Proposed new clinicopathological surrogate definitions of luminal A and luminal B (HER2-negative) intrinsic breast cancer subtypes

Abstract

Introduction: The St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013 recognized substantial progress in the pathological characterization of breast cancer subtypes. A useful surrogate definition was developed to distinguish luminal A-like breast cancer from luminal B-like disease based on a combination of estrogen receptor (ER), progesterone receptor (PgR) and Ki-67 status, without a requirement for molecular diagnostics. Differences depend upon the choice of the threshold value for Ki-67 and the requirement for substantial PgR positivity. We aimed to verify the suitability of the new surrogate definitions of luminal subtypes in terms of distant disease control in a large series of patients.

Methods: We studied 9,415 women with a median follow-up of 8.1 years who (1) had ER-positive, human epidermal growth factor receptor 2 (HER2)-negative early breast cancer and (2) had undergone surgery at the European Institute of Oncology between 1994 and 2006. We evaluated distant disease-free survival of patients with "low" (<14%), "intermediate" (14% to 19%) or "high" (≥20%) Ki-67 positivity stratified by PgR expression (negative or low versus high). We calculated the cumulative incidence of distant events, considered competing events and performed multivariable analysis adjusted for pathologic tumor stage, pathologic node stage, tumor grade, peritumoral vascular invasion and menopausal status.

Results: Lack of substantial PgR positivity was associated with poorer outcomes only for patients with an intermediate Ki-67 level (P<0.001). The 4,890 patients (51.9%) with low Ki-67 level (any PgR expression level) or with intermediate Ki-67 level but substantial PgR positivity had comparably good outcomes and thus may represent a most advantageous grouping of those with luminal A-like disease.

Conclusions: The updated pathological definition of intrinsic molecular subtypes may maximize the number of patients classified as having the luminal A-like intrinsic subtype of breast cancer and for whom the use of cytotoxic drugs could mostly be avoided.

Figure 1

Patient distribution according to molecular characteristics and St Gallen 2011 and 2013 intrinsic molecular subtype definitions. The threshold of ≥14% for Ki-67 in the 2011 definition was derived from comparison with gene array data as a prognostic factor [3,4], whereas the threshold of ≥20% in the 2013 definition was approved by the majority of the expert consensus panel [1]. HER2, Human epidermal growth factor receptor 2; PgR, Progesterone receptor.

Figure 2

Distant disease-free survival according to Ki-67 and progesterone receptor expression levels. (a) Patients with low Ki-67 levels (<14%). (b) Patients with intermediate Ki-67 expression (14% to 19%). (c) Patients with high Ki-67 expression (≥20%). (d) Number of patients and 10-year cumulative incidence. (e) Multivariable analysis. PgR, Progesterone receptor. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using a multivariable Cox proportional hazards regression model adjusted for pathologic tumor stage, pathologic node stage, tumor grade, peritumoral vascular invasion and menopausal status.

Figure 3

Distant disease-free survival according to our proposed intrinsic molecular subtypes. Hazards ratios (HRs) and 95% confidence intervals (CIs) were calculated using a multivariable Cox proportional hazards regression model adjusted for pathologic tumor stage, pathologic node stage, tumor grade, peritumoral vascular invasion and menopausal status.

Figure 4

Distant disease-free survival according to our proposed new definitions of intrinsic molecular subtypes and tumor grades. Hazard ratios and 95% confidence intervals (CIs) were calculated using a multivariable Cox proportional hazards regression model adjusted for pathologic tumor stage, pathologic node stage, tumor grade, peritumoral vascular invasion and menopausal status. P = 0.33 for the interaction between tumor grade and new intrinsic molecular subtype. The percentages of patients with luminal A–like breast cancer who received adjuvant chemotherapy were 9.5% (G1), 24.73% (G2) and 42.0% (G3). Similarly, 21.3%, 38.7% and 60.5% of patients with G1, G2 and G3, luminal B–like breast cancer, respectively, received adjuvant chemotherapy.