Identification of Early-Stage Breast Cancer with a Minimal Risk of Recurrence by the Breast Cancer Index.

PURPOSE

This study assessed the prognostic ability of the breast cancer index (BCI) to identify patients at a minimal risk (<5%) of 10-year distant recurrence (DR) who are unlikely to benefit from adjuvant endocrine therapy.

EXPERIMENTAL DESIGN

This prospective translational study included postmenopausal patients with early-stage, hormone receptor-positive N0 breast cancer from the Stockholm (STO-3) trial who underwent surgery alone ("untreated") or surgery plus adjuvant tamoxifen ("treated") and from the Netherlands Cancer Registry (surgery alone). The primary endpoint was time to DR. An adjusted BCI model with an additional cutpoint was developed, which stratified patients into four prognostic risk groups.

RESULTS

Across cohorts, 16% to 22% of patients were classified as minimal risk of 10-year DR. In the Stockholm untreated cohort (n = 283), risks in the minimal-, low-, intermediate-, and high-risk groups were 2.3%, 15.5% [hazard ratio, 4.71 (95% confidence interval, 1.09-20.29) vs. minimal risk], 19.8% [6.97 (1.61-30.18)], and 35.9% [13.21 (3.07-56.76)], respectively (P < 0.001). In the Stockholm treated cohort (n = 317), risks were 4.3%, 5.0% [1.16 (0.35-3.85)], 11.7% [2.45 (0.74-8.14)], and 21.1% [5.27 (1.72-16.16); P < 0.001]. In the Netherlands Cancer Registry cohort (n = 1245), risks were 4.5%, 7.5% [subdistribution hazard ratio, 1.67 (95% confidence interval, 0.81-3.45)], 10.3% [2.40 (1.14-5.03)], and 13.1% [3.13 (1.50-6.55); P = 0.005]. BCI risk scores provided additional independent information over standard prognostic factors (likelihood ratio, χ2 = 7.98; P = 0.004).

CONCLUSIONS

The adjusted BCI model identified women with early-stage, hormone receptor-positive N0 breast cancer at a minimal risk of DR who may consider de-escalating adjuvant endocrine therapy.