Purpose: To review the preclinical and clinical studies relevant to the prognosis and prognostic associations of BRCA1- and BRCA2-associated breast carcinomas, with an emphasis on research methodology.
Methods: Reports of relevant studies obtained from a MEDLINE search, and references from these articles, were critically reviewed.
Results: Consistent associations with both favorable (medullary or atypical medullary carcinoma) and unfavorable (high tumor grade, hormone receptor negativity, somatic p53 mutation) prognostic characteristics have been found for BRCA1-associated breast carcinomas. Inconsistent results have been demonstrated for prognostic associations of BRCA2-associated breast tumors. Clinical studies that have directly assessed the prognosis of these tumors have not shown a clear effect of BRCA1 or BRCA2 mutation, but no study has used optimal methodology. In vitro and animal model data suggest a possible influence of these mutations on response to agents that cause double-strand DNA breaks, but clinical data are limited.
Conclusion: The elucidation of an identifiable subgroup of breast carcinomas that result from germline mutations in BRCA1 or BRCA2 may be an important step toward genotype-based understanding of prognosis and choice of therapy in this disease. However, currently there are inadequate data to support use of BRCA1 or BRCA2 status to counsel individuals regarding their prognosis or to select treatment. Well-designed studies of population-based inception cohorts of breast cancer patients, which have adequate sample size and complete follow-up, and which use objective outcome criteria and blinding of outcome assessment, are required to optimally address this question.