Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer

Abstract

Background: Gene expression profiling of breast cancer has identified two biologically distinct estrogen receptor (ER)-positive subtypes of breast cancer: luminal A and luminal B. Luminal B tumors have higher proliferation and poorer prognosis than luminal A tumors. In this study, we developed a clinically practical immunohistochemistry assay to distinguish luminal B from luminal A tumors and investigated its ability to separate tumors according to breast cancer recurrence-free and disease-specific survival.

Methods: Tumors from a cohort of 357 patients with invasive breast carcinomas were subtyped by gene expression profile. Hormone receptor status, HER2 status, and the Ki67 index (percentage of Ki67-positive cancer nuclei) were determined immunohistochemically. Receiver operating characteristic curves were used to determine the Ki67 cut point to distinguish luminal B from luminal A tumors. The prognostic value of the immunohistochemical assignment for breast cancer recurrence-free and disease-specific survival was investigated with an independent tissue microarray series of 4046 breast cancers by use of Kaplan-Meier curves and multivariable Cox regression.

Results: Gene expression profiling classified 101 (28%) of the 357 tumors as luminal A and 69 (19%) as luminal B. The best Ki67 index cut point to distinguish luminal B from luminal A tumors was 13.25%. In an independent cohort of 4046 patients with breast cancer, 2847 had hormone receptor-positive tumors. When HER2 immunohistochemistry and the Ki67 index were used to subtype these 2847 tumors, we classified 1530 (59%, 95% confidence interval [CI] = 57% to 61%) as luminal A, 846 (33%, 95% CI = 31% to 34%) as luminal B, and 222 (9%, 95% CI = 7% to 10%) as luminal-HER2 positive. Luminal B and luminal-HER2-positive breast cancers were statistically significantly associated with poor breast cancer recurrence-free and disease-specific survival in all adjuvant systemic treatment categories. Of particular relevance are women who received tamoxifen as their sole adjuvant systemic therapy, among whom the 10-year breast cancer-specific survival was 79% (95% CI = 76% to 83%) for luminal A, 64% (95% CI = 59% to 70%) for luminal B, and 57% (95% CI = 47% to 69%) for luminal-HER2 subtypes.

Conclusion: Expression of ER, progesterone receptor, and HER2 proteins and the Ki67 index appear to distinguish luminal A from luminal B breast cancer subtypes.

Figure 1

Flow diagram of the strategy for development and implementation of a breast cancer specimen classifier. A) Development with breast cancer specimens from the University of British Columbia and Washington University at St Louis (ie, the UBC-WashU series). B) Implementation or application to breast cancer specimens from the British Columbia Cancer Agency (ie, the BCCA series). ER = estrogen receptor; qRT-PCR = quantitative reverse transcription–polymerase chain reaction; IHC = immunohistochemistry; PR = progesterone receptor.

Figure 2

The x and y axes of ROC curve are true positive rate and false positive rate respectively. True positive rate equals to sensitivity and false positive rate is 1-specificity. Establishment of Ki67 cut point. True positive rate equals to sensitivity and false positive rate is 1-specificity. A) ROC analysis of 144 luminal A and B tumors with Ki67 IHC data to identify luminal B tumors as defined by a 50-gene classifier. Gene expression data for the classifier were obtained by quantitative reverse transcription–polymerase chain reaction. The selected best cut point for the Ki67 index was 13.25%. B) ROC analysis that was confined to 127 luminal A and B tumors with Spearman rank correlation coefficients of more than 0.1. CI = confidence interval; ROC = receiver operating characteristic; IHS = immunohistochemistry.

Figure 3

Univariate survival by breast cancer subtype among 943 patients with lymph node–negative, hormone receptor–positive breast cancer who received no adjuvant systemic therapy. A) Relapse-free survival. B) Breast cancer–specific survival. CI = confidence interval.

Figure 4

Univariate survival by breast cancer subtype among 976 patients with hormone receptor–positive breast cancer who received tamoxifen as their sole adjuvant systemic therapy. A) Relapse-free survival among all 976 patients. B) Breast cancer–specific survival among all 974 patients (two patients with unknown cause of death were excluded). C) Relapse-free survival among 287 patients with lymph node–negative disease. D) Breast cancer–specific survival among 287 patients with lymph node–negative disease. E) Relapse-free survival among 627 patients with lymph node–positive disease. F) Breast cancer–specific survival among 625 patients with lymph node–positive disease (two patients with unknown cause of death were excluded). CI = confidence interval.

Figure 5

Univariate survival by breast cancer subtype among 196 patients with hormone receptor–positive breast cancer who were treated with both tamoxifen and chemotherapy (doxorubicin and cyclophosphamide; fluorouracil, doxorubicin, and cyclophosphamide; or cyclophosphamide, methotrexate, and fluorouracil) as adjuvant systemic treatments. A) Relapse-free survival. B) Breast cancer–specific survival. CI = confidence interval.