| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Correspondence |
Regional Medical Physics Department, South Cleveland Hospital , Middlesbrough TS7 8RZ, UK
(The Editors do not hold themselves responsible for opinions expressed by correspondents)
The Editor—Sir,
I read with great interest the recent publication by Vollans and Wilkinson [1]. I have adopted a similar method of measurement of calibration of iridium-192 sources, with three major differences. First, the sources are used for high dose rate brachytherapy. Second, I have measured dose rate in water. Third, the assessment is against test reports that quote activity of each source. The results of calibration have been compared with test reports from the manufacturer over a 4-year period, including 18 source exchanges. The mean agreement was 100.7%. Results suggest test reports are acceptable for clinical use.
All sources were used in high dose rate brachytherapy based on Varisource, a remote afterloading treatment unit supplied by Varian (Varian Medical Systems, Crawley, UK). Varisource and its associated treatment planning system, Cadplan BT, were delivered in 1996. A method of measuring dose rates in water was developed to verify dose rates under clinical conditions and thus to validate the source activity and the planning software. A polymethylmethacrylate gauge was designed and constructed. Interstitial needles were spaced 120° around circles of 20 mm and 50 mm, with the centre on the axis of a Farmer chamber. The gauge was immersed in water at room temperature. Corrections for attenuation in the sleeve and needle were determined experimentally. A series of measurements was made to establish the maximum response of the chamber as the position of the source in the needle was varied.
An attempt was made to follow national guidelines [2], which call for measurement of air kerma rate (AKR) in air using a gauge. However, the choice of a factor relating AKR to nominal activity compromised this method: if the factor recommended by the source supplier is accepted, does the measurement validate the nominal activity and what is the relationship to absorbed dose rate in the treatment condition?
The ratio of measured dose rate in water to that calculated from the activity quoted in the test report was assessed. The maximum was 102.2%, the minimum was 97.2% and the mean was 100.7%. This implies a small systematic discrepancy with random errors up to 3.5%. As expected, the random errors are larger than those associated with external beam therapy. For these short treatment distances, neither chamber nor source size can be neglected. Non-uniformity of linear source activity contributes small errors. Autoradiographs of the initial sources did not reveal large non-uniformities. Typical irradiation times of 2–5 min were not affected by the electrometer leakage rate, which was negligible.
As part of the routine quality assurance programme, dose rate in water was measured each month. The results were as expected, allowing for radioactive decay. The evidence suggests consistency between measured activity derived from dose rates in water and activity in the manufacturer's test report. Since this method was introduced in 1996, 18 sources have been calibrated, nominally one source per calendar quarter. 177 high dose rate treatments have been performed.
The method outlined by Vollans and Wilkinson [1] has been extended. A similar conclusion has been reached that the manufacturer's test report should be used for clinical dosimetry.
Received for publication August 29, 2000. Accepted for publication September 25, 2000.
References
North Western Medical Physics, Christie Hospital, Manchester M20 4BX, UK
The Editor—Sir,
The letter by Dr Plane raises two concerns that call for further comment. First, one must distinguish between measurements that are done to calibrate a source, that is to determine the magnitude of the source specification quantity, and measurements that are done to verify the outcome of a dose computation. For many years now there have been strong recommendations that brachytherapy sources should be specified in terms of an emission quantity [1–4], and the joint BIR/IPSM Working Party [5] endorsed the use of reference air kerma rate (RAKR) for this purpose. It is, therefore, somewhat disappointing to find that the source supplier for the Varisource machine continues to use a content quantity, namely activity. It is also disappointing to find that some treatment planning systems, including it would seem Cadplan BT, require sources to be specified in terms of activity. There are two arguments against using activity for source specification. First, there is potential for confusion between the true activity and the apparent (or equivalent) activity, where the latter is the activity of a hypothetical unfiltered point source of the radionuclide that would give the same emission rate at some reference point. Ironically, both the true activity and the apparent activity of a brachytherapy source will be determined by working backwards from an emission quantity measurement, but derivation of the former would involve making an additional correction for the source size and the net effect of attenuation and scattering in the sheathing material. The second argument against using activity is that there seems to be no universally accepted value for the iridium-192 air kerma rate constant. For example, Dutreix et al [6] quote 0.1157 µGy h-1 MBq-1 m2, whereas Godden [7] quotes 0.111 µGy h-1 MBq-1 m2. Clearly, if a source manufacturer uses one of these values in going from an air kerma rate measurement to a source activity specification and a computer software package uses the other in the reverse process for clinical dosimetry, then a systematic error in excess of 4% is introduced. These uncertainties and potential errors are avoided by using RAKR as the specification quantity.
The second concern raised by Dr Plane's letter relates to the use of a Farmer-type chamber in high fluence gradients such as those that exist in the close vicinity of a brachytherapy source. To a good approximation, the effect is to produce a reading that is indicative of the average emission rate over the volume occupied by the detector rather than of the emission rate at a precisely defined point (e.g. the geometric centre of the chamber). If the requirement is to determine the air kerma rate (or dose rate) at a point, then a geometry correction has to be applied. For a small (approaching a point) source, a Farmer-type chamber (e.g. 24 mm long by 6 mm internal diameter) and a source-to-detector centre distance of 20 mm, the correction factor for a measurement in air is approximately 1.10 using the Kondo and Randolph data [8] or 1.12 using the non-isotropic data from Bielajew [9]. These are not insignificant adjustments, and it is not obvious by how much they would change, or in which direction even, for measurements that are done in water. An appropriate factor for use with measurements done in a water tank would have to be determined by a Monte Carlo calculation. The problem is not reduced by irradiating the detector from different directions. The size of the required correction reduces with increasing distance but it is still about 2%, and therefore still significant, for accurate dosimetry, at a source-to-detector distance of 50 mm. Anyone making measurements with a Farmer-type chamber in these conditions must, therefore, exercise extreme caution.
Received for publication November 27, 2000. Accepted for publication December 21, 2000.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
