Is there any role for new prognostic markers in breast cancer?

Abstract

Breast cancer is heterogeneous in its development, progress and response to treatment. In the initial stages, surgery is usually the fi rst treatment option; however, the risk of recurrence and death has been associated with certain clinical and biological variables, including nodal involvement, histological grade, degree of endocrine responsiveness, HER-2 amplifi cation and tumour size [1]. For over ten years, recognition of the intrinsic subtypes of breast cancer has allowed the disease to be classifi ed into different types, with the use of adjuvant therapies tailored to the biological profi le of each type [2‐4]. Different genomic platforms have been developed to help clinicians and patients make the best decision regarding adjuvant therapy. The Oncotype DX platform estimated the risk of recurrence in tumours with hormone sensitivity criteria, and defi nes more precisely which patients will not benefi t from receiving a particular chemotherapy regimen and those that will. The TransGEICAM study in 107 hormone receptor-positive and lymph node-negative patients detected a change in the type of treatment in 32% of cases: 21% from chemohormone therapy to hormone therapy and only 11% from endocrine therapy to the combined chemotherapy and hormone therapy. High histological grade and elevated Ki67 were associated with increased recommendation for chemotherapy, whereas progesterone receptor expression was associated with a recommendation for hormone therapy alone [5]. Other genomic signatures provide a dichotomous outcome of good or poor prognosis and, on this basis, indicate a recommendation to not administer chemotherapy in the fi rst case while this treatment should be given in the second. However, validation of these platforms was made on the basis of retrospective randomised studies using treatment regimens which are scarcely used today. Moreover, some researchers have shown that the clinical criteria supporting the decision to give a complementary treatment do not always match the information obtained from DNA expression arrays; for this reason two large multicentre studies, MINDACT and TAILORx, were designed in which patients have been randomised to receive endocrine therapy or a combination according to clinical or genomic tumour characteristics. The results of these trials are pending. Additionally, the use of genomic signatures has other limitations. Firstly, they provide a static view of the disease