Intended for healthcare professionals


Link between breast cancer and ATM gene is strong

BMJ 1999; 318 doi: (Published 06 February 1999) Cite this as: BMJ 1999;318:400
  1. Michael Swift, Director,
  2. Yun Su, Epidemiologist
  1. Institute for the Genetic Analysis of Common Diseases, New York Medical College, Hawthorne, New York 10532, USA

    EDITOR—We welcome Lavin's recognition of the important role that the ataxia-telangiectasia (ATM) gene plays in breast cancer despite the attention given to two other genes, BRCA1/BRCA2, which are numerically less important.1 It is important to emphasise that the evidence linking the ataxia-telangiectasia gene to breast cancer is more than suggestive. As well as the two smaller studies referenced by Lavin, two large epidemiological studies have clearly indicated the increased risk of breast cancer among blood relatives of patients with ataxia-telengiectasia. 2 3

    In 1996 a study using genotyping technology and the statistically powerful unbiased index test method for testing gene-disease associations provided compelling evidence that carriers of the ataxia-telangiectasia gene have a 3.8-fold increased risk of breast cancer compared with non-carriers.4 With a P value of 0.0001, it is unlikely that this result arose by chance. More precise estimates of the magnitude of this risk will become available as more breast cancer cases in families with ataxia-telangiectasia are genotyped.

    There is no conflict between these studies and the study by Fitzgerald et al,5 as Lavin suggests. The apparent lack of association found by Fitzgerald et al is probably because they studied only women who developed breast cancer before the age of 40, based on a speculation by Easton that is not supported by any published data. Furthermore, women with BRCA1/BRCA2 mutations accounted for a substantial proportion (13%) of Fitzgerald's study sample. An unmatched comparison group was used; it is highly unlikely that the distribution of the most important factors that affect gene frequency—ethnicity and social class—in their blood donor “control group” resembled that in the group of breast cancer patients.

    Such studies also have low statistical power when the population frequency of heterozygotes is about 1-2%. Additionally, they sought ATM mutations through the protein truncation test, which typically fails to detect about half of all such mutations.

    In conclusion, the evidence supporting the association between mutations at the ataxia-telangiectasia locus and breast cancer is strong. No study with sound methods has refuted this association.


    View Abstract