with the proton range(10, 11) and because imaging must be performed quickly after treatment, which ideally require the treatment room to be equipped with a PET camera. More promising is the use of prompt gamma rays for in vivo dosimetry(12 – 16), but development of a detection system incurs in high technical difficulties which have not yet been overcome. In a recent work(17) the secondary neutral radiation field at CAL was characterised. Fluence and dose maps were determined across the treatment room to identify the origin and directionality of secondaries, together with spectral fluences at positions of interest to determine strategy in monitoring. The relationship of secondary radiation with RMW and RS was also investigated and results were used to set up a system which represents an alternative method for indirect in vivo dosimetry. The system has been used since for quality control. The present study reports the
# The Author 2013. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
*Corresponding author: [email protected]
A. CARNICER ET AL.
implementation of the system and presents updated data on the results obtained. MATERIALS AND METHODS
The MCNPX code(20) was used, starting from the geometry developed in previous works(19, 21). In this work the geometry has been completed to include the whole treatment room, the beam line bunker and the maze (Figure 1). The following tallies were scored: † †
The proton dose deposition at the water phantom. The uncertainty of the maximum dose was 2 %. The photon and neutron H*(10) at the position of the CET62 and Studsvik detectors, respectively. Uncertainties were ,4 %.
The RS and RMW were modelled as a unique Plexiglas sheet. Nineteen simulations were run using a common phase space file ( psf ), varying the thickness of the Plexiglas sheet in steps of 0.8 mm, which corresponds to the RMW steps. The proton dose D at a depth z was then obtained for any SOBP from an appropriate weight of single simulations: Pn p D ðzÞ Pn i i ð1Þ DðzÞ ¼ i¼0 i¼0 pi Equation (1) was also applied to H*(10). More details are given in the original paper(17).
Dosimetric measurements Neutron and photon ambient dose equivalent H*(10), proton dose D and proton dose rate D/MU were measured during treatment. Photon and neutron H*(10) were measured, respectively, with the previously mentioned large-volume ionisation chamber CET62 (SAPHYMO, France) and with a portable monitor Studsvik model 2202D connected to a scaling counter ECS1 (Nardeux, France). The mean standard deviation for 100 patients (four sessions per patient) was 1 % for photons and 1.5 % for neutrons (type A evaluation). Their energy window covers the range from 10 keV to 10 MeV and from thermal to 17 MeV, respectively. The Studsvik detector was placed on the table next to the wall of the entrance maze, right below the CET62 ionisation chamber. Calibration was performed in terms of H*(10) in the energies of 137Cs for the photon ionisation chamber and 252Cf for the neutron detector. The proton dose D was measured in the water tank at the centre of the spread-out Bragg Peak (SOBP), using the ionisation chamber Far West Technology IC18 and following TRS-398 report(18). MU is the charge collected in the air-filled ionisation chambers located in box 2 and is proportional to the dose. At CAL the quantity D/MU is always determined from simulations, with accuracy better than 3 %(19). The value of D/MU calculated for a given SOBP is normalised to that of the SOBP used for daily calibration. It is characterised by a proton range and modulation width in water of 32 and 16 mm, respectively, and a value of D/MU of 1.37 cGy MU21.
Modulation Measurements and simulations were performed for a large range of clinical SOBPs, which cover the whole proton dose rate D/MU and range employed at CAL (0.4 –1.5 cGy MU21 and 5 to 32 mm, respectively). In particular, integrated photon and neutron H*(10) were simultaneously acquired during treatment, together with the proton dose D and the proton dose rate D/MU. Data from up to 66 and 177 patients were acquired for photons and neutrons, respectively. The same quantities were determined also with MC simulation for 100 different SOBPs. RESULTS A strong correlation was found between H*(10)/D and D/MU both for photons and neutrons and from both measurements and simulations (Figure 2). All four data sets were fit to a power function with a correlation coefficient better than 0.99 except for the experimental neutron data set (0.95) due to higher reading fluctuation. H*(10)/D decreases up to a factor of 5 for increasing values of D/MU. Calculations and measurements differ by a constant factor all along the range of D/MU values, of 2.3 for neutrons and of 0.5 for photons. The modulation of the beam is satisfactorily achieved by simulations since calculated curves show the same behaviour than experimental ones. CET62 ionisation chamber can be operated from the control room and provides higher
Page 2 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
The proton beam line at CAL delivers low-energy protons of 65 MeV and is used to treat eye tumours. A single Tantalum scattering foil is located in the beam line bunker, which is separated from the treatment room by a barite-enriched concrete wall. The nozzle holds two boxes in the treatment room. The first box contains the RS and the RMW. The second box houses two air-filled parallel-plate-type transmission ionisation chambers and two collimators. At the end of the second box the patient-specific accessories, such as the collimator and possible filters or compensators, are found. A water tank is placed on the position of the patient for dosimetric measurements. The treatment room is equipped with a large-volume ionisation chamber fixed on the wall next to the maze, 2.5 m above the floor and from the nozzle, at the level of box 1 (Figure 1). This chamber is connected to a computer in the treatment room, which enables the measurements of integrated H*(10) to be made from there during treatment.
Monte Carlo simulation
AN IN VIVO DOSIMETRY SYSTEM FOR PROTON THERAPY
Figure 1. Whole geometry of the simulation visualised with Moritz, showing the treatment room, the beam line bunker and the maze.
Figure 4. Differences between the calculated and the theoretical D/MU as a function of D/MU for 1660 patients.
Figure 2. H*(10)/D as a function of D/MU for photons (at CET62) and neutrons (at Studsvik), both for measurements and simulations.
correlation than Studsvik even if less data were acquired. For this reason the chamber was used to set up a system to verify the proton dose D and proton dose rate D/MU from the integrated photon H*(10) acquired during treatment. To this end a user interface was developed to store the value of H*(10) and retrieve the corresponding D/MU value from the curve. The relative difference between the theoretical value of D/MU entered by the user and this latter value is also provided. If a high difference is obtained then either the delivered dose is wrong or the RMW and/or the RS are not correct. This system has been used at CAL since its implementation and routinely adopted as an additional tool for quality control. Up to now .1660 measurements have been performed.
The relative differences between theoretical D/MU and those obtained from H*(10) show a Gaussian distribution centred on 0 with a standard deviation of 3 % (Figure 3). These values are in addition independent of the value of D/MU as shown in Figure 4. The advantage of this tool is that it provides a verification of D/MU and D independent of the beam monitoring. The second ionisation chamber in box 2 is electronically independent of the first for control purposes, but a failure in the selection of RMWor RS would not be detected as both chambers would provide the same reading. No differences were found for different patients treated with similar dose and dose rates. The system is thus not able to detect changes in secondary particle production coming from the patient-specific accessories such as the personal final collimator, filters or compensators. The ionisation chamber is indeed too far from the patient location and this contribution is in addition very small when compared with that of the RMW and RS. Therefore, a complete verification system is not yet achieved. Detecting this component
Page 3 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
Figure 3. Histogram of the differences between the calculated and the theoretical D/MU for 1660 patients.
A. CARNICER ET AL.
would require a smaller detector size and a closer position. Even more difficult is the detection of nuclear events coming from the patient itself and is the final step towards a true in vivo dosimetry. CONCLUSION
9.
10.
11.
12.
13.
FUNDING This work was supported by l’Agence Nationale de la Recherche (ANR), France [PROUESSE Contract No. ANR-09-COSI-010-01].
14.
REFERENCES 1. Agosteo, S., Birattari, C., Caravaggio, M., Silari, M. and Tosi, G. Secondary neutron and photon dose in proton therapy. Radiother. Oncol. 48, 293 –305 (1998). 2. Tayama, R., Fujita, Y., Tadokoro, M., Fujimaki, H., Sakae, T. and Terunuma, T. Measurement of neutron dose distribution for a passive scattering nozzle at the Proton Medical Research Center (PMRC). Nucl. Instrum. Methods Phys. Res. A 564, 532–536 (2006). 3. Zheng, Y., Fontenot, J., Taddei, P., Mirkovic, D. and Newhauser, W. Monte Carlo simulations of neutron spectral fluence, radiation weighting factor and ambient dose equivalent for a passively scattered proton therapy unit. Phys. Med. Biol. 53, 187– 201 (2008). 4. Trompier, F., Delacroix, S., Vabre, I., Joussard, F. and Proust, J. Secondary exposure for 73 and 200 MeV proton therapy. Radiat. Prot. Dosim. 125(1–4), 349–354 (2007). 5. Yonai, S. et al. Measurement of neutron ambient dose equivalent in passive carbon-ion and proton radiotherapies. Med. Phys. 35, 4782– 4792 (2008). 6. Clasie, B., Wroe, A., Kooy, H., Depauw, N., Flanz, J., Paganetti, H. and Rosenfeld, A. Assessment of out-offield absorbed dose and equivalent dose in proton fields. Med. Phys. 37(1), 311 –321 (2010). 7. Wang, X., Sahoo, N., Zhu, R. X., Zullo, J. R. and Gillin, M. T. Measurement of neutron dose equivalent
15.
16. 17.
18.
19.
20. 21.
Page 4 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
Neutron and photon H*(10)/D are strongly correlated with D/MU by a power function. This correlation was used as a system for proton dose and proton dose rate verification. Up to now the system is able to predict possible errors in the modulation of the beam uniquely due to incorrect choice of RMW and/or RS. These errors would not be detected by the double ionisation chamber, conventionally used as a verification system. The method is easy to set up, requires simple and low-cost instrumentation, and can be implemented both for photon and neutron detection. This is a first step towards an in vivo dosimetry and further work is in progress to develop a full verification system.
8.
and its dependence on beam configuration for a passive scattering proton delivery system. Int. J. Radiat. Oncol. Biol. Phys. 76(5), 1563– 1570 (2010). Pe´rez-Andujar, A., Newhauser, W. D. and DeLuca, P. M. Neutron production from beam-modifying devices in a modern double scattering proton therapy beam delivery system. Phys. Med. Biol. 54, 993– 1008 (2009). Polf, J. C. and Newhauser, W. D. Calculations of neutron dose equivalent exposures from range-modulated proton therapy beams. Phys. Med. Biol. 50, 3859–3873 (2005). Fourkal, E., Fan, J. and Veltchev, I. Absolute dose reconstruction in proton therapy using PET imaging modality: feasibility study. Phys. Med. Biol. 54, N217– N228 (2009). Vecchio, S., Attanasi, F., Belcari, N., Camarda, M., Cirrone, G. A. P. and Cuttone, G. A PET prototype for in-beam monitoring of proton therapy. In: IEEE Nucl. Sci. Symp. Conf. Rec. N24-362, Honolulu, HI, pp. 1607–1611 (2007). Min, C. H., Kim, C. H., Youn, M. Y. and Kim, J. W. Prompt gamma measurements for locating the dose falloff region in the proton therapy. Appl. Phys. Lett. 89, 183517 (2006). Min, C. H., Kim, J. W., Youn, M. Y. and Kim, C. H. Determination of distal dose edge location by measuring right-angled prompt gamma-rays from a 38 MeV proton beam. Nucl. Instrum. Methods Phys. Res. A 580, 562– 565 (2007). Min, C. H., Lee, H. R., Kim, C. H. and Lee, S. B. Development of array-type prompt gamma measurement system for in vivo range verification in proton therapy. Med. Phys. 39(4), 2100–2107 (2012). Roellinghoff, F. et al. Design of a Compton camera for 3D prompt-gamma imaging during ion beam therapy. Nucl. Instrum. Methods Phys. Res. A 648(Suppl. 1), S20– S23 (2011). Smeets, J. et al. Prompt gamma imaging with a slit camera for real-time range control in proton therapy. Phys. Med. Biol. 57(11), 3371–3405 (2012). Carnicer, A., Letellier, V., Rucka, G., Angellier, G., Sauerwein, W. and He´rault, J. Study of the secondary neutral radiation in proton therapy: toward an indirect in vivo dosimetry. Med. Phys. 39(12), 7303–7316 (2012). International Atomic Energy Agency. Absorbed dose determination in external beam radiotherapy: An international code of practice for dosimetry based on standards of absorbed dose to water. Technical Report Series No. 398. IAEA (2000). He´rault, J., Iborra, N., Serrano, B. and Chauvel, P. Spread-out Bragg peak and monitor units calculation with the Monte Carlo Code MCNPX. Med. Phys. 34, 680– 688 (2007). Pelowitz, D. B. MCNPX user’s manual version 2.5.0. Report No. LA-CP-05-0369 . Los Alamos National Laboratory (2005). He´rault, J., Iborra, N., Serrano, B. and Chauvel, P. Monte Carlo simulation of a proton therapy platform devoted to ocular melanoma. Med. Phys. 32, 910–919 (2005).
doi:10.1093/rpd/nct284
AN INDIRECT IN VIVO DOSIMETRY SYSTEM FOR OCULAR PROTON THERAPY A. Carnicer1, V. Letellier2, G. Rucka3, G. Angellier1, W. Sauerwein4 and J. He´rault1, * 1 Cyclotron Biome´dical, Centre Antoine Lacassagne, 227 Avenue de la Lanterne, 06200 Nice, France 2 Centre de protonthe´rapie, Institut Curie, Campus Universitaire d’Orsay, baˆtiment 101, 91898 Orsay Cedex, France 3 Centre de radiothe´rapie St Louis, Hoˆpital de la Croix Rouge, rue Andre´ Blondel, 83100 Toulon France 4 Universita¨tsklinikum Essen, Strahlenklink, 45122 Essen, Germany
Secondary radiation, particularly neutron radiation, is a cause of concern in proton therapy. However, one can take advantage of its presence by using it to retrieve useful information on the primary proton beam. At the Centre Antoine Lacassagne the secondary radiation in the treatment room has been studied in function of the beam modulation. A strong correlation was found between the secondary ambient dose equivalent per proton dose H*(10)/D and proton dose rate D/MU. A large volume ionisation chamber fixed on the wall at 2.5 m from the nozzle was used with an in-house computer interface to retrieve the value of D/MU derived from the measurement of photon H*(10) integrated over treatment time, using the correlation curve. This system enables the verification of D and D/MU to be made independently of the monitoring of the primary beam and represents a first step towards an alternative in vivo dosimetry in proton therapy.
INTRODUCTION Proton therapy facilities are polluted by secondary radiation. The role of neutrons in secondary cancer induction is a matter of concern and many data in the literature report on the levels of the neutron dose equivalent or ambient dose equivalent per proton dose H/D or H*(10)/D across the treatment room(1 – 9). Variations in the energy beam and beam delivery type lead to a wide range of reported values. But even for a similar type of facility and energy beam significant variations can be found depending on the configuration of the nozzle. Indeed, secondary particles are produced by proton interaction with the elements encountered all along the beam axis, especially those with a high-Z material. Therefore, each facility is characterised by a particular secondary radiation field. Usually, brass collimators that limit the section of the beam are the main source of production of secondaries. In passive scattering delivery systems another important source are the elements inserted into the beam to spread it laterally and in depth. The scattering foils made of high-Z materials that are used to enlarge and homogenise the beam represent a large contribution. The range shifter (RS) and the range modulating wheel (RMW), employed for beam depth arrangement, also produce a significant contribution even when they are composed of a low-Z material. To limit the supply of secondaries the scattering system at the Centre Antoine Lacassagne (CAL) in Nice is placed outside the treatment room and the RS and RMW are entirely made of Plexiglas. An interesting point is that for a particular treatment, the individual elements used to conform the beam to
the patient’s tumour size and shape will contribute to the secondary radiation field around the nozzle. Therefore, secondary radiation could in theory provide useful information on the patient-specific beam components, if the relationship between both is well known. This subject has been indirectly addressed in some few works(3, 7, 9), but no practical information for a possible quality control system is provided. An ongoing line of research in proton and ion beam therapy is the development of a system for in vivo dosimetry, in particular for range and path verification. The only method currently used in routine clinics is PET imaging, but this technique is limited because annihilation photons do not completely correlate with the proton range(10, 11) and because imaging must be performed quickly after treatment, which ideally require the treatment room to be equipped with a PET camera. More promising is the use of prompt gamma rays for in vivo dosimetry(12 – 16), but development of a detection system incurs in high technical difficulties which have not yet been overcome. In a recent work(17) the secondary neutral radiation field at CAL was characterised. Fluence and dose maps were determined across the treatment room to identify the origin and directionality of secondaries, together with spectral fluences at positions of interest to determine strategy in monitoring. The relationship of secondary radiation with RMW and RS was also investigated and results were used to set up a system which represents an alternative method for indirect in vivo dosimetry. The system has been used since for quality control. The present study reports the
# The Author 2013. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
*Corresponding author: [email protected]
A. CARNICER ET AL.
implementation of the system and presents updated data on the results obtained. MATERIALS AND METHODS
The MCNPX code(20) was used, starting from the geometry developed in previous works(19, 21). In this work the geometry has been completed to include the whole treatment room, the beam line bunker and the maze (Figure 1). The following tallies were scored: † †
The proton dose deposition at the water phantom. The uncertainty of the maximum dose was 2 %. The photon and neutron H*(10) at the position of the CET62 and Studsvik detectors, respectively. Uncertainties were ,4 %.
The RS and RMW were modelled as a unique Plexiglas sheet. Nineteen simulations were run using a common phase space file ( psf ), varying the thickness of the Plexiglas sheet in steps of 0.8 mm, which corresponds to the RMW steps. The proton dose D at a depth z was then obtained for any SOBP from an appropriate weight of single simulations: Pn p D ðzÞ Pn i i ð1Þ DðzÞ ¼ i¼0 i¼0 pi Equation (1) was also applied to H*(10). More details are given in the original paper(17).
Dosimetric measurements Neutron and photon ambient dose equivalent H*(10), proton dose D and proton dose rate D/MU were measured during treatment. Photon and neutron H*(10) were measured, respectively, with the previously mentioned large-volume ionisation chamber CET62 (SAPHYMO, France) and with a portable monitor Studsvik model 2202D connected to a scaling counter ECS1 (Nardeux, France). The mean standard deviation for 100 patients (four sessions per patient) was 1 % for photons and 1.5 % for neutrons (type A evaluation). Their energy window covers the range from 10 keV to 10 MeV and from thermal to 17 MeV, respectively. The Studsvik detector was placed on the table next to the wall of the entrance maze, right below the CET62 ionisation chamber. Calibration was performed in terms of H*(10) in the energies of 137Cs for the photon ionisation chamber and 252Cf for the neutron detector. The proton dose D was measured in the water tank at the centre of the spread-out Bragg Peak (SOBP), using the ionisation chamber Far West Technology IC18 and following TRS-398 report(18). MU is the charge collected in the air-filled ionisation chambers located in box 2 and is proportional to the dose. At CAL the quantity D/MU is always determined from simulations, with accuracy better than 3 %(19). The value of D/MU calculated for a given SOBP is normalised to that of the SOBP used for daily calibration. It is characterised by a proton range and modulation width in water of 32 and 16 mm, respectively, and a value of D/MU of 1.37 cGy MU21.
Modulation Measurements and simulations were performed for a large range of clinical SOBPs, which cover the whole proton dose rate D/MU and range employed at CAL (0.4 –1.5 cGy MU21 and 5 to 32 mm, respectively). In particular, integrated photon and neutron H*(10) were simultaneously acquired during treatment, together with the proton dose D and the proton dose rate D/MU. Data from up to 66 and 177 patients were acquired for photons and neutrons, respectively. The same quantities were determined also with MC simulation for 100 different SOBPs. RESULTS A strong correlation was found between H*(10)/D and D/MU both for photons and neutrons and from both measurements and simulations (Figure 2). All four data sets were fit to a power function with a correlation coefficient better than 0.99 except for the experimental neutron data set (0.95) due to higher reading fluctuation. H*(10)/D decreases up to a factor of 5 for increasing values of D/MU. Calculations and measurements differ by a constant factor all along the range of D/MU values, of 2.3 for neutrons and of 0.5 for photons. The modulation of the beam is satisfactorily achieved by simulations since calculated curves show the same behaviour than experimental ones. CET62 ionisation chamber can be operated from the control room and provides higher
Page 2 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
The proton beam line at CAL delivers low-energy protons of 65 MeV and is used to treat eye tumours. A single Tantalum scattering foil is located in the beam line bunker, which is separated from the treatment room by a barite-enriched concrete wall. The nozzle holds two boxes in the treatment room. The first box contains the RS and the RMW. The second box houses two air-filled parallel-plate-type transmission ionisation chambers and two collimators. At the end of the second box the patient-specific accessories, such as the collimator and possible filters or compensators, are found. A water tank is placed on the position of the patient for dosimetric measurements. The treatment room is equipped with a large-volume ionisation chamber fixed on the wall next to the maze, 2.5 m above the floor and from the nozzle, at the level of box 1 (Figure 1). This chamber is connected to a computer in the treatment room, which enables the measurements of integrated H*(10) to be made from there during treatment.
Monte Carlo simulation
AN IN VIVO DOSIMETRY SYSTEM FOR PROTON THERAPY
Figure 1. Whole geometry of the simulation visualised with Moritz, showing the treatment room, the beam line bunker and the maze.
Figure 4. Differences between the calculated and the theoretical D/MU as a function of D/MU for 1660 patients.
Figure 2. H*(10)/D as a function of D/MU for photons (at CET62) and neutrons (at Studsvik), both for measurements and simulations.
correlation than Studsvik even if less data were acquired. For this reason the chamber was used to set up a system to verify the proton dose D and proton dose rate D/MU from the integrated photon H*(10) acquired during treatment. To this end a user interface was developed to store the value of H*(10) and retrieve the corresponding D/MU value from the curve. The relative difference between the theoretical value of D/MU entered by the user and this latter value is also provided. If a high difference is obtained then either the delivered dose is wrong or the RMW and/or the RS are not correct. This system has been used at CAL since its implementation and routinely adopted as an additional tool for quality control. Up to now .1660 measurements have been performed.
The relative differences between theoretical D/MU and those obtained from H*(10) show a Gaussian distribution centred on 0 with a standard deviation of 3 % (Figure 3). These values are in addition independent of the value of D/MU as shown in Figure 4. The advantage of this tool is that it provides a verification of D/MU and D independent of the beam monitoring. The second ionisation chamber in box 2 is electronically independent of the first for control purposes, but a failure in the selection of RMWor RS would not be detected as both chambers would provide the same reading. No differences were found for different patients treated with similar dose and dose rates. The system is thus not able to detect changes in secondary particle production coming from the patient-specific accessories such as the personal final collimator, filters or compensators. The ionisation chamber is indeed too far from the patient location and this contribution is in addition very small when compared with that of the RMW and RS. Therefore, a complete verification system is not yet achieved. Detecting this component
Page 3 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
Figure 3. Histogram of the differences between the calculated and the theoretical D/MU for 1660 patients.
A. CARNICER ET AL.
would require a smaller detector size and a closer position. Even more difficult is the detection of nuclear events coming from the patient itself and is the final step towards a true in vivo dosimetry. CONCLUSION
9.
10.
11.
12.
13.
FUNDING This work was supported by l’Agence Nationale de la Recherche (ANR), France [PROUESSE Contract No. ANR-09-COSI-010-01].
14.
REFERENCES 1. Agosteo, S., Birattari, C., Caravaggio, M., Silari, M. and Tosi, G. Secondary neutron and photon dose in proton therapy. Radiother. Oncol. 48, 293 –305 (1998). 2. Tayama, R., Fujita, Y., Tadokoro, M., Fujimaki, H., Sakae, T. and Terunuma, T. Measurement of neutron dose distribution for a passive scattering nozzle at the Proton Medical Research Center (PMRC). Nucl. Instrum. Methods Phys. Res. A 564, 532–536 (2006). 3. Zheng, Y., Fontenot, J., Taddei, P., Mirkovic, D. and Newhauser, W. Monte Carlo simulations of neutron spectral fluence, radiation weighting factor and ambient dose equivalent for a passively scattered proton therapy unit. Phys. Med. Biol. 53, 187– 201 (2008). 4. Trompier, F., Delacroix, S., Vabre, I., Joussard, F. and Proust, J. Secondary exposure for 73 and 200 MeV proton therapy. Radiat. Prot. Dosim. 125(1–4), 349–354 (2007). 5. Yonai, S. et al. Measurement of neutron ambient dose equivalent in passive carbon-ion and proton radiotherapies. Med. Phys. 35, 4782– 4792 (2008). 6. Clasie, B., Wroe, A., Kooy, H., Depauw, N., Flanz, J., Paganetti, H. and Rosenfeld, A. Assessment of out-offield absorbed dose and equivalent dose in proton fields. Med. Phys. 37(1), 311 –321 (2010). 7. Wang, X., Sahoo, N., Zhu, R. X., Zullo, J. R. and Gillin, M. T. Measurement of neutron dose equivalent
15.
16. 17.
18.
19.
20. 21.
Page 4 of 4
Downloaded from http://rpd.oxfordjournals.org/ at Univ of Southern California on April 7, 2014
Neutron and photon H*(10)/D are strongly correlated with D/MU by a power function. This correlation was used as a system for proton dose and proton dose rate verification. Up to now the system is able to predict possible errors in the modulation of the beam uniquely due to incorrect choice of RMW and/or RS. These errors would not be detected by the double ionisation chamber, conventionally used as a verification system. The method is easy to set up, requires simple and low-cost instrumentation, and can be implemented both for photon and neutron detection. This is a first step towards an in vivo dosimetry and further work is in progress to develop a full verification system.
8.
and its dependence on beam configuration for a passive scattering proton delivery system. Int. J. Radiat. Oncol. Biol. Phys. 76(5), 1563– 1570 (2010). Pe´rez-Andujar, A., Newhauser, W. D. and DeLuca, P. M. Neutron production from beam-modifying devices in a modern double scattering proton therapy beam delivery system. Phys. Med. Biol. 54, 993– 1008 (2009). Polf, J. C. and Newhauser, W. D. Calculations of neutron dose equivalent exposures from range-modulated proton therapy beams. Phys. Med. Biol. 50, 3859–3873 (2005). Fourkal, E., Fan, J. and Veltchev, I. Absolute dose reconstruction in proton therapy using PET imaging modality: feasibility study. Phys. Med. Biol. 54, N217– N228 (2009). Vecchio, S., Attanasi, F., Belcari, N., Camarda, M., Cirrone, G. A. P. and Cuttone, G. A PET prototype for in-beam monitoring of proton therapy. In: IEEE Nucl. Sci. Symp. Conf. Rec. N24-362, Honolulu, HI, pp. 1607–1611 (2007). Min, C. H., Kim, C. H., Youn, M. Y. and Kim, J. W. Prompt gamma measurements for locating the dose falloff region in the proton therapy. Appl. Phys. Lett. 89, 183517 (2006). Min, C. H., Kim, J. W., Youn, M. Y. and Kim, C. H. Determination of distal dose edge location by measuring right-angled prompt gamma-rays from a 38 MeV proton beam. Nucl. Instrum. Methods Phys. Res. A 580, 562– 565 (2007). Min, C. H., Lee, H. R., Kim, C. H. and Lee, S. B. Development of array-type prompt gamma measurement system for in vivo range verification in proton therapy. Med. Phys. 39(4), 2100–2107 (2012). Roellinghoff, F. et al. Design of a Compton camera for 3D prompt-gamma imaging during ion beam therapy. Nucl. Instrum. Methods Phys. Res. A 648(Suppl. 1), S20– S23 (2011). Smeets, J. et al. Prompt gamma imaging with a slit camera for real-time range control in proton therapy. Phys. Med. Biol. 57(11), 3371–3405 (2012). Carnicer, A., Letellier, V., Rucka, G., Angellier, G., Sauerwein, W. and He´rault, J. Study of the secondary neutral radiation in proton therapy: toward an indirect in vivo dosimetry. Med. Phys. 39(12), 7303–7316 (2012). International Atomic Energy Agency. Absorbed dose determination in external beam radiotherapy: An international code of practice for dosimetry based on standards of absorbed dose to water. Technical Report Series No. 398. IAEA (2000). He´rault, J., Iborra, N., Serrano, B. and Chauvel, P. Spread-out Bragg peak and monitor units calculation with the Monte Carlo Code MCNPX. Med. Phys. 34, 680– 688 (2007). Pelowitz, D. B. MCNPX user’s manual version 2.5.0. Report No. LA-CP-05-0369 . Los Alamos National Laboratory (2005). He´rault, J., Iborra, N., Serrano, B. and Chauvel, P. Monte Carlo simulation of a proton therapy platform devoted to ocular melanoma. Med. Phys. 32, 910–919 (2005).
