| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Correspondence |
We thank Dr Heyes and his colleagues for sending us an advance copy of their paper [1], in which they report a relative biological effectiveness (RBE) of 4 for mammographic X-ray energies relative to the much higher energies of atom bomb radiation. We do not presume to comment on the biological aspects of their work, but we note that their results agree well with those of two other research groups, although all such work from these three groups has been done on in vitro systems rather than in vivo.
However, we are concerned with the conclusions the authors reach on radiation risk estimation in mammography, in which low energy X-rays are used, and especially in breast screening programmes. Heyes et al [1] state that "risk estimates for radiation-induced cancer – principally derived from the atomic bomb survivor study (ABSS) – are based on the effects of high energy gamma-rays and thus the implication is that the risks of radiation-induced breast cancer arising from mammography may be higher than that assumed based on standard risks estimates". We have to point out that their statement, at least as far as it concerns mammography in the UK, including the UK National Health Service breast screening programme (NHSBSP), is incorrect. Later in their paper, Heyes et al [1] acknowledge that the risk estimates we have used are in fact based on studies of individuals receiving medical exposures, but appear to overlook this in the calculations and conclusions of their paper. They conclude that because they find the biological effect of mammographic X-rays to be four times greater per unit dose than atomic bomb spectrum exposures in their experiments, current risk estimation in mammography should be multiplied by four. We disagree because the various risk factors we have used at different times have all been derived from studies of medical exposures at energies much closer to those of mammographic X-rays than to atomic bomb exposures.
Breast cancer induction risk factors used by Law in a series of papers from 1987 to 1993 were stated in the first of those papers [2] to be based solely on studies of patients receiving multiple chest fluoroscopies or treatment for post-partum mastitis. From 1995 to 2002, the same author and both present writers used risk factor estimations provided by National Radiological Protection Board (NRPB) (now the radiation division of the Health Protection Agency), which we understand are also based entirely on medical exposures. In 2003, Young et al used very slightly different risk factors, also from NRPB, and also based solely on medical exposure of North American women [3]. Thus, no contribution to any of these three sets of figures has been derived from the ABSS.
Modern mammography is most commonly performed at around 28 kV to 30 kV peak on X-ray sets having molybdenum targets and filters, or with other target/filter combinations, giving somewhat higher mean photon energies. The mean photon energy for the Mo/Mo combination at 28 kV to 30 kV will be within the range of 15 keV to 20 keV [4]. For other target/filter combinations in use in mammography, the mean photon energy will be slightly higher.
For the multiple chest fluoroscopies, it is difficult to estimate mean photon energy precisely, but in the Massachusetts study exposures were made at 60–80 kVp with 0–1 mm Al filtration, giving a mean photon energy of about 40 keV [4].
Photon energies with the post-partum mastitis treatments are even more difficult to estimate precisely. Initially these were made at tube voltages ranging from 120 kV with 3 mm Al filtration to 200 kV with 1 mm Al + 0.5 mm Cu [6]. Later treatments ranged up to 250 kV (filtration not known) [7]. Verhaeggen et al have calculated mean photon energies for a range of orthovoltage therapy beams. For example, 100 kV with 1 mm Al gave a mean photon energy of 47 keV, while 250 kV with 1 mm Al +1 mm Cu gave a mean of 114 keV. Thus, the mean photon energy for all the post-partum mastitis patients is very likely to have been less than 100 keV.
Heyes et al [1] do not estimate a mean photon energy for the atomic bomb spectrum, but in an earlier paper [9] they describe how they simulate that spectrum using 6 MV photons (82%) combined with 10–14 MeV electrons (18%), giving an equivalent photon energy which must be greater than 2 MeV.
Thus the current risk factors for breast cancer induction are based on photon energies much closer to those of mammography than the atom bomb exposures.
It is our view that the work of Heyes et al does not imply any need to modify current risk factors used in mammography on grounds of RBE at this time. (Those risk factors may require revision for other reasons). It would be unfortunate if readers were to assume without further consideration that mammographic or breast screening risks must be substantially revised upward on the basis of the paper by Heyes et al.
Yours etc.,
1 Edinburgh University, Department of Medical Physics, Chancellor's Building, Little France Crescent, Edinburgh, EH16 4SB, 2 Quality Assurance Reference Centre, 9 Kingfisher Way, Silverlink Business Park, Wallsend, NE28 9ND, 3 Radiation Protection Service, St Luke's Wing, Royal Surrey County Hospital, Guildford, GU2 7XX, UK
References
Related articles in BJR:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
