[10] Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal, node-positive, ER-positive breast cancer (S8814,INT0100).

Albain K, Barlow W, Shak S, Hortobagyi G, Livingston R, Yeh I, Ravdin P, Yoshizawa C, Baehner F, Davidson N, Sledge G, Winer E, Hudis C, Ingle J, Perez E, Pritchard K, Shepherd L, Allred C, Osborne K, Hayes D. Southwest Oncology Group and The Breast Cancer Intergroup of NA, San Antonio, TX

Objectives: The 21-gene Recurrence Score assay (RS) is prognostic for women with node-negative, ER-positive breast cancer (BC) treated with tamoxifen (T) alone and a high RS predicts a large, additional chemotherapy benefit. There are no data in a node-positive population with a T-alone control. The 2 primary objectives of this analysis were to determine if the RS 1) provides prognostic data for DFS in the T-alone control arm and 2) predicts a group that does not benefit from chemotherapy followed by T, despite positive nodes.
Patients and Methods: The phase III trial S8814 for postmenopausal women with node-positive, ER-positive BC showed CAF added DFS and OS benefit to T at 10 years, especially if T followed CAF (CAF-T). Optional specimen banking yielded tumor blocks or unstained slides for RT-PCR analyses of the 21 genes for the RS by Genomic Health, Inc. Analyses for DFS and OS were conducted by the SWOG Statistical Center using the T and CAF-T arms of S8814.
Results: 45% of S8814 participants provided specimens, with RNA sufficient for RT-PCR on 367 patients (T, 148; CAF-T, 219). The RS risk distribution was 40% low (<18), 28% intermediate (18-30), and 32% high (31). The RS was prognostic for DFS in the T-alone arm (p=0.006), with similar effects in the 1-3 and 4+ nodal subsets and for OS. There was no apparent DFS or OS benefit to CAF-T vs T in the low RS group, whereas there was a large CAF benefit in the high RS subset. Ten-year Kaplan-Meier DFS estimates (95% CI) and p-values for stratified logrank tests for T vs CAF-T in the 3 RS groups were (1) Low: 60% (40%-76%) vs. 64% (50%-75%), p=0.97; (2) Intermediate: 49% (32%-63%) vs. 63% (48%-74%), p=0.48; and (3) High: 43% (28%-57%) vs. 55% (40%-67%), p=0.03. In a time-dependent Cox model (employed due to non-proportional hazards), continuous RS score was prognostic in T alone in the first 5 years (p<0.001), but not subsequently (p=0.80), adjusting for nodes. For prediction of CAF benefit, the RS by treatment interaction was significant in the first 5 years for DFS (p=0.029), but not after 5 years (p=0.58), with nodal status strongly prognostic for both time periods. Results were similar for OS. Although the effects of RS on hazard rates were concentrated during the first 5 years, cumulative effects on DFS and OS were still observed at 10 years.
Conclusions: The RS is prognostic for T-treated patients with positive nodes. It is predictive of added CAF benefit in those patients whose tumors have a high RS. A low RS may define a group of women with positive nodes who do not appear to benefit from anthracycline-based adjuvant chemotherapy.

Thursday, December 13, 2007 9:15 AM

General Session 1 (9:30 AM-10:30 AM)