Using real-world data from the FLEX registry, MammaPrint, a 70-gene signature for risk of distant recurrence, was found to be predictive of chemotherapy benefit and distant recurrence-free interval prognosis in early-stage breast cancer.

“The MINDACT trial established the ‘negative predictive value’ of MammaPrint, but it wasn’t designed or powered to answer whether MammaPrint genomic high-risk results would predict chemotherapy benefit—its ‘positive predictive value,’” corresponding author William Audeh, MD, MS, Chief Medical Officer at Agendia, a molecular diagnostics company focused on breast cancer, told MedPage Today.

“To answer this question,” Dr. Audeh says, “we studied patients who had been enrolled in our prospective FLEX study—the largest and most diverse prospective real-world data cohort in early breast cancer, involving more than 20,000 patients—who had been treated with chemotherapy or endocrine therapy alone. We analyzed MammaPrint genomic risk in addition to clinical features, the type of therapy given, and 5-year outcomes, such as freedom from metastasis.”

Study design and patient characteristics

This analysis of the FLEX trial used data from 1407 patients with HR-positive/HER2-negative early-stage breast cancer who received adjuvant endocrine therapy alone (n=859) or endocrine therapy with chemotherapy (n=548).¹ Patients were diagnosed at 93 institutions in the U.S. between 2012 and 2021, with a median follow-up of 5 years after diagnosis. Propensity-score matching was used to balance tumor stage, lymph node status, and menopausal status between the 2 cohorts.

The final analysis included 1002 propensity-score matched patients (501 in each treatment group), with an average age of 59 years. Roughly 7 out of 10 (70.1%) were postmenopausal, and 79.2% were White. MammaPrint ultra-low cancer was reported in 12.2% of patients, low in 35.8%, high 1 (H1) in 41.1%, and high 2 (H2), a distinct, higher-risk category than H1, in 10.9%.

Patients receiving endocrine therapy alone were more likely to be White (85.4%) and have MammaPrint low-risk tumors (61.1%), and were less likely to have grade 3 tumors (6.4%). Patients receiving both endocrine therapy and chemotherapy were less likely to be White (73.1%) and have MammaPrint low-risk tumors (10.6%) but were more likely to have grade 3 tumors (29.3%).

There were significant differences in the distribution of tumor stage (P=.026), grade (P<.001), and menopausal status (P<.001) across the MammaPrint Index. A larger proportion of pre/perimenopausal patients had low-risk MammaPrint Index values compared with postmenopausal patients (27.0% versus 19.5%, respectively). Cancers classified as MammaPrint high risk were more likely to be grade 3. Of grade 1 cancers, 27.1% were either H1 or H2 risk, and 10.6% of grade 3 cancers were ultra-low or low risk. Grade 2 cancers were distributed evenly between the low-risk and high-risk subgroups (48.7% and 51.3%, respectively).

Predicting chemotherapy benefit 

Regression analysis revealed that the MammaPrint Index was predictive of DRFI outcome for the endocrine therapy group (R2=0.99, P<.001) and the endocrine therapy plus chemotherapy group (R2=0.90, P<.001). The risk of a DRFI event rose significantly as MammaPrint Index risk increased among patients in the endocrine therapy only group.

In this same group, patients with ultra-low-risk cancers had a 5-year DRFI event risk ranging from 0.6% to 2.2%. Patients with low-risk cancers had a risk of a DRFI event between 2.2% and 4.5%, whereas risk for those with H1 and H2 tumors ranged from 5.6% to 14.6% and from 14.8% to 24.8%, respectively.

Among patients in the endocrine therapy plus chemotherapy group, those with ultra-low-risk cancers had a 5-year DRFI event risk ranging from 0.1% to 1.0%. Patients with low-risk cancers had a risk of a DRFI event between 1.0% and 2.1%. Risk for those with H1 tumors ranged from 2.6% to 6.4%, rising to 6.5% to 10.6% for those with H2 tumors.

Multivariate analysis revealed a significant interaction between the MammaPrint Index and chemotherapy treatment (hazard ratio [HR] 0.15, 95% confidence interval [CI] 0.02 to 0.97; P=.047), indicating that higher MammaPrint Index risk is predictive of the relative benefit of chemotherapy. There was also a significant association between pre/perimenopausal status and chemotherapy benefit (HR 0.08, 95% CI 0.01 to 0.74; P=.025). However, a number of factors—including age, tumor stage, positive nodal status, and higher-grade tumors—failed to show statistical significance in predicting a benefit from chemotherapy.

Limitations and study conclusions

Limitations of this study include its observational nature, the lack of follow-up beyond 5 years, and the limited data available on the use of ovarian function suppression and cyclin-dependent kinase (CDK) 4/6 inhibitors, which are current standard-of-care treatments for early-stage breast cancer. Despite these limitations, however, the data demonstrate a significant chemotherapy benefit among patients from the FLEX registry with MammaPrint high-risk, HR-positive/HER2-negative early breast cancer.

“We observed that MammaPrint high risk clearly identified those HR-positive breast cancers that would benefit from chemotherapy,” Dr. Audeh noted. “We hope to assess the utility of gene expression profiling with our platform as a predictive biomarker for the benefit of other emerging therapies in breast cancer, such as CDK4/6 inhibitors and selective endocrine degraders.”

Dr. Audeh noted that gene expression profiling can be used to glean other clinically important insights, including endocrine therapy duration and chemotherapy selection.

“This study provides solid evidence of MammaPrint’s chemo-predictive utility and adds to our body of knowledge of MammaPrint as both a predictive and prognostic biomarker in early-stage breast cancer,” Dr. Audeh concluded. “We expect to gain many more valuable insights from FLEX, with the goal of improving outcomes for women with newly diagnosed breast cancer.”

1. Brufsky AM, Hoskins KF, Conter HJ, et al. MammaPrint predicts chemotherapy benefit in HR+HER2- early breast cancer: FLEX Registry real-world data. JNCI Cancer Spectr. 2025;9(5):pkaf079. doi:10.1093/jncics/pkaf079

See article by Erin M. Burns via MedPage Today