MammaPrint Finds Breast Cancer Patients Who Can Skip Chemo

Alexander M. Castellino, PhD

August 26, 2016

In breast cancer patients who were judged to be at high risk by clinical assessment, and who were thus destined for chemotherapy, use of the 70-gene signature test (MammaPrint, Agendia) found that half (49%) of these women could skip chemotherapy.

This finding comes from the large European study called MINDACT (Microarray in Node-Negative and 1 to 3 Positive Lymph Node Disease May Avoid Chemotherapy), which was published online August 25 in the New England Journal of Medicine.

These results were first presented at the 2016 annual meeting of the American Association for Cancer Research and were reported at the time by Medscape Medical News.

In the critical group of patients at high clinical risk but at low genomic risk, as identified with MammaPrint, the use of adjuvant chemotherapy was associated with a 5-year survival rate without distant metastasis that was 1.5% higher in comparison with patients who did not receive adjuvant chemotherapy — 95.9% vs 94.4%. Corresponding rates of 5-year disease-free survival and overall survival were also higher — 2.8% and 1.4%, respectively. The study was not powered to determine significance between the groups, but the magnitude of chemotherapy benefit appeared modest in consideration of its inconvenience, risks, and costs, the researchers note.

"The study confirms that it is possible to identify patients who are not going to benefit from adjuvant chemotherapy in a significant way," co–principal investigator of MINDACT Martine Piccart, MD, PhD, head of the Department of Medicine at the Jules Bordet Institute in Brussels, and cofounder and chair of the Breast International Group, told Medscape Medical News.

This is important because oncologists are anxious when they have to decide which of their patients with early-stage disease may benefit from adjuvant chemotherapy and for whom surgery and endocrine therapy is sufficient, she explained.

"The disease can be cured when we take the right decision at the beginning. When the disease returns, we can treat it, but not cure it," she said.

"MINDACT shows that a high-technology genomic assay can provide greater precision in establishing prognostic significance," Clifford A. Hudis, MD, CEO of the American Society of Clinical Oncology, told Medscape Medical News. "In future, such tools may even be used as part of staging.

"However, the question MINDACT now raises is how good of a prognostic significance is required to determine if a patient requires chemotherapy," Dr Hudis added.

He was referring to the 5-year survival rates without distant metastasis reported above and the 1.5% difference between patients who did not receive chemotherapy and those who did (94.4% vs 95.9).

"[A] difference of 1.5 percentage points, if real, might mean more to one patient than to another. Thus, the stated difference does not precisely exclude a benefit that clinicians and patients might find meaningful," he writes in an accompanying editorial coauthored with Maura Dickler, MD, of the Memorial Sloan Kettering Cancer Center and Weill Medical College, New York City.

Dr Hudis also notes the previous results for another genomic test, the 21-gene signature (Oncotype Dx, Genomic Health, Inc), in the TAILORx (Trial Assessing Individualized Options for Treatment) study, which may be of higher prognostic significance. That trial showed a 5-year survival without distant metastasis of 99.3% of patients with hormone receptor–positive, HER2-negative, node-negative breast cancer who had a recurrence score of <10 — a low-risk group.

The TAILORx results were published last year. At the time, Dr Hudis commented in an accompanying editorial: "This result [freedom from recurrence of breast cancer of 99.3% of the group with a low-recurrence score group] is numerically good enough to exclude any potential meaningful benefit for additional chemotherapy."

In a recent interview with Medscape Medical News, Dr Hudis pointed out that prognostic significance might matter if the bar is set high enough. "At a certain point, good prognosis wins the day," he said.

"We may expect to see more recurrences in 10 years, but this would not be prevented by chemotherapy," lead author and study chair of TAILORx Joseph A. Sparano, MD, of the Department of Oncology at Montefiore Medical Center and associate director for clinical research at the Albert Einstein Cancer Center, New York City, told Medscape Medical News.


"To convince clinicians and patients that chemotherapy is not needed in otherwise healthy younger patients would generally require setting either a high bar or a randomized design that would exclude a meaningful benefit for the intervention," Dr Hudis and Dr Dickler write in their editorial.

MINDACT set the bar high, they comment. The investigators enrolled nearly 6700 patients with early-stage breast cancer. Clinical and pathologic features, as well as information from MammaPrint, were used to stratify the patients into four groups — two concordant groups (high clinical and high genomic risk, low clinical and low genomic risk) and two discordant groups (high clinical and low genomic risk, low clinical and high genomic risk).

Dr Piccart explained that patients were evaluated by two methods. Clinical risk was determined on the basis of a computerized program and was not subjective. Determination of genomic risk (high or low) was based on the MammaPrint done on frozen tissue sample. MammaPrint no longer requires a frozen sample but is performed on fixed paraffin tissue samples.

Patients in the concordant groups were treated according to their risk category — no adjuvant chemotherapy for the low-risk group, and adjuvant chemotherapy added to endocrine therapy following surgery in the high-risk group. Patients in the discordant groups were randomly allocated to receive chemotherapy or not.

By comparing outcomes for patients whose treatment was guided by the MammaPrint test vs patients whose treatment was guided by conventional methods, MINDACT "tested the test," note the editorialists.

The primary endpoint of MINDACT was to show noninferiority against a predefined benchmark of a 5-year metastasis-free survival rate of 92% in patients at high clinical risk for whom chemotherapy was omitted because of their low genomic risk. That endpoint was achieved. The 5-year survival without distant metastasis was 94.7% (95% confidence interval, 92.5 - 96.2%), which established noninferiority.

TAILORx was also conducted in patients with early-stage breast cancer. It was a randomized study in which approximately 10,200 patients were clinically managed on the basis of breast cancer recurrence score. Patients in the low-risk group were given hormone therapy, and those in the high-risk group were given hormonal therapy and adjuvant chemotherapy. Patients in the intermediate-risk group were randomly allocated to receive either hormonal therapy or hormonal therapy and chemotherapy.

Although TAILORx confirmed that Oncotype Dx was of high prognostic significance, the power of Oncotype Dx resides in its predictive power, which has been shown in some retrospective analyses, according to Dr Hudis.

In TAILORx, the predictive power of Oncotype Dx was the primary endpoint of the prospective study — that is, whether patients in the intermediate group benefited from the addition of chemotherapy to hormonal therapy. Dr Hudis pointed out that these results have yet to be published. The same is true for the results from another study in patients with node-positive disease (RxPONDER); in that study, the benefits of chemotherapy are being evaluated in patients with a recurrence score of ≤25.

Dr Sparano explained that TAILORx is an event-driven trial. "If a significant advantage for chemotherapy emerges, or when there are a sufficient number of events for full information to exclude a benefit, the Data Monitoring Safety Board will release the data in accordance with the trial's prespecified statistical plan," he said.

Which Test to Use in Clinical Practice?

MammaPrint and Oncotype Dx are among two of the more widely used tests in clinical practice. Other tests, such as the Breast Cancer Index (7-gene signature), EndoPredict (12-gene signature), and Mammostrat (5-gene signature), are also available but are less frequently used.

Oncotype Dx has been the predominant test of choice in the United States, Dr Sparano commented. It has always been carried out on paraffin-embedded tissue, which allows retrospective testing. A high level of evidence from prospective-retrospective analysis of biospecimens derived from randomized trials demonstrated the clinical utility for the assay.

MammPrint can now compete on both points ― a newer version of the test, which has been available for about 2 years, utilizes paraffin-embedded tissue, and the MINDACT results now provide high-level clinical trial data.

Dr Piccart predicts that the findings from MINDACT will increase the use of MammaPrint. "Our data show that MammaPrint need not be ordered for every patient that enters the office," she said. If the patient is at clinically low risk, the test should not be ordered, because there are no advantages to giving adjuvant chemotherapy to these patients, she pointed out.

In patients at high clinical risk, ordering the test would identify the approximately 46% of patients with low genomic risk who can forgo chemotherapy, she indicated.

Dr Sparano agreed that the MammaPrint is a binary test that identifies patients on the basis of two risk categories, whereas Oncotype Dx categorizes patients on the basis of recurrence scores into low risk (<18), intermediate risk (18 to 30), and high risk (≥31). "It is in the midrange group of patients where the uncertainty lies," he said.

According to Dr Piccart, the low-risk group represents about 16% of patients, whereas the intermediate-risk group represents 67% of patients — the population for which definitive evidence is lacking with TAILORx.

However, Dr Sparano told Medscape Medical News that the patients for whom clinicians order these tests are not necessarily the same types of patients who are studied in prospective trials. "They tend to be ordered by clinicians in circumstances where uncertainly exists about the role of chemotherapy ― usually patients with tumors between 1 and 2 cm and/or intermediate grade," he said.

In a press release, the manufacturer of the MammaPrint test emphasized that it is "the only genomic assay with level 1A evidence to help physicians more accurately predict risk of distant metastasis in patients with early-stage breast cancer."

MammaPrint and Oncotype Dx both cost approximately $4000 in the United States.

Dr Hudis hopes for less expensive tests. Commenting on TAILORx, he writes: "This multigene assay is unlikely to be the only test that can provide a prediction of chemotherapy benefit.... A less expensive and broadly distributed test would be valuable globally."

In their editorial, Dr Hudis and Dr Dickler write: "On the basis of the MINDACT study, clinicians may consider ordering the 70-gene signature for patients in line for chemotherapy who hope to forgo it on the basis of low genomic risk.

"What doctors and their patients do with the results of such testing will be highly individualized — and will inevitably be finessed by findings from future studies," they add.

The MINDACT authors agree, commenting: "[A] risk-benefit assessment and decisions with respect to the use of adjuvant chemotherapy are highly variable among physicians, and even national and international guidelines differ in their recommendations.

"Ultimately, the decision to receive or forgo chemotherapy (or any other treatment) lies with each patient who is properly informed about the potential side effects and the potential benefits of such treatment. For the same risk-benefit scenario, different patients may make different decisions," they state.

Dr Piccart said that MINDACT cost around €45 million ($51 million), of which €7 Million came from the European Union and the remainder from grants from various cancer foundations and from the pharmaceutical industry.

Dr Piccart has disclosed no relevant financial relationships. Several coauthors have relationships with pharmaceutical companies, which are listed in the original article. Dr Hudis and Dr Dickler have disclosed no relevant financial relationships. Dr Sparano holds a patent related to tests to predict responsiveness to chemotherapy treatment options in patients with cancer.

New Engl J Med. Published online August 25, 2016. Abstract, Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.