Many women with high-risk breast cancer by conventional criteria avoided adjuvant chemotherapy with no increased risk of metastatic disease by use of a cancer-genetics test, according to data reported here.

Patients at high clinical/pathologic risk but low risk by the 70-gene MammaPrint assayhad a 5-year distant metastasis-free survival (DMFS) of 94.4% without chemotherapy versus 95.9% with chemotherapy, a difference that did not reach statistical significance (HR 0.78, 95% CI 0.50-1.21). In patients with a low clinical risk but high genetic risk, the 5-year DMFS was 95% to 96% with or without chemotherapy.

Among the 3,348 high-risk patients, use of the gene assay was associated with a 46% reduction in prescriptions for chemotherapy and 14% reduction for all 6,648 patients in the study, Martine Piccart, MD, PhD, of the Jules Bordet Institute in Brussels, Belgium, said at the American Association for Cancer Research (AACR) meeting.

Acknowledging the 22% reduction in the hazard ratio with chemotherapy for the clinical high-risk/genetic low-risk group, Piccart said, “The statistician will tell you that it is not totally impossible that there is a very small chemotherapy benefit, which would translate into a very small absolute benefit that would not justify the risks of chemotherapy.”

In the U.S., breast cancer specialists have access to MammaPrint and the OncoType DXgene test, the latter being more widely used. AACR President Nancy E. Davidson, MD, of the University of Pittsburgh, declined to speculate as to whether the findings could influence clinical practice. Instead, she said the data will help inform decision making for clinicians and their patients.

Piccart said the key issue is that breast cancer patients should have access to a validated gene test to help with clinical decision making.

Three decades of clinical trial results have shown that adjuvant chemotherapy improves survival in early breast cancer by 2% to 12% in absolute terms. However, effective therapy confers its own risks, including secondary cancers, cardiac toxicity, early menopause, and an adverse impact on cognitive function. Chemotherapy also adds to the cost of care

One approach to maximize the benefits and minimize the risks of chemotherapy is to use clinical/pathologic risk criteria, such as a freely available prognostic model validated more than a decade ago as a decision-making aid for adjuvant chemotherapy. More recently, tumor gene expression profiling has added another method to define breast cancer risk.

Whether tumor genomics profiling could improve risk prediction versus clinicopathologic criteria remained unproven, leading to the international MINDACT (Microarray In Node Negative Disease May Avoid Chemotherapy) trial to evaluate the strategies in clinical practice.

“Our hypothesis was that the MammaPrint genomic assay would outperform the clinical criteria by reducing the prescription of adjuvant chemotherapy without impairing patient outcomes,” said Piccart.

From 2007 to 2011, investigators in nine countries enrolled 6,693 who had undergone surgery for early breast cancer. Initially, only patients with node-negative disease status were eligible, but the protocol was subsequently amended to allow women with as many as three positive nodes after the gene assay was validated in such patients.

Every patient was evaluated by the clinicopathologic prognostic model and the genetic test. By consensus, investigators defined “clinical low risk” as a patient who had an estimated 10-year disease-specific survival of 88% without chemotherapy or endocrine therapy for estrogen receptor (ER)-positive disease or 92% for ER-negative disease.

Patients who had low-risk disease by both methods did not receive chemotherapy, and all patients who were high risk by both methods did get chemotherapy. The 3,348 patients with discordant results (high by one method, low by the other) formed the focal point of the investigation.

The patients with discordant results received radiation therapy and were randomized clinical or genetic risk assessment to decide whether they would receive adjuvant chemotherapy. The primary endpoint was 5-year DMFS.

The primary analysis involved the patients with high clinical risk/low genetic risk, randomized to no adjuvant chemotherapy. To reject the null hypothesis, the lower boundary of 95% confidence intervals (CI) had to exceed 92.0%

The patients with discordant results had a median age of 55; 80% had node-negative disease, 58% had T1 tumors, 88% had ER-positive tumors, and 10% had HER2-positive tumors.

After a median follow-up of 5 years, DMFS was 97.6% for patients who had low-risk disease by clinical/pathologic factors and the gene test versus 90.6% for those who had high-risk disease by both criteria. Piccart said women with high-risk disease tended to have larger tumors and were more likely to have node-positive and HER2-positive disease.

For the primary analysis, women with high clinical/low genetic risk had a 5-year DMFS of 94.7% without chemotherapy (95% CI 92.5%-96.2%), which allowed investigators to reject the null hypothesis.

By intention-to-treat analysis, the clinical high-risk/genetic low-risk group had a 5-year DMFS of 94.4% without chemotherapy and 95.9% with chemotherapy. For the clinical low-risk/genetic high-risk group, 5-year DMFS was 95.8% with chemotherapy and 95.0% without.