• Breast Cancer
• Oncology Genomics

This session explored the use of molecular profiling in three stages of breast cancer development: cacinoma in situ, primary, and metastatic. The treatment of in situ carcinoma of the breast is controversial. In the molecular progression from normal tissue to invasion, in situ carcinoma is a late-occurring step associated with an increased risk of the subsequent development of invasive carcinoma [Bombonati A, Sgroi DC. J Pathol 2011]. Lawrence J. Solin, MD, Einstein Medical Center, Philadelphia, Pennsylvania, USA, presented the currently available data on molecular profiling of in situ carcinoma of the breast.

Predictors of low-risk ductal carcinoma in situ (DCIS) are poorly defined using clinical and pathologic characteristics. Randomized trials of radiation and tamoxifen have shown a reduced risk for patients with DCIS. Molecular profiling for DCIS may shed light into its underlying tumor biology, potentially allowing for improved risk assessment and individualized treatment decisions.

During the same session, Antonio C. Wolff, MD, The Johns Hopkins Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA, provided an overview of molecular profiling of breast cancer. From 1980 to 2005, decisions about adjuvant chemotherapy were based on anatomy, nodal status, and tumor size, and adjuvant endocrine treatment was given if the tumor was estrogen receptor (ER)-positive. Since 2005, there has been a greater focus on stratifying breast cancer therapy decisions according to the biologic subtypes, including ER-positive/human epidermal growth factor receptor 2 (HER2)-negative, HER2-positive, and triple-negative breast cancer. While these phenotypes inform the choice of therapy, they do not offer predictive information on which patients will benefit from the treatment.

The American Society of Clinical Oncology (ASCO) tumor marker guidelines for breast cancer recommends measuring ER and progesterone receptor (PR) on primary invasive breast cancers to determine which patients are candidates for endocrine therapy [Harris L et al. J Clin Oncol 2007] and evaluate HER2 expression to guide selection of trastuzumab therapy. While adjuvant endocrine therapy is standard of care in patients with tumors that express ER and/or PR, the indication for adjuvant chemotherapy is less clear-cut. It is well established that ER-positive breast tumors derive less benefit from chemotherapy compared with ER-negative tumors [Berry DA et al. JAMA 2006].

The ASCO guidelines suggest use of the 21-gene Oncotype DX recurrence score (RS) assay to help identify subsets of patients with ER-positive breast cancers who may benefit from the addition of chemotherapy to endocrine therapy. Another prognostic multigene signature that may have utility in predicting for adjuvant chemotherapy benefit is the FDA-approved 70-gene MammaPrint signature. Unlike the 21-gene Oncotype RS where the genes were preselected, MammaPrint was developed using an unsupervised hierarchical clustering approach whereby the high-risk gene signature predicted for poor outcomes in tumors of all subtypes. Both these assays were tested retrospectively; however, the Oncotype RS was evaluated retrospectively in a prospectively assembled clinical trial (Table 1). Both these multigene signatures are currently undergoing prospective validation in large ongoing studies. The TAILORx trial [NCT00310180] will study the utility of the 21-gene RS signature to predict for chemotherapy benefit in the intermediate score range, while the MINDACT trial [NCT00433589] will study the outcomes of patients with discordant risk assessments when using the 70-gene signature and clinicopathologic features using the Adjuvant! Online program.

Lajos Pusztai, MD, DPhil, Yale Cancer Center, New Haven, Connecticut, USA, reported on the current use of genomic and molecular tests in routine practice and clinical research in metastatic breast cancer (MBC). In the management of MBC, there is no genomic test that is used routinely at this time. Measuring circulating tumor cells may help assess prognosis of MBC, but this currently has limited impact on the choice of therapy.

Discordance between HER2 or ER/PR receptor status in primary and metastatic lesions has been reported as 15% to 30%, depending on the study. This discordance—due to both technical reasons and true biologic changes in the cancer—is expected. It is not possible at present to distinguish between these two causes or to accurately estimate their contribution to differences in receptor status. [Pusztai L et al. Oncologist 2010]. Clinicians could consider repeating HER2 and/or ER/PR measurements on the metastatic tumors, if possible, as this may potentially aid treatment decisions.

Several genomic-based tests are being evaluated as predictive markers of therapeutic efficacy in MBC. These include the development of gene expression signatures to predict chemotherapy response, the identification of genomic abnormalities that have been successfully targeted in other cancers, and the development of gene predictors based on drug mechanism-of-action.

Molecular markers are being increasingly used in clinical trials for patient selection criteria or as an enrichment development. Targetable abnormalities appear to be individually rare in breast cancer, however they are present in up to 50% of breast cancers collectively. Due to low marker prevalence, Dr. Pusztai proposed that multiple tests should be performed at once and the results should be used to triage patients to targeted therapies. The Yale Molecular Analysis Prior to Investigational Therapy program is currently being conducted in a series of Phase 2 trials. These are biomarker-driven adaptive trials comparing the outcomes of patients with MBC with and without a specific gene mutation treated with specific targeted therapies. Dr. Pusztai concluded, “One of the most important future challenges is to design experimental and informatics tools that could guide how to combine targeted agents to match the multiple abnormalities that individual cancers have.”

• © 2013 MD Conference Express®