Australas Phys Eng Sci Med (2011) 34:529–533 DOI 10.1007/s13246-011-0095-z
TECHNICAL PAPER
Lessons learned from a HDR brachytherapy well ionisation chamber calibration error Claire Dempsey
Received: 4 April 2011 / Accepted: 8 August 2011 / Published online: 20 August 2011 Ó Australasian College of Physical Scientists and Engineers in Medicine 2011
Abstract The outcomes of a recent brachytherapy welltype ionization chamber calibration error are given in the hope that other brachytherapy treatment centres may better understand the importance of each entry stated in a well chamber calibration certificate. A Nucletron Source Dosimetry System (SDS) PTW well-type ionization chamber was sent for a biennial calibration in September 2010. Upon calibration of the chamber, it was discovered that the previous calibration (in July 2008) contained a ?2.6% error in the chamber calibration coefficient. Investigation of the information on the 2008 well chamber calibration certificate indicated the source of the error, which could or should have been detected by both the calibration laboratory and/or the radiation therapy department upon return of the chamber. Consideration must be given to all values and conditions given on the calibration certificate when accepting a ionization chamber back from a calibration laboratory. The issue of whether the source strength from the source calibration certificate or the measured source strength from the calibrated ionization chamber should be entered into the treatment unit is also raised. Keywords
Brachytherapy Chamber Calibration
C. Dempsey (&) Department of Radiation Oncology, Calvary Mater Newcastle Hospital, Locked Bag 7, Hunter Regional Mail Centre, Newcastle, NSW 2310, Australia e-mail:
[email protected] C. Dempsey School of Health Sciences, University of Newcastle, Newcastle, NSW, Australia
Introduction High dose rate (HDR) brachytherapy treatments are commonly undertaken using an Iridium-192 (192Ir) source with an initial activity of between 370 and 518 GBq (10–14 Ci). These sources are replaced on average every 3 months, with source decay extending treatment times by approximately 60% by the end of 3 months. Each Nucletron 192Ir HDR source comes with a calibration certificate issued from Mallinckrodt Medical (Netherlands) on behalf of Nucletron, with a similar certificate issued for Varian 192Ir HDR sources. This certificate states the material, dimensions, serial number and activity of the source to which it relates. Whilst the source activity is measured by the manufacturer using a calibrated ionization chamber, it is best practice and recommended by several publications [1–6] to confirm the source activity. Source strength verification checks are generally conducted using one of two methods; an in-air checking ‘jig’ or a well-type ionization chamber [7]. Each method requires a calibrated ionization chamber to convert the measured reading into a reference air kerma strength, SK, (cGy m2 h-1) or reference air kerma rate (RAKR), KR (Gy s-1 at 1 m). This calibration should be undertaken by an internationally accredited calibration laboratory. In Australia, well-type ionisation chamber calibration for 192Ir is not available, so radiation therapy departments must send secondary standard chambers overseas for calibration. The problem in deriving an air kerma calibration factor of an ionization chamber for use with HDR 192Ir is not as straight-forward as calibration of an ionization chamber for use in external beam irradiation conditions. The most important part of the emitted photon spectrum for an 192Ir HDR source lies in the gap between the standards for x- and
123
530
gamma rays (137Cs and 60Co) established at primary standards laboratories [8]. To overcome this problem, several intercomparison techniques have been developed [9–12] which rely on a predictable response of the ionization chamber between standard energies surrounding the 192Ir spectrum. In 2004, the National Physical Laboratory (NPL) developed a spherical graphite-walled cavity ionisation chamber which was established as the United Kingdom (UK) national primary standard for 192Ir [13]. This chamber has been compared against the standard interpolation techniques in other calibration laboratories, including the Australian Radiation Protection and Nuclear Safety Agency (ARPANSA) [14] and the French Laboratorie National Henri Becquerel (LNHB) [15]. The primary standard can be used for calibration of an 192Ir source, which in turn can be used to calibrate a secondary standard ionisation chamber (either a thimble chamber or well-type chamber). This reduces the requirement for multiple interpolation measurements and associated uncertainty. However, this primary standard for 192Ir is only available for NPL ionization chamber calibrations. All other accredited dosimetry calibration laboratories (ADCL) still rely on an interpolation method of ionization chamber calibration. On the back of the development of the NPL air kerma standard, the Institute of Physics and Engineering in Medicine (IPEM) developed a new code of practice for measurement of the reference air kerma rate for HDR 192Ir brachytherapy sources [16]. The IPEM made several recommendations regarding verification of the manufacturer’s source strength certificate for 192Ir sources. The IPEM recommends the continuing use of the reference air kerma rate (KR) as a means of specifying source strength, as favoured by the International Commission on Radiological Units (ICRU) [17, 18]. The IPEM code of practice acknowledge the comprehensiveness of the American Association of Physicists in Medicine (AAPM) task group 43 (TG43) protocol for brachytherapy treatment planning, which is used in commercial treatment planning systems. The AAPM TG43 uses the quantity air kerma strength (SK), however the IPEM (and NPL) still specify the 192Ir strength as KR. Whilst correction of this requires a simple unit conversion, it would be more appropriate to determine the quality air kerma strength directly from measurements. The code of practice was drafted to help align practices in the UK for HDR 192Ir source measurements. The IPEM specifically recommended the use of well chambers for source calibration/verification, stating it is ‘‘an easily implementable method with the best achievable accuracy’’. There is no current Australasian code or protocol that specifies a method or recommends a particular published
123
Australas Phys Eng Sci Med (2011) 34:529–533
guideline to follow. The previous Australasian College of Physical Scientists and Engineers in Medicine (ACPSEM) position paper was written in 1997, before HDR brachytherapy became well-established [6]. A survey of Australian radiotherapy centres in 2009 indicated that 18 centres used a well-ionisation chamber for source output verification, with three centres using an ionization chamber in an in-air jig. Two centres reported using a well-ionisation chamber as a primary measurement and an in-air jig for a secondary check [19]. There were large variations in the protocol followed by departments, with half of respondents using IAEA TecDoc 1274 [8] protocol to determine source output with some using AAPM TG 43 [3] or associated AAPM documents and others using in-house developed procedures.
Materials In 2006 the Newcastle Calvary Mater Hospital began treating patients with HDR brachytherapy. Acceptance and commissioning of the unit and all of its components were completed at this time. The source strength verification system chosen was a Nucletron branded well-chamber. The Nucletron Source Dosimetry System (SDS) consists of a PTW (Physikalisch-Technische Werkstatten) well-type ionization chamber (Model No. 077.092). This type of ionization chamber comprises a cylindrical air-vented active volume of about 200 cm3 with a PMMA tube insert, a collecting electrode, an external electrode and a guard electrode. The 192Ir source is driven into the PMMA tube insert with a high voltage (?300 V) connected to the ionization chamber via an electrometer. The radiation of the source ionizes the air in the chamber, producing a detectable current related to the source strength. The chamber calibration certificate that came with the well-ionisation chamber was measured by PTW Freiburg. The chamber was calibrated with a standard Nucletron microSelectron HDR192Ir source on a comparison measurement using a well-type chamber (No. 33002) interpolative standard traceable to NIST (National Institute of Standards and Technology) for air kerma. The uncertainty of this measurement was stated at ±2%. The certificate was dated 17th February 2005 with current recommendations for a chamber calibration to be performed every 2 years. In June 2008, the well-ionisation chamber was sent away for another calibration. It was decided that a truly ‘independent’ calibration would be best, so the chamber was sent to a different ADCL instead of to PTW Freiburg, which is associated with the chamber manufacturer. This time, the chamber was calibrated using a standard Varian 192 Ir source (manufactured by Alpha-Omega) on a comparison measurement using a well-type chamber (No.
Australas Phys Eng Sci Med (2011) 34:529–533
33002) interpolative standard traceable to NIST interpolative standard for air kerma. The uncertainty of this measurement was stated at ±2.6%. In September 2010, the well-type ionization chamber was sent for another biennial calibration to the same ADCL that was used in 2008. Upon calibration of the chamber, the laboratory discovered that the previous calibration (dated July 2008) contained a ?2.6% error in the chamber calibration coefficient. Investigation of the information on the 2008 well-chamber calibration certificate indicated the source of the error, which could or should have been detected by both the calibration laboratory and/or the radiation therapy department upon its’ return.
Discussion When the well-chamber was returned to the department in 2008, the difference between the original calibration factor (from PTW Freiburg) and the new calibration factor (from another ADCL) was noticed. The original calibration factor (9.355 9 107 cGy m2 h-1 A-1) was 3.73% lower than the new calibration factor (9.704 9 107 cGy m2 h-1 A-1). There is a different mindset when searching for the cause of something which is most likely an error, and something which could be an error, or could be a normal variation for this type of chamber with a change in calibration centre. It might be feasible (based on the calibration uncertainty of each calibration centre), that there could be a discrepancy between the calibrations of this extent. The apparent shift in calibration factor value fell within the reported accuracy range, if both uncertainties were added together. The standard calibration conditions differed to the original (and Australian standard conditions) with a normalisation temperature of 22°C instead of 20°C. This difference was easily accounted for in the temperature/ pressure correction factor equation. Whilst the original certificate stated that the chamber is vented to atmospheric communication (i.e. open), it was not noticed that the new certificate stated the chamber was sealed to atmospheric communication (hence no air density correction factor was applied). If this had been detected at the time the new certificate was returned to the department then the error could have been easily corrected at that point. However, it was not standard practice in the department to check the status of the atmospheric communication on a calibration certificate. If the calibration factor had the air density correction applied to it, the calibration factor should have been 9.454 9 105 Gy m2 h-1 A-1, a difference of 1.06% from the original PTW Freiburg calibration. Prior to realising the calibration error, both the old and new calibration factors were used to calculate the source
531
strength until March 2010 when it had been established that the differences in calibration factors still generated source strength differences (from the source calibration report) well within the ±3% tolerance. Applying the corrected calibration factor to all source strength calculations from September 2008 to 2010 yielded a maximum difference of 2.19% from the source calibration certificate. Before the 2008 well-chamber calibration, the measured source strength was used as the definitive source strength for both the treatment planning computer and treatment console. With the discrepancy between the well-chamber calibration values, it was decided that the source calibration certificate should be the reference source strength in the treatment planning computer and treatment console. This is realistically a more appropriate value as it can be traced directly back to the source manufacturer and does not rely on referencing an in-house determined value. Due to the shift in protocol (to use the source certificate value), the error in the 2008 well-chamber calibration has not had a clinical impact on any patient treatment. The ADCL was swift in their notification to us when the error was detected. Corrective action protocols required that they evaluate the calibrator response, and immediately determine if this human error occurred systematically in other calibrations. After comprehensive review of database tables no other instance of this error was found in their calibration history. The ADCL was well aware the chamber is open to atmospheric communication, and the mistake in the 2008 calibration was purely accidental, not intentional or the result of ignorance or incompetence. An evaluation of preventative measures is currently underway, including discussions with all ADCLs to increase vigilance in looking for these kind of subtle errors. In addition, the ADCL software database manager is investigating the feasibility of hard-coding certain known sealed chambers to remove the subjective ability of the calibrator to over-ride the default open to atmospheric condition in the software calibration code. It is unlikely that this type of error would be caught by an advisory committee member, who counter-signs the calibration certificate, as the calibration coefficient was within the expect tolerance for this chamber model, and this signature stage does not review the raw data generated during calibration. The status of the ionisation chamber atmospheric communication appears to be a condition that is over-looked by many medical physicists. As well as the operator and advisory committee member at the ADCL, more than three physicists at the Newcastle Calvary Mater Hospital were involved in trying to determine the source of the discrepancy between the original and 2008 well-chamber calibration and none of these discovered the problem. Various details of the differences between the PTW Freiburg
123
532
calibration and the ADCL calibration were studied, and a secondary check of the well-chamber (by comparison with a thimble ionisation chamber in air) was also attempted. During this time, a study was also completed into the dependence of the well-chamber calibration factor on the activity of the source used for the calibration, particularly for the PTW well ionization chamber [20]. Higher calibration factors for the same source type but different source strengths were measured. Although the results of this indicated a difference between HDR and PDR source activities of approximately 1%, it should be noted that the calibration of the well chamber at PTW Freiburg was completed using a source activity of 5.0 Ci and the calibration at the ADCL was completed using a source activity of 6.0 Ci. The Newcastle Calvary Mater Hospital uses the well-chamber to measure source strengths of a new 192Ir source in the order of 10–12 Ci (double that of the calibration source). The recommendation is now that the well ionization chamber calibration occurs at a source activity as close as possible to the source activity that is used in the department. The use of a Varian (Alpha-Omega) 192Ir HDR source for calibration of a well-chamber designated to be used for Nucletron microSelectron 192Ir HDR source verification is another issue that must be addressed. The Varian and nucletron 192Ir HDR sources have different physical geometries (of both the source itself and the encapsulation) which could result in differing calibration coefficients for the same well-chamber. A literature search revealed little information on the effect a different 192Ir source may have on the calibration coefficient. Stump et al. [10] in 2002 performed comparative measurements on the Varian VariSource (VS2000) and the Nucletron microSelectron source and found that at distances greater than 1 cm the anisotropy was similar, deeming it acceptable to use either source for a well-chamber calibration. Ideally, using a ADCL that provides a calibration service with a 192 Ir source from the same manufacturer as used in the clinical setting is recommended. The issue of whether the source strength from the source calibration certificate or the measured source strength from the calibrated ionization chamber should be entered into the treatment planning system and treatment console is also raised. The 2009 survey of Australasian radiation therapy centres indicated that nine centres used in-house measurements as the source strength value for treatment planning while ten centres used the value quoted on the manufacturer’s certificate. Different centres will have varying rationales for applying the manufacturer or measured source strength in the planning system. Some manufacturers are now questioning the validity of adjusting the source strength in the treatment planning system to match the treatment unit itself. The source activity in the treatment planning computer is an
123
Australas Phys Eng Sci Med (2011) 34:529–533
‘associated source strength’, a value used purely for attachment of dwell times to a treatment prescription. Once the treatment plan is sent to the treatment console the ‘associated source strength’ and ‘actual source strength’ (determined by the actual treatment time on the console) are compared and the dwell times modified. Whilst a single, non-decaying source strength value in a treatment planning system may negate the need to check the accuracy and maintain currency of the source strength, it does increase the complexity of any pre-treatment quality assurance on the treatment plan. This would point toward the critical step in defining the correct source strength lies in the value used in the treatment unit itself as this is ultimately the value used in patient treatments. The associated uncertainty with a source certificate calibration is given at ±5% (for a Nucletron 192Ir source) and the uncertainty in the well-chamber calibration is ±3%. The well-chamber uncertainty may be smaller than the source certificate but if an error was to occur with the departmental ionisation chamber calibration, the extent of the error would be far greater than an error on an individual source calibration. This is purely due to the length of time that an ionisation chamber is used in a department before re-calibration, with ionisation chamber calibrations recommended to be performed every 2 years. A single 192Ir source is used clinically for 3 months before replacement, thus any manufacturers source certificate errors would only affect patients in that 3 month period.
Conclusion Consideration must be given to all values and conditions given on the calibration certificate when accepting an ionization chamber back from a calibration laboratory. For a department to ensure that errors like this do not occur, it is suggested to keep a database of all secondary standard calibrations, with the calibration conditions used and the ‘ideal’ calibration conditions required. Each time a secondary standard chamber (or electrometer) is sent for re-calibration, the new values should be checked and validated against previous and ideal values. Any discrepancies (over 1.5%) must be immediately followed up with the standards laboratory that completed the calibration. An Australasian standard protocol should be developed, that indicates the ideal conditions for source strength measurement (ionisation chamber type with desired calibration conditions) and gives firm recommendations for the origin of the source strength in the treatment unit. It is the authors’ opinion that the source strength from the source certificate should be used in the treatment planning computer and treatment console, with any measurements completed ‘‘in-house’’ serving the purpose of ‘‘verification’’.
Australas Phys Eng Sci Med (2011) 34:529–533
References 1. Kutcher GJ, Coia L, Gillin M, Hanson WF, Leibel S, Morton RJ, Palta JR, Purdy JA, Reinstein LE, Svensson GK, Weller M, Wingfield L (1994) Comprehensive QA for radiation oncology: report of the AAPM radiation therapy committee task group 40. Med Phys 21(4):581–618 2. Glasgow GP, Bourland JD, Grigsby PW, Meli JA, Weaver KA (1993) Remote afterloading technology: report of the AAPM task group 41. Published for the American Association of Physicists in Medicine by the American Institute of Physics, New York 3. Nath R, Anderson LL, Luxton G, Weaver KA, Williamson JF, Meigooni AS (1995) Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM radiation therapy committee task group no 43. Med Phys 22:209–234 4. Baltas D, Sakelliou L, Zamboglou N (2007) The physics of modern brachytherapy for oncology, series in medical physics and biomedical engineering. CRC Press, Taylor and Francis Group, p 585, IBSN-13: 978-0-7503-0708-6 5. Nath R, Anderson LL, Meli JA, Olch AJ, Stitt JA, Williamson JF (1997) Code of practice for brachytherapy physics: report of the AAPM radiation therapy committee task group no 56. Med Phys 24(10):1557–1598 6. ACPSEM Position Paper (1997) Recommendations for the safe use of external beams and sealed brachytherapy sources in radiation therapy. Australasian College of Physical Scientists and Engineers in Medicine, Melbourne 7. Nair MTK, Cheng M-C (1997) HDR Source calibration methods and discrepancies. Int J Radiat Oncol Biol Phys 38(1):207–211 8. IAEA (2002) Calibration of photon and beta ray sources used in brachytherapy, International Atomic Energy Agency, IAEATECDOC-1274, Vienna 9. Goetsch SJ, Attix FH, Pearson DW, Thomadsen BR (1991) Calibration of 192Ir high-dose-rate afterloading systems. Med Phys 18(3):462–467 10. Stump KE, DeWerd LA, Micka JA, Anderson DR (2002) Calibration of new high dose rate 192Ir sources. Med Phys 29(7): 1483–1488
533 11. Bu¨ermann L, Kramer HM, Schrader H, Selbach HJ (1994) Activity determination of 192Ir solid sources by ionization chamber measurements using calculated corrections for selfabsorption. Nucl Instrum Methods Phys Res A 339:369–376 12. Loftus TP (1980) Standardization of iridium-192 gamma-ray sources in terms of exposure. J Res Natl Bur Stand 85:19–25 13. Sander T, Nutbrown RF (2006) The NPL air kerma primary standard TH100C for high dose rate 192Ir brachytherapy sources, NPL Report DQL-RD 004, http://publications.npl.co.uk 14. Butler D, Haworth A, Sander T, Todd S (2008) Comparison of 192 Ir air kerma calibration coefficients derived at ARPANSA using the interpolation method and the national physical laboratory using a direct measurement, Australas. Phys Eng Sci Med 31(4):332–338 15. Douysset G, Sander T, Gouriou J, Nutbrown R (2008) Comparison of air kerma standards of LNE-LNHB and NPL for 192Ir HDR brachytherapy sources: EUROMET project no 814. Phys Med Biol 53:N85–N97 16. Bidmead AM, Sander T, Locks SM, Lee CD, Aird EGA, Nutbrown RF, Flynn A (2010) The IPEM code of practice for determination of the reference air kerma rate for HDR 192Ir brachytherapy sources based on the NPL air kerma standard. Phys Med Biol 55:3145–3159 17. ICRU (1985) Dose and volume specification for reporting intracavitary therapy in gynaecology, ICRU Report No 38, Washington, DC 18. ICRU (1997) Dose and volume specification for reporting interstitial therapy, ICRU Report No 58, Bethesda, MD 19. Brown RL, Butler DJ (2010) The 2009 survey of therapy equipment and dosimetry practices in Australian radiotherapy centres, Australas. Phys Eng Sci Med 33(4):285–297 20. Douysset G, Ostrowsky A, Delaunay F (2008) Some unexpected behaviours of PTW/Nucletron well-type ionization chambers. Phys Med Biol 53:269–275
123
