Classification systems that categorize breast cancers based on estrogen receptor (ER), progesterone receptor (PR), and HER2 expression levels may obscure the heterogeneity of other key tumor features, according to a new study published in The Oncologist. Some breast tumors that share a single clinical label, such as triple-negative breast cancer (TNBC), may in fact represent a diverse collection of molecular subtypes.
Many breast cancer treatment decisions hinge on whether tumors test positive or negative for ER, PR, and/or HER2, yet the criteria for interpreting a test result can vary. The American Society of Clinical Oncology (ASCO) and the College of American Pathologists (CAP) recommend thresholds of <1 percent staining to classify a tumor as hormone receptor (HR)-negative, and similarly restrictive criteria for HER2 negativity. Historically, however, definitions of HR negativity have reached nearly 10 percent, and patients with borderline (1 to 9 percent) HR staining have been enrolled in TNBC trials.
Most TNBCs are believed to share a similar molecular tumor subtype called Basal-like. In the current study, researchers showed that the degree of overlap between the TNBC and Basal-like categories changes substantially with different definitions of "negative" HR staining.
Lisa A. Carey, MD, of Lineberger Comprehensive Cancer Center at the University of North Carolina in Chapel Hill, described the results of the study conducted in collaboration with Angelo di Leo, of Istituto Toscano Tumori in Prato, Italy, on behalf of the U.S. National Cancer Institute and the Breast International Group. "Including tumors with borderline HR staining in the definition of triple-negative breast cancer significantly diminished the proportion of Basal-like molecular subtypes. The optimal threshold for enriching for Basal-like breast cancer is <1 percent for either hormone receptor, in alignment with current ASCO/CAP guidelines."
The study included 1,557 breast tumor samples that were tested for ER, PR, and HER2 expression by standard clinical assays. The same tumors were also tested for their molecular features and classified into 1 of 5 molecular subtypes: Luminal A, Luminal B, HER2-Enriched, Basal-like, and Normal-like.
Among TNBCs with <1 percent HR staining, the most common molecular subtype was Basal-like (73 percent), followed by HER2-Enriched (17 percent). By comparison, TNBC tumors with borderline HR staining had a much wider mix of molecular subtypes, including Luminal A/B (44 percent), HER2-Enriched (31 percent), and Basal-like (18 percent).
"Our findings show that borderline HR-expressing tumors are heterogeneous and do not fit well into distinct molecular categories," Dr. Carey said. "This raises the question of whether 'borderline' HR staining should instead be considered 'indeterminate' and require additional assays to clarify the underlying biology."
Gabriel Hortobágyi, MD, FACP, professor of medicine in the department of breast medical oncology at The University of Texas MD Anderson Cancer Center, and a member of The Oncologist's editorial board, noted, "The manuscript by Cheang et al. provides strong confirmatory evidence of the prognostic and predictive value of intrinsic subtypes and provides guidance about how we can enrich patient groups by the use of quantitative hormone receptor and HER2 expression, and where more sophisticated technology is needed to understand the biology of breast cancers. The days of breast cancer as a single entity are clearly over. We are well on our way to personalized breast cancer therapy."