inDEPTH: Who Are the Women With Breast Cancer Survival Rates Near 100%?

Fatima Cardoso, MD Director Breast Cancer Unit Champalimaud Clinical Cancer Center Lisbon, Portugal

MINDACT was conducted to answer, “Can we define prognosis better by using traditional pathology, the biomarkers—like ER, PR, HER2—plus the genomic tests, like in this case, MamaPrint. Together, can we make it better than when we use traditional pathology alone?” The 6,600 patients enrolled in the trial all had their risk assessed by the traditional biomarkers and by the genomic test. We had a clinical pathological risk based on age, menopausal status, size of the tumor, and nodal status since MINDACT included node negative and 1, 2, and 3 positive nodes, ER, PR, HER2, and MammaPrint. With that, you could have three situations. 1) Both methods would say that the risk was low, in which case the patient was proposed to receive endocrine therapy alone. 2) Both methods would say that the risk of recurrence is high, and the patient was proposed to have chemotherapy followed by endocrine therapy. 3) One would say the risk was high and the other would say the risk is low, in which case we randomized for the method and followed the genomic risk or the clinical risk and then proposed chemotherapy or not according to that risk so that we would know which is more frequently correct, the genomic risk or the clinical pathological risk. We now have a medium follow-up of 8 years.

MINDACT was one of the first de-escalation studies, because we wanted to prove that by adding the genomic risk, we would prescribe less chemotherapy in ER+, HER2- disease.  For de-escalation studies, it is difficult to define the proper endpoint, and sometimes the proper endpoint cannot be a comparative. And this was one of the cases; we had a predefined threshold and we said, “If we’ve reached that threshold, then we reach our primary endpoint.” We chose distant disease-free survival, because it is much more closely related to overall survival than just disease-free survival. We really wanted to know just distant metastasis because that’s what eventually kills a patient. We wanted to have at least 92% at 5 years of distant metastasis-free survival when we do not give chemotherapy based on genomic risks.

These tests are expensive, between 2,500 to 3,000 Euros. That’s added to all the other expenses that someone with breast cancer already has. However, these genomic tests are cost-effective through prescribing an estimated 40% less chemotherapy, which comes not only with the costs of treatment, but the consequences of the treatment, like not going to work and being on sick leave. Many studies of health technology assessments show that they are cost effective.

William Audeh, MD, MS Chief Medical Officer Agendia

MammaPrint predicts the likelihood of distant metastatic recurrence in a breast cancer, based upon the expression levels of 70 genes that are important in the metastatic process. MammaPrint was validated in the MINDACT trial. BluePrint is made up of 80 genes that assess the signaling pathways driving breast cancer growth. We recognize three subtypes: luminal, HER2, and basal. Together, we perform MammaPrint and BluePrint on the same breast cancer sample and produce a coordinated, comprehensive classification.

While immunohistochemistry looks at protein products of the genes and draws assumptions based upon the presence or absence of key proteins, the mere presence of those proteins doesn’t tell you all about the actual biology and the inner workings of breast cancer. As we add to the information provided by pathology by further defining, for example, hormone positive breast cancers into multiple types based upon their gene expression, each has clinical relevance for patient management.

Recently, the large AFFINITY trial—a HER2+ trial to which we also applied MammaPrint and BluePrint—found a proportion of HER2+ breast cancers that were genomically more dominantly basal in their intrinsic type. HER2-targeted treatment wasn’t as effective as it would have been for a true HER2 type defined genomically. About one-half of the clinically high-risk women who were routinely receiving chemotherapy, who based on clinical features would have been given chemotherapy, did not need it and did just as well without it.

Our next steps are to apply MammaPrint and BluePrint, as well as additional new signatures of prediction of response to therapy, to new therapies coming into the clinic. Strong data from the I-SPY trial indicates that with the MammaPrint assay in hormone-positive breast cancers, we can correlate the level of the MammaPrint risk index with the likelihood of responding to new therapies. Once you separate the subgroups within breast cancer into more biologically and clinically relevant groups, you see a very different rate of response for certain drugs, and we hope to avoid the use of drugs that are unlikely to be helpful.  As we look at the entire transcriptome—which is an aspect of the nearly 10,000-patient FLEX trial—we are finding additional signatures and pathways that are important to understanding the clinical differences in breast cancer.

In the low-risk range of MammaPrint, we can identify women who are ultra-low risk and appear to do extremely well even with less that the full 5 years of endocrine therapy, which is the standard of care. We certainly wouldn’t recommend that the standard of care not be recommended. However, ultra-low MammaPrint risk patients—we have shown in four cohorts—have an exceptional outcome out to 10 or 20 years with less than the full 5 years. Such women may be able to stop early if they are having a great deal of toxicity. We’ve also shown in the NSABP B-42 trial that MammaPrint can identify women who benefit from extended endocrine therapy out to 10 years. MammaPrint offers a whole range of information that we believe can guide women and their physicians to make the best decisions for them.