THIS REPORT considers potential hazards from radiation-induced activation of patients and materials in a treatment room secondary to radiotherapy. Because of major differences in the nature of the activation with the equipment used for the various types of radiotherapy beams, separate sections address photon, proton, and neutron facilities. While the potential hazards tend to be small or very small, this report does not include hazards to service personnel working in the treatment unit where activation doses, even for photon units, could be very high. Some of the contents of this report were presented by the American Association of Physicists in Medicine (AAPM) in a white paper submitted to the U.S. Nuclear Regulatory Commission (NRC). The white paper was prepared by an ad hoc panel. That white paper was reviewed and approved by the AAPM Radiation Safety Subcommittee, the Therapy Physics Committee, and subsequently by the AAPM Science Council. These committees felt that an expansion of that white paper would be beneficial as an educational document for the membership. The final report of this task group has also been approved by the Radiation Safety Subcommittee, the Therapy Physics Committee, and by the Science Council of the AAPM.
The Therapy Physics Committee charged the task group:
* to review information available both in the literature and from non-published sources addressing induced radioactivity produced by radiotherapy accelerators using photon, neutron, or proton beams;
* to assess the hazards to therapy staff, physics staff, engineering staff, patients, and patients’ family members presented by this induced radioactivity; and
* to make recommendations for minimizing such hazards following the ALARA principle and for the operational health physics aspects of activation.
A. POTENTIAL HAZARDS DUE TO ACTIVATION IN PHOTON RADIATION THERAPY FACILITIES
External-beam radiation therapy (EBRT) using high-energy photons (x-rays) produced by radiotherapy accelerators is the most common radiation modality for treating cancer. High-energy photons offer several advantages over lower-energy photons in terms of better dose distribution for deep-seated tumors, lower skin dose, less sensitivity to tissue heterogeneities, etc. However, for energies above 6 MeV and mostly for beams produced above 10 MV, some of the materials in the accelerator, the treatment room, and the radiotherapy patient become radioactive due primarily to photonuclear reactions and neutron-capture reactions, exposing therapy staff and patients to unwanted radiation dose. Most, but not all, of the radionuclides produced are short-lived. Isotopes that are neutron-deficient will likely undergo decay by positron emission or electron capture, and isotopes with excess neutrons will likely decay by β− emission. Since many of these decays result in isotopes in an excited state, gamma rays are emitted that can cause exposure to those present in the treatment room.
High-energy photons produce radioactive products primarily by (γ,n) reactions by releasing a neutron from a target nuclide. Other less common processes are photo-disintegration—(γ,2n), (γ,p), etc.—in the treatment head, the patient, and the surrounding materials in the room. The released neutrons are slowed down by the surrounding materials in the treatment room, allowing them to undergo (n,γ) reactions that result in activation of materials located away from the path of the original high-energy photon beams. These processes induce radioactivity in the irradiated materials, a process known as “photon activation.” Materials that may become activated include accelerator components, the patient couch, treatment accessories, the building’s materials, the patient, and air.
Recent advances in computer-controlled multi-leaf collimators (MLC), inverse planning software tools, and imaging sciences have led to the development of intensity-modulated radiation therapy (IMRT), which delivers dose distributions that are highly conformal to the target. IMRT gives the ability to deliver higher tumor doses concomitant with lower doses to normal tissue, and it has become the standard of care at many treatment sites. However, IMRT requires a substantial increase in monitor units for the same patient dose. This is primarily due to the low monitor-unit efficiency of IMRT compared to conventional EBRT. Similarly, with the introduction of the high-throughput IMRT techniques, such as volumetric modulated arc therapy modalities [e.g., VMAT™ (Elekta AB, Stockholm, Sweden) and RAPIDARC™ (Varian Medical Systems, Palo Alto, CA USA)], the number of patients to be treated per day on an individual machine may increase substantially. Some techniques also include much higher monitor-unit rates than conventional treatments (Welsh et al. 2007). Thus, the occupational and patient exposures due to activation could be higher; however, most IMRT uses beam energies below 10 MV. As some patients who would have been treated with high-energy beams are moved to IMRT, the exposure due to activation is likely to decrease. In addition, with the introduction of image-guided radiation therapy (IGRT) and stereotactic body radiation therapy (SBRT), many treatments are now designed with a higher dose per fraction and a smaller number of fractions, sometimes even a single fraction. This also leads to a potentially longer beam-on time per day for the same number of patient treatments and potentially an increase in activation in the treatment room. Many SBRT fields do use higher-energy beams that could lead to unwanted exposures through activation. While on a given day the number of monitor units may increase with SBRT, due to the lower total dose that accompanies the few fractions used, the net effect may be to reduce the activation exposure. This document summarizes various relevant aspects of photon activation in medical linear accelerators using high-energy x-rays.
ACTIVATION IN THE THERAPY ROOM AFTER IRRADIATIONS
The sources of activation in the treatment room can be classified into five components: (a) linear accelerator, (b) patient support system, (c) building materials, (d) patient, and (e) air. A number of investigators have reported on the characteristics and the levels of radiation in the treatment room. The type and significance of radioactive species vary depending upon the accelerator, the location in the room in relation to the angle of the accelerator head, the collimator opening, and the irradiation history and room construction. Howell et al. (2009) note that the neutron fluence varies by about a factor of three between the various manufacturers, but the ratio depends on the nominal energy. Because the threshold for isotope production covers a range, the actual activation products depend on the beam energy and spectrum. In addition, the patient also can undergo some activation (Wielopolski et al. 1987). The principal radionuclides produced by the photon activation process decay primarily through beta (positive and negative) emissions followed by gamma rays, with half-lives from a few minutes to several days, as listed in Table 1.
In 1977, Stranden reported that, in the somewhat extreme case of a 45-MV photon beam from a betatron, the radioactive species that are produced due to activation of the patient and in-room air primarily include 15O, 14O, 11C, 13N, and 30P (Stranden 1977).
In 1988, Ahlgren and Olsson, using gamma spectroscopy, detected the following radionuclides after typical use of an 18-MV beam: 24Na, 28Al, 54Mn, 56Mn, 57Ni, 53Fe, 59Fe, 58Co, 62Cu, 64Cu, 82Br, 122Sb, and 187W (produced in the accelerator head); 28Al, 53Fe, and 56Mn (produced in steel wedges); 27Mg, 28Al, 120Sb, 122Sb, and 124Sb (produced in lead wedges); 15O, 34mCl, and 38Cl (produced in PVC block-holding trays); and 53Fe, 56Mn, and 28Al (produced in concrete walls) (Ahlgren and Olsson 1988). Ateia et al. (2008a) confirmed the presence of most of these radionuclides. They measured the dose rates at the surface of the accelerator head and the wedges to project and estimate that the dose to a therapist’s fingers may be about 2 mGy y−1. They also identified 511 keV photons, assumedly from 14 N in the air and 62Cu in the accelerator head.
In 1991, Almen et al. reported that the principal nuclides produced during linear accelerator operation between 13 and 17 MV are 28Al, 56Mn, 64Cu, and 187W, of which the origins of 28Al, 64Cu, and 187 W are believed to be from aluminum components on the treatment couch, copper in the target and flattening filter, and tungsten in the target, respectively (Almen et al. 1991). The exposure of the technicians was primarily the result of leakage and scatter radiation penetrating the walls of the treatment room. For one accelerator, the absorbed dose rate in the treatment room was measured continuously between 0.5 min and 48 h after end of treatment. Immediately after treatment with high-energy photons, the radiation was dominated by 28Al and 62Cu (T1/2 = 2.3 and 9.7 min, respectively) and later by radionuclides with longer half-lives, 187W and 57Ni (T1/2 = 24 and 36 h, respectively). Due to these radionuclides, the radioactivity in the accelerator will build up, and the therapists will therefore be irradiated every time they enter the treatment room and not only directly after a treatment with high-energy photons. Similarly, Perrin et al. (2003) identified 28Al, 62Cu, 56Mn, and 187W as the dominant activation products at the beam exit window of a linear accelerator (LINAC) after a 2,000-MU, 18-MV x-ray irradiation. In addition, Rawlinson et al. (2002) identified 24Na and 122Sb 1.0 m lateral to the isocenter of a LINAC operating at 18 MV that decay with half-lives of 15 h and 2.7 d, respectively. However, they failed to observe at the studied location the copper and tungsten isotopes produced mainly in the target material and collimators.
In 2006, Fischer et al. (2006) identified activation products by in situ gamma spectroscopy at the isocenter of a medical linear accelerator shortly after termination of a high-energy photon beam irradiation with 15 × 15 cm field. Spectra were recorded either with an open or closed collimator. While some activation products disappear from the spectrum with closed collimator or exhibit reduced count rates, others remain with identical intensity. The former isotopes were neutron-deficient and mostly decay by positron emission or electron capture; the latter had neutron excess and decay by β− emission. This new finding was consistent with the assumption that photons in the primary beam produce activation products by (γ,n) reactions in the treatment head, and subsequently the neutrons created in these processes undergo (n,γ) reactions, creating activation products in a much larger area. These finding are also in accordance with Kalef-Ezra (2011). After 25-MV x-ray irradiations, Guo and Ziemer identified 52V, 49Cr, 51Cr, 56Mn, 53Fe, 57Ni, and 60Co in wedges made of iron, chromium, nickel, magnesium, and silicon, and 61Co, 57Ni, 51Cr, and 62Cu in wedges made of tungsten, nickel, and copper (Guo and Ziemer 2004). They also found 187W and 111mIn, 111Sn, 51Cr, 105Cd, and 203Pb in blocks made of Lipowitz’s metal (see Table 2 for the composition).
In 2007, Konefal et al. presented spectroscopy data for two different medical accelerators, a Siemens Primus (Siemens Healthcare, Erlangen, Germany) and a Varian Clinac-2300 (Konefal et al. 2008). The radioisotopes 187W, 56Mn, 28Al, 57Ni, 38Cl, 57Co, and 198Au were identified as a consequence of these reactions. Moreover, an increased photon fluence rate behind the door of the accelerator bunker in the operator console room was observed during emission of the 20-MV x-rays from the Varian Clinac-2300 as well as in the case of the 15-MV x-ray beam from the Siemens Primus. No increased radiation was observed with the 6-MV x-ray beam. The measurements indicated the presence of neutrons in the operator console room during emission of the 15-MV x-ray beam from the Primus Siemens as well as the 20-MV x-rays and the 22-MeV electrons from the Varian Clinac-2300 accelerator.
Also in 2007, Bednarz et al. (2007) used Monte Carlo simulations to plot neutron production locations inside the accelerator head components. Aided by data on the composition of the components, this neutron fluence map can help identify specific locations of photon and neutron activation and give one insight into ways to reduce such activation in the future.
In 2008, Fischer et al. obtained sequences of in situ gamma spectra, accompanied by continuous dose rate measurements, at the isocenters of four different brands of high-energy medical linear accelerators shortly after beam-off in order to study the effects of photon activation (Fischer et al. 2008). Spectral analysis revealed up to 20 different radionuclides per machine, with a total of 21 found isotopes having half-lives between 2.3 min and 5.3 y. Important isotopes as judged from activity, dose rate, and half-life were 28Al, 54Mn, 56Mn, 57Ni, 60Co, 62Cu, 64Cu, 124Sb, 187W, and 196Au. Petrović et al. (2011) validated most of this list and added 24Na.
DOSE TO PERSONNEL
In 1977, Stranden reported a whole body dose of 4 μGy to a patient due to activation of the patient’s body when a treatment of 2.5 Gy was delivered with 45-MV x-rays (Stranden 1977). These dose would have been lower with the use of lower-energy treatment beams. The corresponding dose to therapists was estimated to be much lower or negligible. These data, while interesting, have little application with respect to modern radiotherapy techniques.
In 1982, Hoffman and Nath reported that the magnitude of the dose due to activation of the patient using 25-MV x-rays was very small to negligible (Hoffman and Nath 1982). The authors reported exposure rates of 3 μSv h−1 at 1 m from the treatment head immediately after a treatment with 25-MV x-rays. As the patient moved out of the room, the exposure rate did not decrease, indicating that most of the radiation was coming from induced activity in the treatment head. The decay curve of exposure was observed to be consistent with a mixture of radioisotopes with half-lives of 2.8 and 12.8 min. The authors surmise that these half-lives may arise from a mixture of the following isotopes: 15O, 53Fe, 13N, 185mW, and 62Cu. Taking an average exposure rate of 1 μSv h−1 at 1 m, the exposure to the technologist (assumed to be at 1 m from the source) for the duration of one patient treatment (assuming an average time of 7.5 min per patient) is 0.13 μSv h−1. Assuming 32 patients per day, five working days per week, and 50 working weeks per year, the yearly exposure was estimated to be 1 mSv. Assuming a uniform rotation among three machines, a technologist working at the 25-MV accelerator was estimated to receive an average exposure of 0.33 mSv y−1, primarily from induced radioactivity in the treatment head. This is approximately the same as was observed on film badges.
In 1984, McGinley et al. (1984) reported that the annual skin dose to therapists due to positron emission was 0.15 mSv from air activation in the treatment room for 25-MV beams. Production rates of radioactive oxygen and nitrogen by high energy x-ray beams from medical accelerators were measured as functions of the accelerator’s energy. It was found that for typical operating conditions of medical accelerators, the dose received by personnel entering the treatment room is negligible compared to the maximum permissible dose limit. The measurements were for ventilation with eight air exchanges per hour and a relatively low delivery rate of 200 MU min−1.
In 1985, O’Brien et al. reported the magnitude of the dose due to air and patient activation was very small to negligible, validating Hoffman and Nath (O’Brien et al., 1985). Measurements were performed on a Therac‐25 linear accelerator (AECL, Quebec, Canada) with a 25-MV photon beam to estimate daily therapist exposure due to induced activity in the treatment room. The authors reported additional exposure to the operators from handling activated shielding blocks and wedges. An approximate 3 μSv h−1 rate was found 30 cm from a stainless steel block after a 500‐MU treatment with 24-MV x-rays. Twenty 4‐Gy fractions given within one hour induced a dose rate of 700 μSv h−1 close to the crosshairs about 1 min after the last exposure and 90 μSv h−1 at the side of the head.
In 1985, LaRiviere reported that 76% of the in-room dose to personnel is due to activation in the accelerator head (LaRiviere 1985). Whole-body doses incurred by therapists working with 18-MV and 24-MV medical linear accelerators were measured at two clinics under normal working conditions with a Geiger–Mueller-based personal dosimeter. The results based on 74 treatments (46 treatments with 24-MV x-ray beams with a mean exposure of 0.017 μSv and 20 treatments with an 18-MV x-ray beam with a mean exposure of 0.027 μSv, compared with eight treatments using a 4-MV unit with a mean post irradiation exposure of 0.034 μSv per fraction attributed to the presence of depleted uranium in the machine collimators) indicate that average monthly doses in the range 0.12 to 0.19 mSv can be expected for 30 daily treatments similar to those monitored. It was also found that 76% of the in-room dose was due to activated components in the treatment head and the accessories for the LINAC operated at 24 MV, the balance of the in-room dose being due to activated concrete.
In 1987, Powell et al. found the dose rate near a primary-beam wall to be about 0.06 mGy h−1, but it fell to about 0.006 mGy h−1 about 20 min after a 30‐min irradiation using 18-MV x-rays from a 20-MeV linear accelerator (Powell et al. 1987). Geiger-Mueller-based dosimeters worn by therapists indicated approximately 6 μSv d−1 due to activation products and suggested avoiding aluminum in items that would be in the beam.
In 1991, Almen et al. measured absorbed dose to the trunk and to the hands of technicians working with accelerators for radiotherapy using TL dosimeters for seven different accelerators (Almen et al. 1991). The contributions from induced activity in the accelerator and from radiation transmitted through the walls of the treatment room have been estimated separately. The total annual absorbed dose to both the trunk and hands was estimated to be 2 mGy, of which induced activity contributes one-third, or about 0.7 mGy. The authors noted that of the seven units tested, the one with the multileaf collimator had relatively low readings. They also noted that the absorbed dose rate at the isocenter of accelerators operated between 13 and 17 MV was dominated initially by 28Al (T1/2 = 2.3 min) and 62Cu (T1/2 = 9.7 min), resulting in a dose rate of approximately 6 μSv h−1 1.0 min after termination of the irradiation. An additional smaller dose-rate was attributed to 64Cu (T1/2 = 12.7 h), 187 W (T1/2 = 24 h), and 57Ni (T1/2 = 36 h).
It should be noted that the above doses were incurred using conventional EBRT, and the doses in the modern era with new techniques could be higher, as reported in the following studies. In 2002, Rawlinson et al. point out that the increased beam-on times that characterize intensity-modulated radiation therapy (IMRT) could lead to an increase in the dose received by radiation therapists due to induced activity (Rawlinson et al. 2002). To examine this, gamma ray spectrometry was used to identify the major isotopes responsible for activation at a representative location in the treatment room of an 18-MV accelerator (Varian Clinac 21EX) operated at 300 monitor units min−1 with a gantry angle of 0o. These were found to be 28Al, 56Mn, and 24Na. The decay of the dose rate measured at this location following irradiation was analyzed in terms of the known half-lives to yield saturation dose rates of 9.6, 12.4, and 6.2 μSv h−1, respectively. A formalism was developed to estimate activation dose (μSv wk−1) due to successive patient irradiation cycles. They derived equations for predicting the dose equivalents in μSv wk−1 to therapists due to activation products in the accelerator head as:
for the longer lived radionuclides, and
for 28Al, where F is the number of patients per week, M is the average number of monitor units per patient, td is the time between the end of the irradiation and the exposure to the therapist in minutes, and tr is the average time from the beginning of the irradiation for a patient through the end for that patient in minutes. The number of fractions only includes those for the high-energy beams on dual-mode accelerators. They estimate a typical therapist dose equivalent of 3 to 4 mSv y−1, but possibly becoming 17 mSv y−1 with high-energy IMRT treatments. They also identified 122Sb with a 2.8‐d half-life that built up over the 5‐d working week. To reduce dose, the authors recommend that IMRT treatments should be delivered at energies lower than 18 MV; that in multi-energy IMRT, high-energy treatments should be scheduled in the latter part of the day; and that equipment manufacturers should strive to minimize activation in the design of high-energy accelerators. These recommendations are also in accordance with Petrović et al. (2011).
In 2002, Evdokimoff et al. measured the induced radioactivity from a Varian Clinac 2100C/D operating at 18 MV (Evdokimoff et al. 2002). In this paper, they determined that the gantry, patient couch, wedges, floor, and air were all activated. The dominant components appeared to have a decay half-life of 2.5 min, except the gantry and port, which had half-lives closer to 10 min. The highest exposure rate, found at the port, was less than 20 μSv h−1. Airborne exposure was also measured in the study. The paper reports that the highest exposure from airborne radioactivity was 3 μSv h−1 on the roof next to the room exhaust. The authors therefore concluded that the activations cause an insignificant amount of exposure to the radiotherapy staff. The authors found no removable contamination from the gantry, patient couch, floor, and wedges due to this activation.
Perrin et al. measured dose rates around an Elekta Precise, finding about 4.5 μSv h−1 at the exit window 2 min following a typical 18-MV treatment (Perrin et al. 2003). They also report 0.4, 1.1, and 3 μSv h−1 at a typical position for a therapist (1 m lateral and 0.5 m away from the gantry at the level of isocenter), at the portal imager and on the patient support assembly, respectively. All dose rates fell to background by the following morning. From these measurements, they estimate a mean equivalent dose per 18-MV patient of 0.264 μSv h−1 and a maximum annual exposure of 2.5 mSv. They observe that this annual reading is not likely because of the rapid decrease in dose rate during the time the therapist is in the room after treatment. The authors also note that the setting on the treatment unit after the treatment greatly affects the dose rate, and by moving the motorized wedge into place, the dose rate falls by 75%. Closing the jaws from 10 cm × 10 cm to 5 cm × 5 cm reduces the dose rate in half.
In 2005, Wang et al. measured the exposure rates around several accelerators, including those equipped for IMRT using multi-leaf collimators, following normal clinical use and extended runs (Wang et al. 2005). They made measurements 100 cm below the source for a 10 cm × 10 cm field using 400 monitor units min−1 with the beam directed toward the floor. They found that the majority of the activation comes from neutrons above the collimators, and thus, using the jaws for field delimitation produced less activation than using the multileaf collimators. They also found that the dose equivalent rates around the unit in the morning increased from Monday through Friday. They found the maximum equivalent-dose rates for the 18-MV beam from a Varian Clinac 2300CD at the isocenter to be about 13 μSv h−1, with a half-life between 2.5 and 7 min. For a 10‐MV beam, the initial maximum rate was 0.3 μSv h−1. The authors present one equation to fit the dose from the three main radionuclides, unlike the two equations from Rawlinson et al. (eqns 1 and 2, above):
where ˙X represents the equivalent-dose rate total from background and the maximum values for the contributions from 28Al, 62Cu, and 187W, respectively. The lambdas stand for the decay constants for 28Al (0.30 min−1), 62Cu (0.071 min−1), and 187W (4.8 × 10−4 min−1). The irradiation time is given by ta, and t is the time since the end of the irradiation. Using eqn (3), the authors calculated that near the collimators, the equivalent dose rates would be about 1.3 μSv h−1 and a little less than 1 μSv h−1 1 m from the isocenter 2 min after delivering 1,000 monitor units. Measurements made following 1,000 monitor units with a 22-MeV electron beam showed no significant activation. In 2007, Vassiliev et al. measured the activation-related equivalent dose rates from a Varian Clinac 21EX accelerator that was operated at 18 MV, with and without a flattening filter (Vassiliev et al. 2007). Their results suggest that a reduction in dose rate of 20 to 30% 10 cm below the MLC can be achieved by using an accelerator without a flattening filter.
In 2008, Fischer et al. reported gamma dose rates at isocenter from the results of spectrum analysis ranging from 0.78 to 3.16 μSv h−1 after beam-off for an 18-MV beam, decaying to values between 0.18 and 0.54 μSv h−1 within 30 min (Fischer et al. 2008). From the graphical presentation of their data, the dose rate after a 2‐d weekend break would be approximately 0.08 μSv h−1. Measured dose rates were systematically higher by up to a factor of 2, which is attributed mainly to beta radiation entering the proportional counter used for the dose measurement, but that was not included in the calculation. No systematic dependence on machine properties or manufacturer could be identified. Assuming realistic working scenarios, the whole-body gamma absorbed dose values for the radiotherapy technologist staff working in the room ranged from 0.62 to 2.53 mSv y−1 depending on the accelerator design. Interestingly, three of the four units tested had a maximum photon energy of 15 MV, for which the projected annual dose exceeded that for the one that could operate at 18 MV. Gründel and Güthoff calculated Ho(10), the equivalent-dose rate at 10 mm deep based on measurements 70 cm from the isocenter every 5 s from 15 s through 200 s after exposures of 20 s and 60 s at 3 Gy min−1 using a 20 cm × 25 cm field size and 20 MV photons (Gründel and Güthoff 2008). They estimated an annual dose of just less than 1 mSv for a therapist’s body. From handling stainless steel wedges, they calculated annual doses to the hands of 362 μSv or 192 μSv, depending on whether the therapist handles the wedge 20 s or 60 s after the exposure terminates. Similarly, Ateia et al. (2008b) made measurements on a Siemens Primus operating at 18 MV and 5 Gy min−1 at isocenter with a 20 cm × 20 cm field size, following a 2.2-min exposure. They made measurements 50 cm lateral to the beam line and also in the center of the beam line (given in parenthesis), finding:
* At the front face, 4 μSv h−1 (10 μSv h−1);
* 15 cm under the jaws, 2.8 μSv h−1 (7 μSv h−1);
* 50 cm under the jaws, 2.6 μSv h−1 (4 μSv h−1);
* At the isocenter, 2.3 μSv h−1 (2.7 μSv h−1); and
* At the floor, 2.2 μSv h−1 (2.3 μSv h−1).
They estimate whole body doses of 3.5 mSv y−1 and a dose to the hands of 5 mSv y−1 based on 20% of the treatments using 18-MV photons. Adjusting for an anticipated decrease in the use of high-energy beams in the future, they projected annual whole body doses closer to 1.5 mSv. Petrović et al. (2011) projected annual whole body doses of 0.25 to 1.7 mSv for an Electa Precise LINAC.
Kalef-Ezra measured dose rates around a Varian Clinac 2100 DHX LINAC operating at 18 MV (Kalef-Ezra 2011). Table 3 highlights some of his findings, in terms of Ho(10). The report contains a great deal of information, and the interested reader is referred to the original for the more complete data. Kalef-Ezra presents many recommendations to reduce the exposure of staff, several of which are included in the recommendations in this report.
Summarizing this information, the activation dose rate in the treatment room starts at a low value at the beginning of a treatment day but goes through cycles of incremental increases with the exposure of high-energy x-rays and subsequent decay, and it may reach a saturation value. Similarly, over the course of a week, the activity increases from Mondays to Saturdays and subsequently decays over the weekends.
The occupational exposure depends upon many factors, including the usage of high-energy beams at the hospital and the amount of time therapists spend in setting up patients. The differing doses projected by authors of the studies cited result from measurements being made on different machines, using different detectors, and making different assumptions about workload and the exposure geometry. For conventional treatment practice, using approximately 50% of the workload with high-energy beams, an annual activation dose to radiation therapists can be up to 4 mSv. This number can be higher if a significant fraction of IMRT treatments is executed with high-energy photons. The higher dose rates available in current accelerators, compared with those for which data has been reported, could be of concern also due to the short half-lived component of the activation dose.
However, whole-body radiation badge readings for therapists usually report a relatively low occupational dose. This is mostly because, contrary to the assumption made by Rawlinson et al. (2002) that a therapist spends all the available time during patient setup in the treatment room near the treatment unit head, much of the time is spent in the treatment room away from the head or in the control room area, particularly with computer-driven units. Ahlgren and Olsson (1988) have also reported maximum estimated dose to the body and the hands of therapists to be 2 mSv y−1, which is well within the acceptable exposure limits to personnel. Even though there is an increased amount of exposure to the body and hands from the radionuclides activated by high-energy accelerators, the dose levels suggest that the risk to personnel and patients is small.
Presently and in the future, with the high proportion of patients receiving IMRT, in which many beams and beam segments are used to optimize dose distributions, low-energy beams (such as 6 MV) are increasingly being used in place of high-energy photons. In addition, all modern accelerators have multi-leaf collimators, and most facilities use these for beam shaping in place of blocks placed on trays at the bottom of the treatment head. Physical wedges, also at the bottom of the treatment head, are increasingly being replaced by dynamic, virtual, or universal wedges that do not require insertion by a therapist. These changes in practice should reduce the dose received by therapists due to activation products below the estimates given above.
DOSE REDUCTION STRATEGIES
1. Use lower-energy beams where possible. The photon activation cross section initially increases with increasing photon energy, reaches a peak called the Giant Dipole Resonance, and then decreases with increasing photon energy. There is a threshold, typically around 8 to 10 MeV (Table 1), below which such a photonuclear interaction does not occur for most of the commonly used materials in the treatment head. Therefore, an obvious way to reduce the dose is to use a lower photon beam energy, ideally below the threshold, such as 6 MV, as most IMRT procedures use today. In addition, neutron-capture interaction cross sections depend heavily on the type of target materials and the neutron energy. However, if the use of higher-energy photons is required for the optimization of patient dose distributions, then a number of methods can be considered to reduce dose to radiation therapy staff, as discussed in the following recommendations;
2. Manufacturers should avoid or minimize the use of materials with high photon and neutron activation cross sections in the construction of accelerator components, treatment couch, and accessories;
3. Restrict the use of high-energy treatment to the latter part of the day to reduce the buildup of activity throughout the day;
4. Consider delayed entry of staff to the treatment room for 30 s for treatments using 15-MV photons and 1 min following a higher-energy beam treatment;
5. Minimize the use of compensators, external wedges, and blocks, using virtual or dynamic wedges and multileaf collimators where possible;
6. Following high-energy photon treatments, close the collimators from the control panel before entering and, if applicable, move any motorized wedge into the beam axis;
7. Move the table and gantry from the control panel rather than from the pendant at the couch following high-energy photon treatments when that can be performed safely;
8. If feasible, assign pregnant staff to lower-energy accelerators;
9. Service on the head of high-energy accelerators should be performed with appropriate hand-held radiation monitors and good radiation protection techniques; and
10. Consider using a hand-held radiation monitor to assess radiation levels around brass or copper caps used for small-field determination of output factors for photon beams above 10 MV (Zhu et al. 2009).
CONCLUSIONS FOR PHOTON-BEAM RADIOTHERAPY
The additional exposure to the patient due to induced radioactivity secondary to external photon beam radiotherapy is negligible compared to the overall radiation exposure as a part of the treatment. The additional whole body exposure to the staff due to induced activity is estimated at about 1-2 mSv y−1. This is well below the allowed exposure limits that have been set at a conservatively safe level. Therefore, the potential hazard to staff from induced radioactivity in the use of high-energy x-rays is considered to be low, and no specific actions applying to all facilities are considered necessary or mandatory. Notwithstanding, facilities may do well to review the personnel dosimetry for staff working with high-energy photon beams, potentially including extremity dosimeters for a period, and consider whether incorporating the above recommendations would be prudent in their setting. The dose reduction strategies suggested above need be followed only if these actions are reasonable and practical in the individual clinics.
B. POTENTIAL HAZARDS DUE TO ACTIVATION IN PROTON RADIATION THERAPY FACILITIES
Forty years after the proton’s discovery, it was suggested that the particle might be well suited for radiotherapy (Thomson 1907; Wilson 1947). This was because protons were considered to be capable of providing superior dose depositional characteristics to photons. It is the fundamental radiological properties of the two types of radiation that account for this. X-rays follow exponential attenuation behavior, whereas protons undergo multiple Coulomb scattering that results in continuous slowing down of the particle until it stops (Bragg 1912). This results in a sharp dose fall-off after localized energy deposition, giving protons an inherent geometrical advantage over most modalities. Recently, the potential to improve the differential between dose delivered to the target volume and dose to normal structures has resulted in the increased interest in particle beam therapy (Orecchia et al. 1998).
PROTON BEAM DELIVERY TECHNIQUES
Energetic proton beams are produced by particle accelerators. A native proton beam, or pencil beam, is described transversely by a primary Gaussian sigma of 3 to 5 mm at higher energies and longitudinally by a pristine Bragg curve. Therapeutic proton beams, however, typically require longitudinal and transverse spreading for use.
Two different methods exist for laterally spreading the beam: physical scattering and electromagnetic scanning. Scattering uses one, or classically two, high-atomic-number (Z) material scatterers to spread the beam transversely (Koehler et al. 1977). Continuous uniform scanning uses electromagnets to sweep the beam across the treatment field (Farr et al. 2008). In some uniform scanning systems, a thin scattering element is still retained to control the size of the scanning spot (Kanai et al. 2006).
Two different methods also exist for longitudinal (energy) beam spreading. A rotating range modulator with a series of thicknesses of range loss (pull-back) material can be used to generate a spread out Bragg peak (SOBP) longitudinally (Koehler et al. 1975). A common material for range modulators is acrylic, although brass ridge filters are sometimes used. Alternatively, a discrete energy layer delivery approach can be used. With a synchrotron, the discrete energies are tuned; cyclotron energy layer delivery requires a range shifting process that is performed in the nozzle for continuous uniform scanning. Graphite and acrylic or other low-Z materials are the choice for range shifters (Farr et al. 2008).
To preserve the proton beam penumbra in both types of systems, final transverse beam collimation is required in proximity to the patient. Range compensation is also performed to account for inhomogeneities, patient curvature, and distal target volume profile. The transverse collimation is performed with an aperture of brass or Lipowitz’s metal [Cerrobend (Marmon Electrical & Plumbing Products-Distribution, Inc., Chicago, IL) and other trade names] as described in Table 2, and the range compensation is performed with acrylic (PMMA) or wax bolus material.
A further development of proton therapy delivery technology is to scan the beam point by point while modulating the intensity between points, or intensity-modulated proton therapy (IMPT) (Pedroni et al. 1995). IMPT systems use an optimized pencil beam profile and planning/calculation system, usually obviating the need for a final aperture and compensator. The IMPT approach also uses energy layer delivery as described for continuous uniform scanning, except the energy shifting or tune is kept close to the accelerator so as to maximize the source distance from the scanning magnets.
ACTIVATED MATERIAL COMPONENTS
For the purposes of this review, there are two major sets of components of materials incidentally activated from a proton therapy accelerator: permanent system components and consumable system components.
Permanent system components
Permanent components remain part of the system until decommissioning. Examples of permanent components are dipole and quadrupole magnets, beam scatterers (Fig. 1), anti-scatter collimators, range modulators, the delivery “snout,” and transmission ionization chambers.
Consumable system components
Consumable components are limited to final apertures, compensators, and patient immobilization material. Both scattering systems and uniform scanning systems activate consumable components. Pencil beam scanning systems typically do not.
ASSESSMENT OF HAZARDS FROM ACTIVATION PRODUCTS
Many reports are available on proton activation of metals, and the reader is referred to them for details (Schneider et al. 1987; Bodemann et al. 1993; Schiekel et al. 1996; Sisterson et al. 1996; Agosteo 2001, 2009; Pelliccioni 2011). There are also reports covering proton activation at therapy centers in particular (Faßbender et al. 1997; Rouse et al. 1997; Sisterson 2002; Moyers and Lesyna 2009).
The Harvard Cyclotron Laboratory (HCL) study concluded that the primary elements used in acrylic, oxygen, and carbon produce short-lived radioisotopes with half-lives of seconds up to 20 min in greatest abundance (Sisterson 2002). This finding is supported by surveys at this clinic that suggest these materials can be discarded when the patient completes the treatment course. The remaining consumable materials of interest are brass and Lipowitz’s metal. Brass is used due to its relative ease of machining compared to harder metals, its efficient stopping power due to its ratio of physical density to Z, and its lower secondary particle production compared to higher-Z elements. Lipowitz’s metal, with its low melting temperature and high density, allows easily customized and relatively thin apertures.
Material activation dependencies
Proton therapy centers generally treat a wide range of cases requiring various proton energies, target volumes, and doses. Useful therapeutic proton energies are in the range from 70 to 250 MeV. Higher energies, such as 300 to 350 MeV, could be of interest in the future to allow proton radiography. Field sizes are in the range of sub-centimeter for eye treatments to 30 cm or more for large areas or long targets. Therapy dose prescriptions range from 10 to 100 Gy with a variety of fractionations, from single-fraction radiosurgery to low-fraction, twice-a-day (BID) treatments. The most common fraction is 1.80 Gy d−1 split among several fields, very similar to that used in photon radiotherapy. Using this common fractionation, an entire average course of proton therapy is usually treated 5 d wk−1 for 5 or 6 wk. All of these parameters, as well as the material composition, determine the level of incidental activation for a particular component placed in the proton beam. The sum of all individual components used, along with their daily activation and decay cycles for the individual radionuclides, form a composite facility activation level.
In summary, the incidental material activation levels are expected to depend on:
* incident proton energy;
* proton dose per fraction;
* total proton dose for treatment course;
* inter-fraction time;
* material composition;
* number of fields treated at the facility;
* field size (physical material cross-section); and
* decay time after treatment course completion.
Proton activation of brass
Brass activation products
The elemental composition of brass varies slightly depending on the standard referenced, but Table 4 gives typical values. A published study of brass activation by proton bombardment was performed for energies ranging from 79 to 350 MeV. Numerous radioisotopes were investigated, including 67Ga, 62,65Zn, 55,56,57,58Co, 52,54Mn, 51Cr, 48V, 47Sc, 22Na, and 7Be (Faßbender et al. 1997). By considering the measured activation cross sections and half-lives of these radioisotopes, their long-term significance may be determined. Long half-life radioisotopes will continue to build up in bombarded materials until the rate of radioisotope production is equaled by the rate of decay. Properties of the longer-lived brass activation isotopes are shown in Table 5.
Changes in activation cross section with proton energy, along with radioisotope half-life, are important determinants of the activation levels observed in brass. Available brass activation cross sections for these long-lived radioisotopes are presented in Table 6. The 56,57,58Co cross-sections vary by as much as 40% over the proton energy range of 75 to 351 MeV and trend toward lower activation at the highest energies. In general, the 54Mn cross sections increase with energy and reach the highest level when proton energies are 351 MeV.
Brass activation magnitude
Short-lived radioisotopes will generally not contribute to the total induced activity in apertures at a proton facility. The total activity will be an amalgamation of the long-lived products reported in Table 4. Using assumptions of a low patient load of two patients treated on 1 d wk−1 with 200 MeV protons for a total 1 h beam-on time, Faßbender et al. (1997) calculated 0.41 GBq y−1 of activity from proton activation in a brass collimator.
A calibrated survey meter ( Model: FH40 G; Thermo Fisher Scientific, Waltham, MA USA) was used by one of the authors at the Indiana University Health Proton Therapy Center to acquire short-term activation data from a “typical” brass patient aperture in active use for prostate cancer treatment. Immediately after the patient field delivery (208 MeV, 0.90 Gy, 60Co equivalent, 6‐cm-diameter field), surface contact indicated 1.8 mSv h−1. After 15 min, the surface contact reading was 0.80 mSv h−1, and the reading at 1 m was 15 μSv h−1. After 1 h, the surface contact reading was 170 μSv h−1, and the reading at 1 m was 8 μSv h−1. These results suggest the appropriateness of ring dosimeters for staff who routinely handle the patient apertures; however, the aperture activation levels decay quickly over the minutes following a patient’s daily treatment. The long-lived isotopes are responsible for the slight elevation over background, and the sum of all activated material generates a level as discussed above. Over 2 y of operation at about half capacity per shift (10 patients per room per 8 h), the highest reported deep-body dose equivalent was 50 μSv in 1 y, and the highest extremity dose reported was 700 μSv based on personnel monitoring. These values project to a maximum annual dose equivalent of 100 μSv for the deep body and 1.5 mSv for extremities. These values are consistent with the above analysis regarding short- and long-term activation products in brass apertures used at the facility. The majority of monthly personnel monitoring reported doses were below the minimum standard for reporting of 10 μSv. Reported doses above this threshold were generally associated with routine handling of activated patient-specific apertures.
Proton activation of Lipowitz’s metal
Lipowitz’s metal is the primary material used for apertures at the James M. Slater, M.D., Proton Treatment and Research Center at Loma Linda University Medical Center (LLUMC). As discussed below, Lipowitz’s metal is composed of metals with higher atomic numbers than brass. This results in longer-lived radioisotopes. Each facility will need to decide the holding time based on their practice; the LLUMC holds apertures for 4 mo before reusing the metal, following a survey to confirm that the radiation level has fallen to background.
Lipowitz’s metal activation products
The elemental composition of Lipowitz’s metal is reported in Table 2. Besides the radionuclides of short half-life formed after a 250-MeV proton irradiation of Lipowitz’s metal, a number of long-lived isotopes were identified spectroscopically (Table 7). The induced activity was compatible with that anticipated by Monte Carlo simulation using the LAHET code [Prael and Lichtenstein (1989)]. An initial 0.6 mSv h−1 surface contact rate was found. The integrated surface dose during the first 15 min post irradiation was less than 0.3 mSv. The annual whole body and shallow equivalent doses to those who handled the apertures were 0.8 mSv and 2.2 mSv, respectively. The activation products of the Lipowitz’s metal decay mainly by electron capture, thus substantially reducing the shallow dose.
Lipowitz’s metal activation levels
The investigators considered the induced activity levels in the Lipowitz’s metal apertures. It is reported that the surface contact dose on a Lipowitz’s metal aperture immediately following a patient’s treatment was 0.6 mSv h−1. Under the same conditions, Landauer Type U dosimeters were allowed to remain in surface contact for 15 min. The report on the returned dosimeter did not exceed the minimum reporting level of 0.3 mSv (Rouse et al. 1997). In addition, the report states that the average personnel dosimetry results for the workers handling the activated Lipowitz’s metal apertures are 0.8 mSv y−1 whole body dose equivalent and 2.2 mSv y−1 shallow dose equivalent for the extremities. The authors also addressed inhalation hazards particular to the working of Lipowitz’s metal. The reader is referred to Rouse et al. (1997) for the discussion.
Proton activation of multileaf collimators
Moskvin et al. simulated activation of a brass aperture and a tungsten MLC secondary to proton treatment (Moskvin et al. 2011). Direct comparison of the dose to personnel becomes difficult because each patient would have one aperture used through his or her course of treatment, while the MLC would be used for all patients. Due to the long half-life of the tungsten progeny 179Ta (1.82 y), the activity in the MLC continues to build over a year of use. Moskvin et al. (2011) calculate a dose equivalent rate at 10‐cm depth at the patient-side surface of the MLC of about 100 μSv h−1. While a therapist’s hands and body would be more distant, the authors suggest that this exposure could exceed regulatory limits. They further note that the dose equivalent rate on the source-side surface of the MLC is about 2.7 mSv h−1, which becomes concerning for service personnel working on the unit.
MANAGEMENT OF ACTIVATED MATERIAL
Incidental material activation is a radiation safety concern at a proton therapy center. Whether the material is permanent or consumable (used and then appropriately discarded or recycled), the component dictates how the material is managed.
Management of permanent system component activation
Permanent system components are monitored routinely for activity as part of the facility radiation protection plan. Unless the facility undergoes decommissioning, disposal is not an issue. Should disposal be necessary, the material must either be retained until decay to background level occurs or properly shipped to a licensed holding facility. Moyers and Lesyna (2009) found that within 20 min of shutdown for a synchrotron, the average exposure rates at 30 cm did not exceed 17 μSv h−1. Readings taken from body dosimeters of maintenance personnel averaged less than 150 μSv y−1, with the maximum less than 1 mSv y−1.
For comparison, measurements from one author’s facility in a proton therapy treatment room using a double passive beam spreading system were acquired with a calibrated NaI survey meter (Ludlum Measurements, Inc., Sweetwater, TX USA). Immediately following a typical patient treatment, readings at the patient treatment position (isocenter) and 1 m from the scattering system (Fig. 1) yielded 140 μSv h−1 and 60 μSv h−1, respectively. Before the next patient treatment, about 20 min later, readings were 70 and 35 μSv h−1 at the same respective locations. Radiation surveys performed before treatments in the morning indicated slightly lower values (the radiation level in the treatment room builds during the day and then decays overnight). However, the values stated here are representative of average room activation levels. The facility radiation protection policy requires compliance with applicable state and federal regulations regarding the facility, including all treatment room components.
Management of consumable component activation
It is expected that patient-specific consumable components will be disposed of or recycled after the patient completes the course of proton therapy and decay to background has occurred. Because of activation from the proton beam, the consumable components require storage and monitoring until they have reached background level. At active centers, these incidentally activated consumable materials represent the most significant area of concern. Apertures are sorted by the month of patient treatment course completion. The decaying apertures are monitored periodically for activity level. After a batch has decayed to background level as determined with a sensitive instrument such as a NaI counter, it is released from the facility, typically to a metal recycling facility due to its monetary value.
SUMMARY AND CONCLUSIONS FOR PROTON BEAM RADIOTHERAPY FACILITIES
The primary role of proton therapy accelerators is to produce useful beams for medical treatments. There are a variety of types of accelerators and therapeutic beam-generation techniques. In common to all is the desire to minimize activation of beam components, whether permanent or consumable. Hence, any activation is incidental to the function of the system.
Fixed components are generally composed of brass, lead, acrylic used for beam modification, and steel inside the accelerator. These components are routinely surveyed to assure staff and patient safety as part of the facility radiation protection plan. Patient-specific beam-modifying components used in passive spreading and uniform scanning are composed of some combination of brass, Lipowitz’s metal, acrylic, and wax. The composition of acrylic and wax leads to short-lived radioisotopes that decay to background within minutes of medical treatment completion; no special requirements are needed for their disposal. Wax can be reused after approximately three hours. After being held for 2 wk, the Lipowitz’s metal is surveyed to confirm decay and released for reuse at the facility. Metal removed from use is held for a 4-mo decay period before being released for recycling by the supplier. Practical experience indicates that, depending on the patient-specific combination of proton-field properties and dosage, the brass apertures decay to background levels within 3–4 mo and can then be released.
It is projected that although the majority of patients treated worldwide with proton therapy are still treated with scattering beam delivery systems (BDS) as described here, there is a transition underway to IMPT types. An important aspect with regard to activation is nozzle efficiency, defined as the ratio of outgoing to incoming numbers of protons. Because scattering systems rely on interactions to produce therapeutic fields, as opposed to magnetic spreading, their efficiency is relatively lower. This is significant because the interactions include undesirable inelastic processes that activate the physical beam-modifying devices. As IMPT uses fewer or no patient-specific modifying devices (apertures/compensators) and may offer an order of magnitude increase in BDS nozzle efficiency (Clasie et al. 2010; Hecksel et al. 2010; Mukherjee et al. 2011), the activation of fixed BDS components is expected to be reduced and activation of consumable components mostly eliminated in the future. Although the activation of fixed components may reduce with increased scanning nozzle efficiency, the workload at scanning facilities may rise to meet the available treatment time, reducing the total effect somewhat.
C. POTENTIAL HAZARDS DUE TO ACTIVATION IN FAST NEUTRON RADIOTHERAPY FACILITIES
Fast neutron therapy beams are used to treat certain types of cancer. The neutrons are produced by bombarding a target, typically beryllium, with protons or deuterons in the 30 to 66-MeV range. Neutrons are emitted after compound nucleus formation, as is the case with proton bombardment, or they originate from the deuteron in a “stripping reaction.” As a result, the neutron fields from protons and deuterons on beryllium have different spectral shapes. In general, the neutron spectra extend continuously up to the maximum energy of the accelerated particle, characterized by a large component in the low-energy range for proton-produced beams and by a peak at about 0.4 times the maximum energy in beams produced by deuterons (Bewley 1989). These variations in neutron spectra have very important consequences on activation of materials caused by fast neutron beams.
In most cases, neutron therapy is delivered externally using techniques similar to photon beam therapy; i.e., the radiation is directed isocentrically from different angles. The neutron beams are produced to achieve the desired field shape in a way analogous to photon beams from medical linear accelerators using collimators made of materials with high attenuation properties.
The neutrons from the primary beam, as well as scattered neutrons, can activate materials in the therapy room, including the treatment equipment (treatment head, collimator, and treatment couch), the patient, the structural materials (building), and room components, as well as any ancillary equipment present in the room during beam delivery.
The extent of the activation is proportional to the isotopic abundance of the target nucleus, the cross section of the nuclear reaction for any given neutron energy, and the distribution of neutron intensity in the room. The level of the exposure from induced activation also depends on the half-life of produced radioisotopes and the energy of emitted gamma-rays. The beta and gamma radiation from the activated materials result in radiation exposure to people entering the therapy room or coming in close contact with activation products.
The hazards due to the activation following neutron irradiation can present themselves by two different pathways: first, as external exposure to γ-emitting isotopes and, second, due to external as well as internal contamination of α-, β- and γ-emitters.
In this section, major sources of activation associated with fast-neutron therapy will be presented, radiation levels in the therapy room after irradiation will be described, and radiation hazards for different population groups will be discussed.
SOURCES OF INDUCED ACTIVITY
Unlike with photon irradiation, all materials become radioactive as a result of neutron interactions. While the activation by thermal neutrons results from (n,γ) neutron-capture reaction, fast neutrons can activate materials through multiple processes: (n,2n), (n,3n), (n,p), and (n,α). A list of the most common activation products in a patient’s body, as well as in various structural materials and metals, is shown in Table 8.
Activation of the therapy equipment
The major sources of induced activity in neutron therapy are the cyclotron itself, various components in the treatment head, and the collimator. Several authors have investigated the activation of the therapy equipment and the most relevant radionuclides, which include 28Al, 56Mn, 24Na, 65Ni, 58Co, 60Co, 64Cu, 66Cu, and 187W (Bonnett 1983; Blake et al. 1988; Bonnett et al. 1988; Finch and Bonnett 1992; Yudelev et al. 1997).
The radiation arising from the parts inside the cyclotron and treatment head usually does not contribute significantly to the exposure levels during normal operation and should be considered only when these components are removed during maintenance, repairs, or decommissioning. However, in the case of a superconducting cyclotron, such as at Wayne State University in Detroit, the iron in the magnet yoke is likely to be the largest source of background activation in the treatment room.
The exposure originating from the beryllium target and inner parts of the treatment head will affect the therapy personnel standing directly in front of the open collimator aperture. For example, Vynckier et al. (1989) registered initial dose rates of 5 mGy h−1 close to the exit window of a cyclotron producing 65-MeV protons on a beryllium target, written as p(65)+Be, equipped with a collimator made primarily of iron. The dose rate was reduced, however, by an order of magnitude following automatic withdrawal of the target post-irradiation (Vynckier et al. 1989).
Selection of proper materials can help in reducing personnel exposure. For instance, the use of aluminum, manganese, cobalt, and copper in the equipment will increase the activation levels and should be avoided, while pure iron with low manganese content is preferred over steel as a collimator material (Bonnett et al. 1980; Finch and Bonnett 1992; Yudelev et al. 1997). The spectroscopic analysis indicated that 52 V, 53Fe, 55Mn, and 56Mn were the main short-term activation products in the collimator made primarily of iron (Vynckier et al. 1989).
The assessment of exposure levels due to activation in neutron therapy is often complicated by the fact that certain materials provide self-shielding and substantially moderate the neutron beam. It was reported that activation of steel is approximately 10 times that of tungsten throughout a working day in neutron beams produced by p(66)+Be or d(48.5)+Be (48.5-MeV deuterons on beryllium), written as d(48.5)+Be (Jones et al. 1997; Maughan et al. 2001). However, in a d(48.5)+Be beam, 30‐cm-thick materials will result in exposure levels from steel 50 to 100 times less than that from tungsten (Maughan et al. 2001).
Activation of building components and air
The residual radiation levels in a neutron therapy room can be reduced by proper selection of the building materials, specifically in the areas bombarded by the primary beam. Limestone concrete aggregate is preferred over concrete containing silicon (Eenmaa et al. 1987; Yudelev et al. 1997).
As shown in Table 8, the main activation products in air due to (n,2n) reactions are 11C, 13 N, and 15O with half-lives of 20.0, 10.4, and 2.04 min, respectively. Ten Haken et al. investigated the air activation by p(66)+Be neutron beams at Fermilab (Ten Haken et al. 1983). They concluded that, with typically 6–8 air changes per hour and the time necessary to access the room of about 0.5 min, there is minimal reason for concern. The neutron therapy facility at Wayne State University is designed for 12 and 31 air changes per hour in the therapy area and in the cyclotron pit, respectively.
Activation of ancillary equipment
Any equipment exposed to neutrons becomes radioactive and must be assessed prior to removal from the treatment area. Finch and Bonnett (1992) measured dose equivalent rate levels of 40 μSv h−1 and 70 μSv h−1 at the surface of tungsten blocks and immobilization masks, respectively, 2 min following a 1 Gy irradiation. They concluded that the main sources of activity were the isotopes of 184Ta and 187W in the blocks and 11C in the masks (Finch and Bonnett 1992).
The issues associated with assessment of activation are complex and should be dealt with on a case-by-case basis. The composition of targets is often not known, which complicates the analysis of measured exposure without determining the spectrum.
Culp suggested a simplified method to predict accelerator-produced metal activation products for the purpose of radiation protection (Culp 2007). The process is based on determining the most likely contaminants from a list of possible reactions for a given beam based on abundance, half-life, and Q-value. Using the method, an initial list of 39 potential contaminants was reduced to 7 and 15 for the (n,γ) and (p,n) reactions, respectively. Radionuclides of concern were determined for stainless steel (55Co, 57Co, 58Co, 61Cu, 55Fe, 54Mn, 56Mn), tungsten (185W, 187W), nickel/copper alloy (61Cu, 64Cu, 66Cu), and titanium/nickel alloy (61Cu, 51Ti, 47V, 48V, 49V).
At a superconducting cyclotron neutron therapy facility in Detroit, a procedure to handle and dispose of radioactive products was established based on existing recommendations by various agencies. All products removed from the cyclotron and treatment area are checked for activation. If exposure levels exceeding twice background are registered, a considerable amount of radioactive contaminants is assumed. Presence of removable activity is assessed using a wipe test in a scintillation counter following decontamination, if necessary. The products are sorted according to their reusability and stored in a secure location. The disposable materials are checked periodically for residual activity prior to their removal. In an attempt to establish guidance for release of neutron-activated products, several categories were identified (Table 9) (Dodge, 2004). In general, equipment used routinely in the neutron therapy room should be dedicated to this room, and all devices should be kept away from the primary beam.
Activation of patients
The patients become somewhat radioactive from the interaction of the neutron beam with tissue. The primary nuclides of concern are 15O (2.04‐min half-life), 13N (9.97 min), and 11C (20.0 min). All are positron emitters. Finch et al. reported on levels of activity at the skin surface induced in patients over the treated area ranging from 105 to 165 μSv h−1 at 2 to 4 min after treatment, decaying to 40 μSv h−1 10 min after treatment (Finch and Bonnett 1992). Tatcher et al. (1987) estimated an initial exposure rate of 23 μSv h−1 from a patient, immediately after a dose of 1.7 Gy of neutrons was delivered. This exposure resulted in a dose of about 0.8 μSv per fraction to a therapist.
The activation of tooth fillings in patients treated with p(66)+Be neutrons was reported by Symons and Jones (1993). Immediately after 2 Gy of absorbed dose, a rate of 50 μSv h−1 was measured at the filled tooth, which was double that measured at the unfilled tooth. The increase of activity was ascribed mainly to 106Ag and 116In, the products of 107Ag(n,2n)106Ag and 116Sn(n,p)116mIn reactions, respectively.
The exposure to surrounding tissues as a result of activation of prosthetic implants was investigated by Kouskoulas and Yudelev (2005) and found to be negligible compared to levels of radiation caused by neutrons. The activities measured 1 min after 1 Gy of d(58)+Be neutrons delivered to three samples of CoCr, Steel, and TiVAl were 0.2, 0.18 and 0.34 GBq, respectively. The authors concluded that due to the presence of long-lived isotopes in some of the prosthesis materials, the activation should be considered only if the device is removed from the patient.
RADIATION LEVELS IN THE THERAPY ROOM AFTER IRRADIATIONS
Radiation levels in the room following neutron activation have been reported by multiple groups (Bonnett et al. 1980, 1988; Bonnett 1983; Rosenberg et al. 1984; Eenmaa et al. 1987; Smathers et al. 1987; Vynckier et al. 1989; Finch and Bonnett 1992; Yudelev et al. 1997). Peaks corresponding to 28Al, 57Co, 60Co, 64Cu, 56Mn, 24Na, and 187W were determined in the activation spectra measured at various locations in the treatment room and at different times post irradiation. Also identified were 15O, 13N, and 11C, although these are more closely related to patient irradiation.
The spatial distribution of activity in the treatment room is not uniform and correlates with proximity to the neutron beam and the neutron source (target), with the highest levels registered around the treatment head. The temporal distribution depends on beam use patterns.
For a typical therapy operation mode, a short rise in fast-decaying activity is observed immediately after the treatment session. This could be correlated with the presence of short-lived activation products such as 15O (2.04‐min half-life) and 28Al (2.2‐min half-life). In several facilities, a 1 to 2 min waiting period is recommended before entering the treatment room.
Over the course of a treatment day, the exposure levels in the room rise from a few μSv h−1 to several times 100 μSv h−1, approaching saturation levels after a few hours of operation, with higher levels immediately after the end of a run. This buildup of activity could be associated with the presence of 56Mn (2.6‐h half-life) among the activation products. This activation pattern was observed at any location in the treatment room, although with higher levels near some equipment and correlated with materials in the collimator.
RADIATION EXPOSURE FOR DIFFERENT POPULATION GROUPS
It is unavoidable that the therapists who set up the patient receive radiation exposure from activated materials in the room. They have to enter the room shortly after patient irradiation, when some of the short-lived activation products have not yet decayed sufficiently. Exposure of this group is monitored, and reported values range from 2.3 μSv to 7.1 μSv per individual and patient field treated (Bonnett et al. 1980; Bonnett 1983; Rosenberg et al. 1984; Eenmaa et al. 1987; Smathers et al. 1987; Vynckier et al. 1989; Finch and Bonnett 1992; Risler et al. 1995; Yudelev et al. 1997).
In the study from Detroit Neutron Therapy Facility using a d(48.5)+Be beam, monthly individual total body exposure on the order of 1 to 2 mSv was reported for therapists (Yudelev et al. 1997). To keep personal exposure of radiation therapists within annual ALARA of 12.50 mSv, each individual was rotated for a period of about 6 wk. About three to four rotations per year were observed.
Following the established working routine and being aware of the radiation environment may help to reduce individual exposure. A study of collective monthly doses received by the radiation therapists demonstrated that gain in experience, together with introduction of treatment techniques allowing remote operation, resulted in a drop of exposure levels (Yudelev et al. 1997). For instance, entering the treatment room to take localization films increases the therapists’ dose almost twofold. The development of remotely controlled electronic portal imaging systems for neutron-beam treatment units could reduce that dose. In addition, it was recommended to observe a 2‐min “cool off” period prior to entering the treatment vault post-irradiation.
Another group routinely exposed to residual radiation in the room includes engineers and physicists servicing the equipment or performing measurements. These personnel are also monitored, and annual values typically below 5 mSv have been reported for individuals. Some individuals could come close to the limit with monthly badge reading of about 3 mSv per month. Doses in the 15 to 20 mSv range for the yearly collective dose have been reported (per NRC definition of the “collective dose” as the sum of all individual doses for the group). Because of the broad nature of the work performed by this group, exposures can vary considerably (Bonnett 1983; Smathers et al. 1987; Vynckier et al. 1989; Risler et al. 1995; Yudelev et al. 1997).
Bonnet drew attention to the potential internal hazard to staff due to ingestion of activation products while handling damaged components during repair and maintenance. He identified isotopes of 184Re and 65Zn, from deuteron activation of tungsten and copper, respectively, in a d(15 MeV)+Be cyclotron at the Western General Hospital MRC Cyclotron Unit, Edinburgh (Bonnett 1983).
Other hospital staff, such as physicians and nurses, interact infrequently with the patient inside the treatment room, and their exposure is considerably less compared to the therapists and the engineering staff. Typically, the exposure of this group, as monitored by radiation badges, doesn’t exceed the levels observed for personnel not involved in neutron therapy.
In addition, some members of the general public, such as visitors or general hospital maintenance staff performing building-related tasks, occasionally may enter the room. In most of these instances, the exposure is very brief, as general room background drops to a few tens of μSv h−1 within minutes after treatments. For longer-lasting tasks to be performed in the room, exposures must be determined on a case-by-case basis.
SUMMARY AND CONCLUSIONS FOR NEUTRON RADIOTHERAPY FACILITIES
The activation and radiation hazards are probably most severe at neutron radiotherapy facilities compared to other modalities. The fact that neutrons, direct or scattered, will activate, to a certain extent, all materials in their path makes it imperative to address the activation issues during the planning and operation of such facilities.
The activation levels in the treatment room during the course of operation may vary, although they follow a common pattern: spikes of high exposure immediately post treatment session due to the short-lived isotopes, and a gradual increase in background caused by the longer half-life products. Materials used in facility design, building, collimator, patient support system, beam stop, etc., should be chosen based on their activation and self-absorption properties. The activation of metals is higher than that of plastic materials. On the other hand, the former are better attenuators due to their higher density.
In order to reduce personnel exposure, a delay of up to 2 min could be introduced prior to entering the treatment vault post-irradiation. This time interval was found to be effective in reducing exposure caused by the isotopes with half-lives up to several minutes. After four to five irradiations, however, the exposure levels follow the build-up of induced activity of the isotopes with half-lives up to several hours and relatively less of the short-lived components.
Patients’ exposure following the activation is negligible compared to the therapeutic doses received. The exposure hazard from these patients to the surroundings is minimal, although it should be taken into account in certain circumstances.
Any equipment and materials removed from the facility are suspect to having been activated and have to be checked prior to storage or disposal according to existing regulations for the handling of radioactive material.
The personnel exposure values reported in the reviewed literature are not insignificant, although well within radiation protection guidelines. In order to keep the individual exposure ALARA, operational protocols specific to the facility should be developed, which include routine training and in-service sessions, periodic personnel rotation, as well as constant individual exposure monitoring.
The personnel exposure is, with few exceptions, from photons and beta particles only. Neutron exposure outside the shielded rooms has been reported and found to be insignificant (Bonnett 1983; Bonnett et al. 1988; Finch and Bonnett 1992; Yudelev et al. 1995).
Agosteo S. Radiation protection at medical accelerators. Radiat Protect Dosim 96: 393–406; 2001.
Agosteo S. Radiation protection constraints for use of proton and ion accelerators in medicine. Radiat Protect Dosim 137: 167–186; 2009. DOI 10.1093/rpd/ncp187.
Ahlgren L, Olsson LE. Induced activity in a high-energy linear accelerator. Phys Medicine Biol 33: 351–354; 1988. DOI 10.1088/0031‐9155/33/3/004.
Almen A, Ahlgren L, Mattsson S. Absorbed dose to technicians due to induced activity in linear accelerators for radiation therapy. Phys Medicine Biol 36: 815–822; 1991. DOI 10.1088/0031‐9155/36/6/009.
Ateia E, Ciraj-Bjelac O, Kovačević M, Belicev P, Cvetković B, Aničin I. Additional dose assessment from activation at high-energy linear accelerator used for radiation therapy. Nucl Technol Radiat Prot 23: 58–64; 2008a.
Ateia E, Ciraj-Bjelac O, Kovačević M, Beličev P, Cvetković B, Aničin I. Additional dose assessment from the activation of high-energy linear accelerators used in radiation therapy. Nucl Technol Radiat Protect 2: 58–64; 2008b.
Bednarz B, Taranenko V, Xu XG. Modeling to characterize neutron activation products in a medical linear accelerator. Med Phys 34: 2469; 2007. DOI 10.1118/1.2760958.
Bewley DK. The physics and radiobiology of fast neutron beams. Bristol: IOP Publishing, Ltd.; 1989.
Blake SW, Finch J, Bonnett DE, Shaw JE. Patient dosimetry and staff protection at the clatterbridge high energy neutron facility. Radiat Protect Dosim 23: 369–372; 1988.
Bodemann R, Lange H-J, Leya I, Michel RR, Schiekel T, Rösel R, Herpers U, Hofmann J, Dittrich B, Suter M, Wölfli W, Holmqvist B, Condé H, Malmborg P. Production of residual nuclei by proton-induced on C, N, O, Mg, Al and S. Nucl Instr Meth 82: 9–31; 1993. DOI 10.1016/0168‐583X(93)95077‐I.
Bonnett DE. Fast neutron therapy at Edinburgh: staff protection. Br J Radiol 56: 665–672; 1983. DOI 10.1259/0007‐1285‐56‐669‐665.
Bonnett DE, Blake SW, Shaw JE, Bewley DK. The Clatterbridge High-energy Neutron Therapy Facility: specification and performance. Br J Radiol 61: 38–46; 1988. DOI 10.1259/0007‐1285‐61‐721‐38.
Bonnett DE, Williams JR, Parnell CJ. The isocentric fast neutron therapy facility at Edinburgh. Br J Radiol 53: 12–20; 1980.
Bragg WH. Studies in radioactivity. London: Macmillan & Co.; 1912.
Clasie B, Wroe A, Kooy H, Depauw N, Flanz J, Paganetti H, Rosenfeld A. Assessment of out-of-field absorbed dose and equivalent dose in proton fields. Med Phys 37: 311–321; 2010. DOI 10.1118/1.3271390.
Culp T. A prediction of accelertaor-produced activation products. Health Phys 92( Suppl. 1): S57–S65; 2007. DOI 10.1097/01.HP.0000253937.78314.e6.
Dodge C. Release of neutron activated materials from broad-spectrum therapeutic cyclotron. Detroit: Wayne State University, School of Medicine; 2004. Thesis.
Eenmaa J, Wootton P, Risler R. Technologist’s exposures and radiation protection aspects of the clinical neutron therapy facility in Seattle. Br J Radiol 60: 310; 1987.
Evdokimoff V, Willins J, Richter H. Induced radioactive potential for a medical accelerator. Health Phys 83: S68–70; 2002. DOI 10.1097/00004032‐200211005‐00007.
Farr JB, Mascia AE, Hsi WC, Allgower CE, Jesseph F, Schreuder AN, Wolanski M, Nichiporov DF, Anferov V. Clinical characterization of a proton beam continuous uniform scanning system with dose layer stacking. Med Phys 35: 4945–4954; 2008. DOI 10.1118/1.2982248.
Faßbender M, Shubin Y, Lunev V, Qaim S. Experimental studies and nuclear model calculations on the formation of radioactive products in interactions of medium energy protons with copper, zinc and brass: estimation of collimator activation in proton therapy facilities. Appl Radiat Isot 48: 1221–1230; 1997. DOI 10.1016/S0969‐8043(97)00102‐4.
Finch J, Bonnett DE. An investigation of the dose equivalent to radiographers from a high-energy neutron therapy facility. Br J Radiol 65: 327–333; 1992. DOI 10.1259/0007‐1285‐65‐772‐327.
Fischer HW, Tabot B, Poppe B. Comparison of activation products and induced dose rates in different high-energy medical linear accelerators. Health Phys 94: 272–278; 2008. DOI 10.1097/01.HP.0000291945.15684.a7.
Fischer HW, Tabot BE, Poppe B. Activation processes in a medical linear accelerator and spatial distribution of activation products. Phys Medicine Biol 51: N461–466; 2006. DOI 10.1088/0031‐9155/51/24/N02.
Gründel M, Güthoff F. Radiation burden of assistant medical technicians at a medical accelerator. Med Phys India 33: 171–174; 2008. DOI 10.4103/0971‐6203.44480.
Guo S, Ziemer PL. Health physics aspects of neutron activated components in a linear accelerator. Health Phys 86: S94–S102; 2004. DOI 10.1097/00004032‐200405002‐00012.
Hecksel D, Anferov V, Fitzek M, Shahnazi K. Influence of beam efficiency through the patient-specific collimator on secondary neutron dose equivalent in double scattering and uniform scanning modes of proton therapy. Med Phys 37: 2910–2917; 2010. DOI 10.1118/1.3431575.
Hoffman RJ, Nath R. On the sources of radiation exposure of technologists in a radiotherapy center with high energy x-ray accelerators. Health Phys 42: 525–526; 1982.
Howell RM, Kry SF, Burgett E, Hertel NE, Followill DS. Secondary neutron spectra from modern Varian, Siemens, and Elekta LINACs with multileaf collimators. Med Phys 36: 4027–4038; 2009. DOI 10.1118/1.3159300.
Jones DTL, Schreuder AN, Symons JE, Binns PJ. Experimental investigation of multiblade trimmer for neutron therapy. Brachyther Internat 1: 59–66; 1997.
Kalef-Ezra JA. Health physics aspects in treatment rooms after 18‐mv x-ray irradiations. Radiat Protect Dosim 147: 281–286; 2011. DOI 10.1093/rpd/ncr361.
Kanai T, Kanematsu N, Minohara S, Komori M, Torikoshi M, Asakura H, Ikeda N, Uno T, Takei Y. Commissioning of a conformal irradiation system for heavy-ion radiotherapy using a layer-stacking method. Med Phys 33: 2989–2997; 2006.
Koehler AM, Schneider RJ, Sisterson JM. Range modulators for protons and heavy ions. Nucl Instrum Meth 131: 437–440; 1975. DOI 10.1016/0029‐554X(75)90430‐9.
Koehler AM, Schneider RJ, Sisterson JM. Flattening of proton dose distributions for large fields. Med Phys 4: 297–301; 1977. DOI 10.1118/1.594317.
Konefal A, Polaczek-Grelik K, Zipper W. Undesirable nuclear reactions and induced radioactivity as a result of the use of the high-energy therapeutic beams generated by medical LINACs. Radiat Protect Dosim 128: 133–145; 2008.
Kouskoulas T, Yudelev M. Activation of metallic prostheses in patients undergoing fast neutron therapy. Med Phys 32: 1891; 2005. DOI 10.1118/1.1997418.
LaRiviere PD. Radiotherapy technologist dose from high-energy electron medical accelerators. Health Phys 49: 1105–1114; 1985.
Maughan RL, Yudelev M, Forman JD, Williams SB, Gries D, Fletcher TM, Chapman W, Blosser EJ, Horste T. Attenuation and activation characteristics of steel and tungsten and the suitability of these materials for use in a fast neutron multileaf collimator. Med Phys 28: 1006–1009; 2001. DOI 10.1118/1.1376135.
McGinley PH, Wright BA, Meding CJ. Dose to radiotherapy technologist from air activation. Med Phys 11: 855–858; 1984. DOI 10.1118/1.595584.
Moskvin V, Cheng CW, Das IJ. Pitfalls of tungsten multileaf collimator in proton beam therapy. Med Phys 38: 6395–6406; 2011. DOI 10.1118/1.3658655.
Moyers MF, Lesyna DA. Exposure from residual radiation after synchrotron shutdown. Radiat Meas 44: 176–181; 2009. DOI 10.1016/j.radmeas.2009.01.009.
Mukherjee B, Lambert J, Hentschel R, Farr J. Explicit estimation of out-of-field neutron and gamma dose equivalents during proton therapy using thermoluminescent-dosimeters. Radiat Meas 46: 1952–1955; 2011.
O’Brien P, Michaels HB, Gillies B, Aldrich JE, Andrew JW. Radiation protection aspects of a new high-energy linear accelerator. Med Phys 12: 101–107; 1985.
Orecchia R, Zurlo A, Loasses A, Krengli M, Tosi G, Zurrida S, Zucali P, Veronesi U. Particle beam therapy (hadrontherapy): basis for interest and clinical experience. European J Cancer 34: 459–468; 1998. DOI 10.1016/S0959‐8049(97)10044‐2.
Pedroni E, Bacher R, Blattmann H, Bohringer T, Coray A, Lomax A, Lin S, Munkel G, Scheib S, Schneider U, Tourovsky H. The 200‐MeV proton therapy project at the Paul Scherrer Institute: conceptual design and practical realization. Med Phys 22: 37–53; 1995. DOI 10.1118/1.597522.
Pelliccioni M. Radiation protection at hadron therapy facilities. Radiat Protect Dosim 146: 407–413; 2011. DOI 10.1093/rpd/ncr242.
Perrin B, Walker A, Mackay R. A model to calculate the induced dose rate around an 18 MV Elekta linear accelerator. Phys Medicine Biol 48: N75–81; 2003. DOI 10.1088/0031-9155/48/5/402.
Petrović N, Krestić-Vesović J, Stojanović D, Ciraj-Bjelac O, Lazarević D, Kovacević M. Contribution of activation products to occupational exposure following treatment using high-energy photons in radiotherapy. Radiat Protect Dosim 143: 109–112; 2011. DOI 10.1093/rpd/ncq290.
Powell NL, Newing A, Bullen MA, Sims C, Leaton SF. A radiation safety survey on a CLINAC‐20 linear accelerator. Phys Med Biol 32: 707–718; 1987. DOI 10.1088/0031‐9155/32/6/004.
Prael RE, Lichtenstein H. User Guide to LCS: The LAHET Code System. LA-UR-89-3 014, Los Alamos National Laboratory, Los Alamos, NM; 1989.
Rawlinson JA, Islam MK, Galbraith DM. Dose to radiation therapists from activation at high-energy accelerators used for conventional and intensity-modulated radiation therapy. Med Phys 29: 598–608; 2002. DOI 10.1118/1.1463063.
Risler R, Wootton P, Ziai F, Jacky J, Emery R, Patrick M, Kalet I. Clinical cyclotron seattle: eleven years of operation. In: Proceedings of the 14th International Conference on Cyclotrons and their Applications - Capetown, South Africa. Singapore: World Scientific Publishing Co.; 1995: 519–522.
Rosenberg I, Awschalom M, Ten Haken RK, Bennett BR. Analysis of personnel exposures in neutron therapy facilities. Health Phys 46: 407–412; 1984. DOI 10.1097/00004032‐198402000‐00016.
Rouse I, Martx M, Glisson C, Siebers J, Elmer T II. Analysis of cerrobend activation produced by a 250‐MeV medical proton accelerator. Health physics of radiation generating machines. College Park, MD: American Institute of Physics; 1997: 425–431.
Schiekel T, Sudbrock F, Herpers U, Gloris M, Lange H-J, Leya I, Michel R, Dittrich-Hannen B, Synal H-A, Suter M, Kubik P, Blann M, Filges D. Nuclide production by proton-induced reactions on elements 6<z<29 in the energy range from 200 MeV to 400 MeV. Nucl Instr Meth B123: 324–329; 1996.
Schneider R, Sisterson J, Koehler A, Klein J, Middleton R. Measurement of cross sections for aluminum‐26 and sodium‐24 induced by protons in aluminum. Nucl Instr Meth B29: 271–274; 1987.
Sisterson J. Selected radiation safety issues at proton therapy facilities. In: 12th Biennial Radiation Protection and Shielding Division Topical Meeting- Santa Fe, NM. LaGrange Park, IL: American Nuclear Society; 2002.
Sisterson J, Kim K, Beverding A, Englert P, Caffee M, Jull A, Donahue D, McHargue L, Castaneda C, Vincent J, Reedy R. Measurement of proton production cross sections of 10 Be and 26 Al from elements found in lunar rocks. Nucl Instr Meth B123: 324–329; 1996.
Smathers JB, Graves RG, Sandel PS, Almond PR, Otte VA, Grant WH. Radiation dose received by TAMVEC neutron therapy staff. Health Phys 35: 271–277; 1987. DOI 10.1097/00004032‐197808000‐00009.
Stranden E. Activity induced in patients by high energy x-ray therapy. Phys Medicine Biol 22: 348–352; 1977. DOI 10.1088/0031‐9155/22/2/014.
Symons JE, Jones TDL. Activation of tooth fillings. Faure, South Africa: iThemba National Laboratory; 1993.
Tatcher M, Rosenberg I, Couch JG. Dose to radiotherapy technologists from activation of patients at a fast neutron therapy facility. Health Phys 53: 311–312; 1987.
Ten Haken RK, Awschalom M, Rosenberg I. Activation of the major constituents of tissue and air by a fast neutron radiation therapy beam. Med Phys 10: 636–641; 1983. DOI 10.1118/1.595326.
Thomson JJ. On rays of positive electricity. Phil Mag 13; 1907.
Vassiliev ON, Titt U, Kry SF, Mohan R, Gillin MT. Radiation safety survey on a flattening filter-free medical accelerator. Radiat Protect Dosim 124: 187–90; 2007.
Vynckier S, Pihet P, Wambersie A. Results of 10 y of radiation protection dosimetry at the neutrontherapy facility in Louvain-la-Neuve. Health Phys 57: 529–535; 1989. DOI 10.1097/00004032‐198910000‐00002.
Wang YZ, Evans MD, Podgorsak EB. Characteristics of induced activity from medical linear accelerators. Med Phys 32: 2899–2910; 2005. DOI 10.1118/1.2001767.
Welsh JS, Mackie TR, Limmer JP. High-energy photons in IMRT: uncertainties and risks for questionable gain. Technol Cancer Research Treatment 6: 147–149; 2007.
Wielopolski L, Meek AG, Moskowitz M, Cohn SH. (γ, n) activation of cancer patients. J Radioanalytical Nucl Chem 114: 187–193; 1987. DOI 10.1007/BF02048890.
Wilson RR. Radiological use of protons. Radiol 47: 487–491; 1947.
Yudelev M, Maughan RL, Dunlap K. Shielding and radiation safety around a superconducting cyclotron neutron therapy facility. Health Phys 69: 130–136; 1995. DOI 10.1097/00004032‐199507000‐00015.
Yudelev M, Maughan RL, Jordan LE, Saxena R. Dose equivalents to neutron therapy facility staff due to induced activation. Health Phys 72: 361–367; 1997. DOI 10.1097/00004032‐199703000‐00003.
Zhu TC, Ahnesjo A, Lam KL, Li XA, Ma CM, Palta JR, Sharpe MB, Thomadsen B, Tailor RC, Group ATPCT. Report of AAPM Therapy Physics Committee Task Group 74: in-air output ratio, sc, for megavoltage photon beams. Med Phys 36: 5261–52691; 2009. DOI 10.1118/1.3227367.
Keywords:© 2014 by the Health Physics Society
accelerators, medical; exposure, occupational; radiation risk; radiation therapy