Use of Bubble Detectors to Characterize Neutron Dose Distribution in a Radiotherapy Treatment Room used for IMRT treatments

Transcription

1 Use of Bubble Detectors to Characterize Neutron Dose Distribution in a Radiotherapy Treatment Room used for IMRT treatments Alana Hudson *1 1 Tom Baker Cancer Centre, Department of Medical Physics, Street NW, Calgary, Alberta, Canada Abstract. Neutrons are an undesirable by-product of megavoltage linear accelerator (linac) units operating at or above 10 MV. This work characterizes the distribution of neutrons produced by a medical radiotherapy treatment unit delivering 10 MV photons. Neutron levels were measured at various locations in a linac treatment room for a newly installed Varian Clinac ix (Varian Medical Systems, Palo Alto, USA) with 6 and 10 MV capabilities. Measurements were made using fast neutron dosimeters (BD-PND) calibrated to Am-Be with sensitivities of ~ 0.07 bubbles/µsv and ~ 1.8 bubbles/µsv (Bubble Technology Industries (BTI), Chalk River, Canada). Because of the potential advantages of using 10 MV for intensity modulated radiotherapy (IMRT) treatments of the prostate and other tumours at depths greater than 10 cm, neutron contamination contributing to the patient dose is very relevant. The results of this work, therefore, have direct clinical implications and neutron contribution to patient dose and neutron shielding were investigated. Patient neutron doses for a 10 MV IMRT fraction ranged between 0.06 and 0.6% of the prescribed x- ray dose at isocentre for closed and s, respectively. At 1 m from isocentre, the dose in the patient plane was found to be 0.036% of the prescribed x-ray dose. Neutron survey results gave neutron doses of about 1.5 µsv per 200 cgy IMRT fraction at the entrance door to the vault and a projected annual neutron dose of 6 msv. Use of 10 MV beams for IMRT treatments is not expected to have significant adverse effect on either the patient or staff doses. KEYWORDS: Neutron contamination, linear accelerator, neutron shielding, neutron survey, bubble detectors * Presenting author, alanahud@cancerboard.ab.ca 1

2 1. Introduction Since the late 1990 s, IMRT has been gaining popularity as the most desirable treatment technique for radiotherapy of certain treatment sites. Due to the increase in complexity of the treatment delivery, and therefore the increase in MU per unit dose required for IMRT versus standard radiotherapy, the total MU workload of the linear accelerator is also increased. While the majority of IMRT treatments use 6 MV photon energy beams, there is interest in higher energy beams due to the increased penetration required for deeper-seated tumours such as prostate tumours. This work examines the workload and radiation safety considerations of moving to IMRT treatments with beam energies of 10 MV or higher as well as the increase in patient dose due to neutrons. 1.1 Increased MU workload for IMRT treatment delivery From our experience at the Tom Baker Cancer Centre, the typical number of monitor units (MUs) required to deliver 200 cgy in an IMRT treatment fraction is on the order of 1200 MU for head and neck patients and 800 MU for prostate patients. This is about 5 times more MUs than if the dose was delivered as a conventional radiotherapy treatment which nominally requires 1 MU per cgy prescribed dose. This increase in monitor units has implications on neutron production and x-ray leakage from the head of the linac. This paper addresses both the increase in patient dose and shielding considerations due to this increase in neutron production and x-ray leakage. 1.2 Neutron bubble detectors Neutron bubble detectors are passive detectors that contain gel with super-heated droplets. When neutrons interact with the droplets, secondary charged particles are produced. When these charged particles deposit their energy in the droplet, they can cause it to be vaporized, i.e. a visible bubble is produced which remains fixed in the polymer and can be counted [1]. Bubble count is proportional to neutron dose through calibration. Neutron measurements were made using BN-PND fast neutron detectors (Bubble Technology Industries (BTI), Chalk River, Canada). Characteristics of these dosimeters have been described elsewhere [1,2,3]. The BN-PND detector calibration factors were determined by the manufacturer using an AmBe source and converting neutron fluence to equivalent dose according to NCRP As a part of this conversion, a radiation weighting factor of 10 was used to account for biological damage that would be induced by the average energy 4.15 MeV AmBe neutrons [4]. Neutrons created in a MV linac photon beam are on the order of 1-2 MeV average energy [1]. For this energy range, the radiation weighting factor has a maximum value of 20 [5]. To account for this difference in radiation weighting factors, the results presented here have been multiplied by 2 to give a more conservative estimate of neutron dose from a radiation safety perspective. 1.3 Neutron measurements At least 3 measurements were made for each setup and location reported in this paper. The bubbles in each detector were read by eye 3-4 times each, rotating the angle of view between readings. 2. Measurements and Results 2.1 Photon interference of neutron measurements in the primary beam BN_PND detectors have been found to over-respond in the primary beam [1]; however, measurements were taken at isocentre in the primary beam to give an upper limit to patient dose due to neutrons. Additionally, several measurements were done in lower energy beams to try to account for the overresponse due to sensitivity to the primary x-ray beam. No bubbles were produced when detectors were irradiated at isocentre in a Co-60 beam, 15 x 15 cm 2 field size, for 10 min. Conversely, approximately 30 bubbles were formed when the detectors were placed in a 6 MV, 15 x 15 cm 2 beam at isocentre and 1000 MU were delivered. This translated to a neutron dose rate of approximately 0.4 µsv/mu of x-rays. Clearly the neutron detectors do have some sensitivity to x-rays although it is small. For the same setup but using a 10 MV x-ray beam, a neutron dose rate of approximately 7 µsv/mu of x-rays was found. If it is assumed that difference in response to x-rays and electrons in the primary beam is the same for 6 MV and 10 MV beams, then the excess dose measured in the 10 MV beam, ~ 6.6 µsv, can be attributed to neutrons while the remainder, ~0.4 µsv, is attributed to 2

3 photons. The neutron component of measured dose in the primary beam, therefore, is on the order of 94%. A complimentary set of measurements were done outside the primary beam at two locations: at isocentre by closing the jaws and MLCs completely offset from the midline and with an 15 cm from isocentre in the toe direction in the patient plane. In these cases, the neutron dose component was found to be around 99% of the measured dose for a 10 MV beam. 2.2 Open vs. closed jaws Measurements were done for both s and fully closed jaws and multileaf collimators (MLCs). It is expected that the neutron dose from an IMRT treatment field fall somewhere between these values as it consists of a field where MLCs are constantly changing. It was found that neutron dose to the patient plane was higher for the versus the closed field which is consistent with the finding of Bourgois et al. where neutron contribution to isocentre increased with field size [2]. 2.3 Patient neutron doses Patient neutron doses were measured at several locations along the treatment couch. All measurements were done with the gantry and collimators at 0. At isocentre, patient neutron doses were found to range between (1.1 ± 0.3) µsv/mu of x-rays when jaws and MLCs were completely closed and offset from the central axis and (11 ± 4) µsv/mu of x-rays for an open 15x15 cm 2 field (Table 1). The neutron doses measured here, in particular those made in a closed field, are consistent with previous measurements [6,7]. In the patient plane away from isocentre towards the feet of the patient, neutron doses were again measured for open and closed fields. The s gave higher neutron doses and are reported here since they give a higher, and therefore more conservative, estimate of patient neutron dose. At 15 cm from isocentre, average neutron dose was found to be (2.8 ± 0.6) µsv/mu of x-rays; at 30 cm it was (2.1 ± 0.5) µsv/mu and at 1 m it was (0.7 ± 0.1) µsv/mu. These doses translate to approximately ,000 µsv per 200 cgy of IMRT x-ray dose delivered or between and 0.6% of the 200 cgy prescription dose. IMRT neutron doses were calculated assuming 500 MU of x-rays are delivered per Gy of prescribed treatment dose. Table 1. Summary of average measured patient neutron doses Position Field setup Measured dose in µsv/mu Approx. dose per 200 cgy conventional fraction (µsv/200 cgy) Isocentre 15x15 cm 2 Isocentre Closed jaws and MLC (offset from isocentre) 15 cm from 15x15 cm 2 isocentre (toe 30 cm from isocentre (toe 100 cm from isocentre (toe 15x15 cm 2 15x15 cm 2 Approx. dose per 200 cgy IMRT fraction (µsv/200 cgy) Percentage of neutron dose to x-ray prescription dose for IMRT (%) 11 ± ± ,000 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

4 2.4 Neutron measurements in the treatment room Treatment head Neutron contamination at the head of the linac for a 10 MV beam was found to be (3.3 ± 0.7) µsv/mu of x-rays Maze and shielding measurements Measurements were made in the maze and outside the treatment room using a worst-case scenario radiation survey approach. The gantry was rotated to 90 such that the head of the machine was closest to the maze. Measurements were done with both jaws and MLCs closed to get maximum neutron production and 5000 MU were delivered. Neutron contamination was measured approximately halfway down the vault maze. Here neutron doses were found to be (0.025 ± 0.005) µsv/mu of x-rays. In this particular vault, the maze is quite short and does not contain excess bends. Nonetheless, the neutron dose falls substantially as you approach the vault door. About 50 cm from the door on the inside of the vault, the neutron dose was measured to be ( ± ) µsv/mu. On the outside of the wooden vault door, the neutron dose was a maximum of ( ± ) µsv/mu. No neutrons were detected in the cabling conduits that go from inside the treatment room to the console area or in the console area where staff normally reside. Based on the measurements made on the outside of the door, the maximum neutron dose in the console area would increase from about 0.3 µsv for a conventional radiotherapy fraction to about 1.5 µsv for an IMRT fraction. The projected annual neutron dose assuming half the treatments are delivered as IMRT would be (6 ± 1) msv/year. 3. Discussion The original shielding calculations for one of the treatment vaults housing a 6/10 MV dual energy linac were revisited and calculations were redone assuming an IMRT factor of 3 for 10 MV, i.e. 5 times more MU for IMRT versus conventional radiotherapy with half the 10 MV dose being delivered as IMRT. The head leakage calculations were found to increase between 170% and 300% for this change in treatment delivery. With existing vault shielding, this led to a negligible increase in predicted annual dose rate for primary barriers and approximately 40% increase in secondary barrier dose rates. The dose rate at the secondary barrier with the highest annual dose became 0.52 msv/year versus 0.39 msv/year if 10 MV was only delivered conventionally. This increase in annual dose causes the barrier to slightly exceed the Nuclear Energy Worker design goal of 0.5 msv/year; however, the design limit is quite conservative and was calculated using a number of conservative assumptions; therefore, worker doses are still expected to be well under the ALARA dose recommendation of 1 msv/year and far below the legal dose limits of 50 msv/year in any one year and 20 msv/year averaged over 5 years. No neutrons were measured at the conduits of the bunker and only µsv/mu of x-rays was measured at the entrance door; therefore, increased neutron production is unlikely to be a shielding concern; however, in the interest of ALARA, a borated polyethylene neutron door could be considered as a replacement to the current wooden door at the entrance to the treatment vault. A similar set of calculations was done for an existing vault housing a 6/15 MV linac. In the case of using IMRT to deliver half the 15 MV dose at isocentre, the increase in annual dose to secondary barriers would be almost 200%. This would lead to more problems meeting ALARA dose constraints and possibly even legal limits if existing shielding was not added. Neutron doses outside the bunker were not measurable so are not a concern in this case. Neutron dose to the patient for a 10 MV IMRT treatment was found to be on the order of 0.6%. Given that the tolerance in x-ray dose output delivered to the patient is 2% and the patient will be receiving significant dose beyond background due to radiotherapy, the increased dose due to neutrons in a 10 MV beam can be considered negligible. It should also be noted that the highest neutron doses are at isocentre which is normally in the prescribed tumour volume. With distance from isocentre, the neutron contribution to dose falls off sharply. In contrast to the 10 MV treatment beam, measurements made in an 18 MV beam gave patient doses up to (700 ± 100) µsv/mu of x-rays which, for an IMRT treatment of 1000 MU per 200 cgy x-ray prescribed, would be about a 35% increase in dose to the patient which is significant and undesirable. 4

5 4. Conclusions The patient neutron contamination dose of an IMRT 10 MV patient treatment was found to be insignificant compared to the prescribed treatment dose. The neutron contribution is not expected to have any adverse effects on the patient for this energy of x-ray beam and therefore, 10 MV is an acceptable choice of treatment beam for IMRT treatments of deeper-seated tumours. Due to the conservative nature of shielding design in Canada, it is expected that shielding would be adequate for a change in practice to include 10 MV IMRT treatments. In contrast to the 10 MV beam, at higher energies such as 18 MV, the shielding may require closer inspection and additional shielding material may need to be added to secondary barriers to account for the extra leakage. Furthermore, neutron dose to the patient also becomes a concern at higher energies and more work would be required to investigate if it is viable for clinical use. Acknowledgements The author wishes to thank Jodi Ploquin, Eduardo Villarreal-Barajas, and Derek Brown for their contributions to this work, and Stephen Lawrence for his support. REFERENCES [1] NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS, Structural Shielding Design and Evaluation for Megavoltage X- and Gamma-Ray Radiotherapy Facilities, Report No. 151, NCRP, Bethesda, MD (2005). [2] BOURGOIS, L., et al., Use of bubble detectors to measure neutron contamination of a medical accelerator photon beam, Radiation Protection Dosimetry 74 (1997) [3] D ERRICO, Francesco, Radiation dosimetry and spectrometry with superheated emulsions, Nuclear Instruments and Methods in Physics Research B 184 (2001) [4] NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS, Protection Against Neutron Radiation, Report No. 38, NCRP, Washington, DC (1971). [5] INTERNATIONAL COMMISSION ON RADIOLOGIAL PROTECTION, Recommendations of the International Commission on Radiological Protection, Publication 60, ICRP, Amsterdam (1990) [6] MCGINLEY, Patton, Shielding Techniques for Radiation Oncology Facilities, 2 nd Ed, Medical Physics Publishing, Madison, WI (2002). [7] MCCALL, Richard, et al., Room scattered neutrons, Med Phys 26 (1999)