Re: HRAS1 Rare Minisatellite Alleles and Breast Cancer in Australian Women Under Age Forty Years

Extract

The report by Firgaira et al. (1) makes the potentially important observation that breast cancer in women under 40 years of age may not demonstrate association with rare alleles of the HRAS1 minisatellite. Breast cancer in younger women may differ genetically and biologically from late-onset disease and, therefore, the lack of association with genetic modifiers of modest effect would be an interesting, but not unexpected, outcome.

Although the authors mention potential biologic differences between cancers in patients of different age groups as one possible basis for the observed difference in rare allele association, they clearly favor a more prosaic explanation: that “new methods” with greater resolving power challenge the original observations and meta-analysis. Under this construction, the previously observed association is merely an artifact of the inability of previous methods to resolve more rare alleles among larger sized, common (low-risk) alleles.

I believe this study has two flaws, one technical and one conceptual, that undermine its conclusions. On the technical side, the newer methods using fluorescence-based genotyping on automated sequencing platforms do resolve alleles better. But a critical issue is the reliability of polymerase chain reaction (PCR) amplification in detecting the larger alleles that sequencers are capable of resolving. The PCR-based method described in the report by Firgaira et al. is not satisfactory in this regard. We know from our own experience that PCR strategies with straightforward cycling protocols can preferentially amplify lower molecular weight alleles at the expense of larger ones (“allele steal”) or they can create artifactual rare alleles from larger rare and common alleles. Our own protocol combines additives (dimethyl sulfoxide and betaine [N,N,N-trimethylglycine]), proofreading enzymes, and a multistage cycling protocol to minimize artifactual loss of larger alleles (2,3).