First National Data on Breast Cancer Subtypes Mark New Era in Biomarker Epidemiology

Get Permission

Lynne Penberthy, MD

Judy Salerno, MD

The SEER program will continue to develop reporting of cancers by these more clinically meaningful categories starting in the coming year.

—Lynne Penberthy, MD

This year’s Report to the Nation on the Status of Cancer contains the first national combined data set on the incidence of four major breast cancer subtypes by race/ethnicity, poverty level, geography, and other factors. The findings show that “there are unique racial/ethnic-specific incidence patterns for breast cancer subtypes, likely because of both biologic and social risk factors, including variation in mammography use,” the authors concluded.

Coauthored each year by the North American Association of Central Cancer Registries, the National Cancer Institute (NCI), the American Cancer Society, and the Centers for Disease Control and Prevention, the report also documents a continuing decline in mortality and incidence for most cancers. It was published online in the Journal of the National Cancer Institute.1

While the subtype findings generally confirmed those of smaller studies, this first gathering of nationwide data on molecular subtypes of a specific cancer is an important landmark, experts say. Covering over 90% of the U.S. population in 42 states and the District of Columbia, it is a first for the surveillance community.

Presentation Precedent

“It sets a precedent for the presentation of cancer surveillance data according to important clinical subcategories,” said Lynne Penberthy, MD, Associate Director of the NCI’s Surveillance Research Program. “The SEER program will continue to develop reporting of cancers by these more clinically meaningful categories starting in the coming year,” she said.

The subtype findings are expected to help researchers more accurately stratify breast cancer by clinically relevant degrees of risk and will help patients understand their risk and the link between biomarker status and risk.

“It is an important first step, really a step forward,” said Judy Salerno, MD, the CEO of Susan G. Komen for the Cure, which has launched a major effort to address breast cancer disparities. “More is needed—to help patients understand therapy options and how they may differ by subtype and to develop a better focus on outcomes by subtype.”

Komen is now holding a series of Disparities Roundtables in urban areas around the country. It also funds disparity research and scholarships for young researchers in the field. “This is a national priority,” she said, “and we need to shine a light on it.”

Subtype Findings

The new data cover the four major breast cancer subtypes, which are based on HER2 and hormone (estrogen and progesterone) receptor status.

Luminal A tumors—those positive for hormone receptors (HR) and negative for HER2—carry the lowest risk. This subtype was most common among all racial/ethnic groups but highest for non-Hispanic whites. It declined with increasing poverty for every group and was generally higher in states with higher rates of mammography screening.

The triple-negative subtype (negative for both hormones and for HER2), which has the worst prognosis, was most common among non-Hispanic blacks and younger women. Non-Hispanic blacks also had the highest rates of distant-stage disease at diagnosis.

While the data suggest differences in access to care, said Dr. Penberthy, they don’t explain all the disparities among racial and ethnic groups. For instance, among all those with the triple-negative subtype, non-Hispanic black women had the most poorly differentiated or undifferentiated tumor cells. “This suggests something else may be going on,” she said.

The other two subtypes are luminal B (positive for both HR and HER2) and HER2-enriched (negative for HR and positive for HER2). There was little ethnic/racial variation in incidence for these two subtypes.

Streamlined Data Collection

Planning is already underway to collect similar data on other cancer subtypes. In lung cancer, a pilot study is now gathering data on the EGFR and ALK markers, Dr. Penberthy said. A bump in lung cancer incidence is expected as more people begin using spiral computed tomography screening, recently approved for Medicare reimbursement.

Work is also in progress to improve and streamline the collection process, an effort that NCI’s Surveillance, Epidemiology, and End Results (SEER) program launched several years ago.

One vital part of that process is the flow of information from oncology practices, pathology laboratories, and hospitals to state registries. Though mandated by state laws, there are gaps in the process, Dr. Penberthy said. Some cancers are not treated in hospitals, and smaller oncology practices may not know about the mandate or may not have the resources to do the reporting.

To facilitate the process, “we are trying to learn how to automate it and to use natural language processing,” she said. The hope is that natural language processing—also known as “text mining”—will allow information to be automatically extracted from pathology reports and sent directly to state registries. If successful, these new methods and technologies will be made available to all registries and will populate data within the SEER public use files.

SEER already has an agreement with Genomic Health, the company that performs the molecular diagnostic assay Oncotype DX for early-stage breast cancer, to have its data on test results sent directly to the SEER registries.

More Treatment Data

Another long-term goal of NCI’s is to begin collecting more data on treatment. Current reporting, limited to whether or not the patient received one or multiple chemotherapy agents, is not sufficiently robust to help us understand disparities in outcomes, Dr. Penberthy said.

Dr. Salerno noted that more data on the use of oral agents, targeted at specific molecular biomarkers, could be especially important because of their cost and the potential for disparities. “We know there is a significant issue with insurance coverage,” she said.

A pilot program at NCI, now being developed, will test ways to capture more data on treatment, including chemotherapy, radiation, and oral agents, directly from oncology practices. “This is a big initiative,” Dr. Penberthy said. “I feel very, very strongly that this is something we have to address. It is a huge challenge…. We have to be inventive.” ■

Disclosure: Drs. Penberthy and Salerno reported no potential conflicts of interest.


1. Kohler BA, Sherman RL, Howlader N, et al: Annual report to the nation on the status of cancer, 1975–2011, featuring incidence of breast cancer subtypes by race/ethnicity, poverty, and state. J Natl Cancer Inst 107(6):djv048, 2015.