21-gene Recurrence Score Does Not Predict Paclitaxel Benefit  

Get Permission

The 21-gene recurrence score significantly predicted the risk of recurrence and death in node-positive, estrogen receptor–positive patients treated with adjuvant chemoendocrine therapy, but it did not predict benefit from the addition of paclitaxel to the regimen in a subset of patients from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-28 trial,1 according to Eleftherios (Terry) P. Mamounas, MD, MPH, FACS, Professor of Surgery at Northeastern Ohio Universities College of Medicine and Medical Director of the Comprehensive Breast Program at MD Anderson Cancer Center, Orlando, Florida, and Chairman of the Breast Committee of the NSABP Operations and Biostatistical Center, Pittsburgh, Pennsylvania.

Study Goals

The goal of this substudy was to evaluate the recurrence score as a predictor of the risk of local-regional recurrence in node-positive, estrogen receptor–positive patients treated with adjuvant endocrine therapy and chemotherapy. The secondary goal was to evaluate recurrence score as a prognostic factor of outcome and as a predictive factor of benefit from the addition of paclitaxel (T) to doxorubicin/cyclophosphamide (AC).

NSABP B-28 compared AC vs AC-T in 3,060 node-positive patients.2 After 5 years, the addition of paclitaxel significantly reduced the annual hazard for disease-free survival events by 17% in the entire population (P = .006).

The current study subgroup included 1,065 estrogen receptor–positive, tamoxifen-treated patients for whom recurrence score was generated, followed for a median of 11.2 years. The recurrence score distribution in this cohort was low in 36%, intermediate in 34% and high in 30%. Older patients and patients with small tumors were significantly more likely to have a low recurrence score.

The recurrence score was significantly correlated with disease-free survival, distant recurrence-free survival, breast cancer–specific survival, and overall survival, and was prognostic regardless of the number of positive nodes (reported at the 2012 Breast Cancer Symposium).3 The ability to predict local-regional recurrence will be presented at the Society for Surgical Oncology 2013 meeting.

Minor Paclitaxel Benefit

In San Antonio, Dr. Mamounas reported that the recurrence score failed to predict the benefit of adding paclitaxel, which offered minor benefit in this study population.

“The overall benefit from paclitaxel in the study cohort was small, and there was low power to detect treatment-by-[recurrence score] interaction,” he said. “We saw this first in the whole B-28 estrogen receptor–positive subgroup (n = 2,007) and then in the subgroup with [recurrence score] data (n = 1,065).”

The hazard ratios for endpoints ranged from 0.86 to 0.90 for the whole B-28 subgroup and from 0.87 to 0.89 for the recurrence score analysis. While some of the risk reduction was statistically significant in the overall subgroup, the differences were small and were similar to the non–statistically significant differences in the recurrence score analysis. Absolute differences in disease-free survival, distant relapse-free interval, and overall survival ranged from 3% to 5% with the addition of paclitaxel, and were not statistically significant, though these were small numbers of patients, he said.

Recurrence Score Subgroups

By recurrence score group, there were no significant differences in any endpoints, although a nonsignificant trend for benefit was observed in patients with intermediate recurrence score and high recurrence score. For overall survival, patients with intermediate recurrence score who received paclitaxel had a nonsignificant 26% reduction in risk (P = .12).

“For all three endpoints, and for the low, intermediate and high [recurrence score] groups, the interaction P values [for treatment effect] were not significant,” he reported. “In some clinicopathologic categories, significant benefit was observed for adding paclitaxel, but again, the test of interaction was not statistically significant.”

An exploratory analysis of the HER2-negative/equivocal subset (n = 937, by HER2 gene expression in the 21-gene panel assay) showed that the prognostic effect of the recurrence score for all endpoints was not driven by HER2 positivity. “The [recurrence score] was not a significant predictor of benefit from the addition of paclitaxel to AC in the HER2-negative/equivocal patients,” he said.

Dr. Mamounas concluded that the recurrence score does not predict benefit from adding paclitaxel to AC, possibly because “the overall benefit from paclitaxel in the study cohort was small and there was low power to detect treatment-by-[recurrence score] interaction.” ■

Disclosure: Dr. Mamounas has served as a paid consultant and speaker for Genomic Health Inc.


1. Mamounas EP, Tang G, Paik S, et al: Association between the 21-gene recurrence score and benefit from addition of adjuvant paclitaxel in node-positive, ER-positive breast cancer patients: Results from NSABP B-28. 2012 San Antonio Breast Cancer Symposium. Abstract S1-10. Presented December 5, 2012.

2. Mamounas EP, Bryant J, Lembersky B, et al: Paclitaxel after doxorubicin plus cyclophosphamide as adjuvant chemotherapy for node-positive breast cancer: Results from NSABP B-28. J Clin Oncol 23:3686-3696, 2005.

3. Mamounas EP, Tang G, Paik S, et al: Prognostic impact of the 21-gene recurrence score on disease-free and overall survival of node-positive, ER-positive patients with breast cancer treated with adjuvant chemotherapy: Results from NSABP B-28. 2012 Breast Cancer Symposium. Abstract 1. Presented September 13, 2011.

Related Articles

Expert Point of View: Clifford A. Hudis, MD and Daniel F. Hayes, MD

Clifford A. Hudis, MD, Chief of the Breast Cancer Medicine Service at Memorial Sloan-Kettering Cancer Center, New York, commented after the presentation, “It strikes me that these findings are parallel to those shown with PAM50 by Liu et al at this meeting.” In that study,1 based on the Cancer and...