The emerging data on genomic predictors in estrogen receptor–positive breast cancer suggest that the best treatment decisions are going to be made by integrating clinical, pathologic, and genomic measures.— Harold J. Burstein, MD, PhD
Genomic assays have had a powerful influence on the management of early-stage breast cancer, particularly estrogen receptor–positive tumors. The mainstay of adjuvant treatment for early-stage, hormone receptor–positive breast cancer has been endocrine therapy, either with tamoxifen and an aromatase inhibitor, or the use of ovarian suppression. Furthermore, we have offered chemotherapy to many patients with estrogen receptor–positive breast cancer in addition to endocrine therapy.
In 1997, the NSABP (National Surgical Adjuvant Breast and Bowel Project) B-20 trial showed that adding chemotherapy to tamoxifen improved outcomes for women with estrogen receptor–positive, node-negative breast cancer, setting the stage for widespread use of chemotherapy in this setting.
Influencing Practice Patterns
That pattern of practice changed with the development of the Oncotype DX 21-gene recurrence score assay. This test, which measures the levels of expression of 21 different genes within estrogen receptor–positive breast cancers, served as a prognostic assay, identifying patients at lower or higher risk of recurrence despite endocrine therapy. Further studies showed that the test could discriminate which tumors warranted adjuvant chemotherapy in addition to endocrine treatment and it showed a cohort of women who received “zero” benefit from the addition of chemotherapy to their treatment program. This was accomplished first for women with node-negative breast cancers, based on the NSABP B-20, and then for women with node-positive breast cancers, based on results from SWOG 8814. Based on these retrospective analyses, genomic testing has become routine for management of estrogen receptor–positive breast cancers and is widely used in cases where there is a question of the need for chemotherapy. To date, hundreds of thousands of women in the United States alone have had their tumors tested for the 21-gene recurrence score.
The use of this genomic tool has markedly influenced practice patterns. Data from individual treatment centers and larger registries have shown that the percentage of women receiving adjuvant chemotherapy drops dramatically based on genomic testing. In addition, recent prospective studies have corroborated that women with estrogen receptor–positive tumors who have very low 21-gene recurrence scores have an excellent short-term prognosis and do not warrant chemotherapy.
In addition to the robust literature emerging for the use of the 21-gene recurrence score, a variety of other molecular signatures have been in development for breast cancer. They include the “intrinsic” subset classifier, which categorizes estrogen receptor–positive tumors as luminal A or luminal B based on the pattern of gene expression. It also includes the MammaPrint assay, a 70-gene assay that dichotomizes breast cancers into “good” or “bad” prognosis.
Who Can Safely Avoid Adjuvant Chemotherapy?
The MINDACT trial1 was reported at the 2016 Annual Meeting of the American Association of Cancer Research and published in The New England Journal of Medicine—its findings are summarized in the November 10, 2016, issue of The ASCO Post. This prospective trial was designed to validate the use of the MammaPrint test for determining the treatment needs of women with early-stage breast cancer. The top-line results serve to validate the overall idea that genomic tests can help figure out which cancers need adjuvant chemotherapy and which do not.
The MINDACT trial included women with estrogen receptor–positive breast cancers but also women with estrogen receptor–negative and HER2-negative cancers, as well as women with HER2-positive breast cancers. There were, however, relatively few women with either HER2-positive cancers or “triple-negative” breast cancers. Given those small numbers, and the fact that data suggest nearly all such cancers would benefit from chemotherapy, or chemotherapy plus anti–HER2-positive treatments in the case of HER2-positive tumors, the MINDACT data speak mostly to women with estrogen receptor–positive, HER2-negative breast cancer.
The MINDACT trial made an assessment of the tumor risk based upon the clinical features of the cancer (tumor size, nodal involvement, and grade), as well as on the genomic score derived from the MammaPrint assay. If both the MammaPrint result and the clinical risk were low, the patient was offered hormonal therapy without chemotherapy. If both the MammaPrint result and the clinical risk were high, the patient received endocrine therapy and chemotherapy. In the cases where there was discordance between the clinical and genomic risk measures, patients were randomly assigned to receive chemotherapy or not.
The principal findings were that patients with low-risk clinical and low-risk genomic tumors had an excellent prognosis without chemotherapy, arguing that chemotherapy can be safely omitted in such cases. In addition, among the tumors that had high-risk clinical features by stage but a low genomic risk, there seemed to be negligible benefit for chemotherapy. An interesting finding here, also seen in studies with the 21-gene recurrence score, is that the small set of tumors with high-grade features but a low genomic risk seems to have a favorable prognosis, even without chemotherapy.
The data in MINDACT provide a powerful confirmation of the importance of genomic testing in the management of early-stage estrogen receptor–positive breast cancer. Multiple studies now show that either the 21-gene recurrence score assay or the MammaPrint assay (based on the MINDACT study) provides information that may shift the management of these cases, often sparing women from unnecessary chemotherapy.
Traditional Pathology vs Genomic Assays
An important remaining question, with implications for use of these technologies around the world, is whether high-quality pathology can substitute for a genomic test. On the one hand, there is substantial overlap between cancers that have low-grade features, with extensive estrogen receptor and progesterone receptor expression, and low measures of proliferation such as Ki67, usually aligning as either a low recurrence score, low MammaPrint, or luminal A–type cancers. By inference, these cases probably do not require chemotherapy.
However, as the MINDACT trial and other related experiences have shown, the concordance between traditional pathology and genomic assays is not perfect. There are not infrequent cases of high-grade cancers with low-risk genomic findings, and occasionally low-grade cancers have unexpected high-risk genomic features. The emerging data on genomic predictors in estrogen receptor–positive breast cancer suggest that the best treatment decisions are going to be made by integrating clinical, pathologic, and genomic measures. ■
Disclosure: Dr. Burstein reported no potential conflicts of interest.