Matthew J.C. Ellis, MB BChir, PhD
For the past 10 years, Matthew J.C. Ellis, MB BChir, PhD, has focused his research on exploring the breast cancer genome in partnerships with academic and government-funded centers in the United States and the United Kingdom.
Ellis, the Anheuser-Busch chair in Medical Oncology at Washington University School of Medicine in St. Louis, Missouri, served as co-leader of a groundbreaking analysis of breast cancer tumors that was published in Nature
in September. The other co-leader was Charles M. Perou, PhD, who heads the Perou Laboratory at the Lineberger Comprehensive Cancer Center at the University of North Carolina at Chapel Hill.
The study, a project of the The Cancer Genome Atlas (TCGA), confirmed the existence of four major subtypes of breast cancer through the most comprehensive analysis to date of genetic mutations in breast tumors. Ellis talked with OncologyLive
about his research.
1. What was the biggest surprise when you analyzed the breast cancers across the six platforms?
We were quite surprised regarding the degree of similarity between basal-like breast cancer and high-grade serous ovarian cancer. Currently, these two diseases are treated differently. While both are treated with taxane drugs, ovarian cancer patients are typically administered a platinumbased agent because a decade or more ago, gynecologic oncologists decided that anthracyclines add little when there is a platinum agent in the regimen. Breast cancer therapy is still focused on the use of anthracyclines such as doxorubicin. Anthracyclines cause heart damage and leukemia, and so we have long been interested in a replacement regimen. The TCGA data therefore suggest that we should explore the hypothesis that a platinumbased regimen without an anthracycline may be just as effective for basal-like breast cancer, or even more effective than the standard regimens with fewer longterm side effects. Some clinicians have begun using these agents based on the reports in the literature. As of yet, however, a clinical trial using platinum-based compounds in place of an anthracycline-based regimen has not been conducted.
2. Is there a concise way to describe potentially optimal treatment regimens for the four subtypes?
Each subtype exhibited some type of feature that upregulated the PI3 kinase pathway. Luminal and HER2- positive tumors showed a high frequency of mutation specifically in the PIK3CA gene. Basal-like tumors tended to upregulate the PI3 kinase pathway by loss of negative regulators such as PTEN and INPP4B. These observations suggest that targeting this pathway with various inhibitors may give us a way to shut down a common deregulated pathway in each subtype.
3. Will your study, do you think, change the accepted paradigms for treating the three phenotypic groups (ie, HER2-positive, ERpositive, and TN)?
We now recognize that different breast cancers are so dramatically distinct that their treatment needs to be reevaluated. This is especially true of hormone-positive tumors where the typical targeted therapy regimen includes tamoxifen or aromatase inhibitors. If we combine inhibition of other pathways such as FGFR, PIK3CA, MDM2 and CDK4/6, we may be able to increase the therapeutic efficacy of endocrine therapy-based regimens.
4. Are there any tests, or would you like to see the development of any tests, that would further elucidate and define the subgroups?
In luminal-type breast cancers, a screen against PIK3CA mutation would be beneficial to see whether patients might be receptive to inhibition of this pathway. A test for EGFR phosphorylation status in HER2-positive disease could also help clinicians distinguish the two subtypes of HER2-positive cancer, and such an approach might be beneficial in predicting the response to trastuzumab, as well. BRCA1 and 2 mutation status could be useful to knowing which triple-negative breast cancer should be treated with PARP inhibitors.
5. What questions about treating breast cancer subtypes remain unanswered?
There are many. Perhaps one to highlight is that one of the standards in the diagnosis of breast cancer is to assess the status of HER2 in the disease. Prior to this work, it was typically assumed that HER2-positive breast cancers were homogeneous enough to warrant acrossthe- board trastuzumab treatment. The TCGA analysis demonstrates that at least two subtypes of the HER2- positive tumor exist. One overexpresses HER2 in combination with other upregulated receptor tyrosine kinases. The other possesses hormone positivity in addition to amplified HER2. This second group appears to express less HER2 with a reduction in signaling, implying that the disease is less addicted to HER2 signaling. This could help to explain why only half of patients with HER2-positive disease respond to trastuzumab therapy.