Improved safety and effectiveness of imaging predicted for MR mammography.

Abstract

Sir, The recent review article by Kneeshaw et al (2003) in the British Journal of Cancer regarding advances made in breast imaging by magnetic resonance is very encouraging and its expedited entry into routine clinical practice should be supported. This is not only due to its anticipated enhanced sensitivity and usefulness, but potentially also for its safety, particularly as may apply to radiation hypersensitivities such as female carriers of ataxia telangiectasia (AT). Ataxia telangiectasia is a genetic cancer predisposition syndrome involving the overexpression of alpha-fetoprotein, immunosuppression of the host, advanced ageing, and a radiation hypersensitivity where AT homozygotes generally succumb to infection or to lymphoma before reaching adulthood (reviewed in Becker-Catania and Gatti, 2001). Female AT heterozygotes are at an increased risk of breast cancer, but this is usually detected after the age of 40 years (Swift et al, 1991; Swift, 2001,). The issue of safety of ionising radiation used for breast cancer screening has been previously raised for AT heterozygotes (Swift et al, 1991). About 1% of the population carries the genetic marker for AT, and 8–10% of all breast cancers appear to be AT heterozygotes (reviewed in Swift, 2001). This indicates the use of nonionising modes of breast imaging could make a significant impact on the incidence of breast cancers. However, other evidence suggests these potential benefits might also be extended into the general female population. A number of years ago while investigating the molecular biology of the AT radiation hypersensitivity, it was discovered that radioresistant DNA synthesis (RDS), the molecular signature of the radiation hypersensitivity of AT, could be modulated (Mirzayans et al, 1995; Mirzayans and Paterson, 2001). When c-myc and alpha-fetoprotein became implicated in the RDS phenotype (Laderoute, unpublished findings), oestrogen was tested for its ability to induce RDS in cells derived from normal individuals. As shown in Figure 1, RDS was induced by estradiol (E2) but not progesterone in a dose-responsive fashion in lymphoblastoid cell lines (LCLs) derived from normal individuals (panel A). The induction was specific to oestrogen as it was blocked in the presence of an excess of tamoxifen (panel C), an oestrogen antagonist (Laderoute, 1998). Furthermore, additional experiments showed the RDS assay could be geared to easily detect low dose ionising radiation exposures around 1 rad or less in radiation hypersensitive cells such as those from AT individuals or AT heterozygotes (Laderoute, 1998). These results raised the possibility that breast imaging techniques using low-dose ionising radiation may induce breast cancers or cause existing breast tumours to progress, if the circulating levels of oestrogen are elevated in women at the time of mammography exposure. This phenomenon might help explain the reduced benefits in the under 50 years age-at-entry group reported in mammogram clinical trials (Miller et al, 1992, 2002), the increased incidence of breast cancer with mammogram screening among 45- to 64- year olds not accountable by earlier detection (White et al, 1990), and the lack of impact of mammography screening for reducing breast-cancer mortality overall (Gotzsche and Olsen, 2000). Figure 1 Estrogen specifically induces RDS in cells from normal individuals. Radioresistant DNA synthesis (RDS) measures the relative failure of cells from AT individuals to inhibit DNA synthesis at 1 h following ionising radiation exposures when compared ... It remains to be seen if the incidence of breast cancer returns to the premammogram screening rates of the 1960 s and 1970 s, as ionising radiation-based breast imaging technologies are replaced by MR imaging and as women decline hormone replacement therapies. One might also expect survival to again correlate inversely with tumour size, which is why early breast cancer imaging methods were introduced in the first place.