Original Study
High Ki-67 Expression and Low Progesterone Receptor Expression Could Independently Lead to a Worse Prognosis for Postmenopausal Patients With Estrogen Receptor-Positive and HER2-Negative Breast Cancer

https://doi.org/10.1016/j.clbc.2014.12.007Get rights and content

Abstract

Background

Accurate classification of luminal A and luminal B characteristics of estrogen receptor (ER)-positive and human epidermal growth factor receptor 2 (HER2)-negative breast cancer is considered clinically important for determining effective adjuvant treatment. Although Ki-67 expression has been identified as an essential constituent for this classification, the role of progesterone receptor (PgR) expression has yet to be fully elucidated. Because PgR expression is influenced by the estrogen milieu, we examined the prognostic significance of PgR expression in immunohistochemical (IHC)-based luminal subtypes defined by Ki-67 expression, taking menopausal status into consideration.

Materials and Methods

We examined 327 surgically removed ER+ and HER2 breast cancer specimens. ER, PgR, and Ki67 expression was determined IHC for semiquantitative measurement. We used 1%, 20%, and 15% as the cutoff value for ER, PgR, and Ki-67, respectively.

Results

Breast cancer with low PgR (≤ 20%) expression was significantly associated with postmenopausal status, a large tumor size, and low ER expression. The low PgR expression subset had significantly worse distant relapse-free survival (DRFS) than the high PgR expression subset (P = .0067). This association was observed consistently in postmenopausal women but not in the premenopausal women. Multivariate analysis demonstrated that high Ki-67 expression (hazard ratio [HR], 3.80; 95% confidence interval [CI], 1.57-10.58; P = .003) and low PgR expression (HR, 2.54; 95% CI, 1.08-6.40; P = .038) were significant independent factors affecting DRFS.

Conclusion

Low PgR expression was independently associated with a poorer prognosis for ER+ and HER2 breast cancer. Determination of PgR expression combined with that of Ki-67 could thus improve the accuracy of IHC-based classification of luminal A and luminal B breast cancer, especially for postmenopausal women.

Introduction

The identification of prognostic factors for estrogen receptor (ER)-positive and human epidermal growth factor receptor 2 (HER2)-negative breast cancer is an important issue for the accurate determination of the indications for chemotherapy. ER+ and HER2 breast cancer can be divided into luminal A and luminal B subtypes according to gene expression profiling.1 Because luminal A cancers are considered to be predominantly estrogen dependent and to possess low proliferative activity and luminal B cancers feature greater proliferative capability and are likely to respond to chemotherapy,2, 3 classification according to these intrinsic subtypes could be useful. In daily clinical practice, Ki-67 expression is used for classification of these subtypes, as reported by Cheang et al,4 and the indications for chemotherapy are usually determined by the Ki-67 expression levels.5 However, because the optimal cutoff point for the Ki-67 labeling index, defined as 13.25% by a previous study, for the identification of luminal B cancers had a sensitivity of only 72% (95% confidence interval [CI], 59%-82%) and specificity of 77% (95% CI, 67%-85%),4 a more exact classification method for use in daily clinical practice is warranted.

The progesterone receptor (PgR) is a downstream molecule induced by transcriptional activation of estrogen signaling.6 Downregulation of PgR can thus be induced by estradiol deficiency but also by cross-talk with growth factor signaling.7 Purdie et al8 examined the prognostic significance of PgR in a population-based study. They found that the PgR groups had significantly poorer disease-free survival (DFS), even for patients with ER+ treated with endocrine therapies (hazard ratio [HR], 1.722; 95% CI, 1.174-2.524; P = .005).8 The associations between PgR negativity and a poorer prognosis were consistently recognized, irrespective of lymph node status or the use of adjuvant chemotherapy. Similarly, quantitatively low PgR expression was significantly associated with a poorer prognosis for patients treated with anastrozole and/or tamoxifen.9 These findings have clearly indicated the prognostic significance of PgR expression in an adjuvant setting.

The observation that most ER+ and PgR cancers are likely to belong to the luminal B subtype10 suggests that a thoroughly validated PgR measurement suitable for general use can be expected to improve the efficacy of discrimination between luminal A and B subtypes. Prat et al11 reported that those with immunohistochemical (IHC)-based luminal A tumors (Ki-67 < 14%) in the low PgR group (≤ 20%) had significantly poorer DFS than did those in the high PgR group (> 20%). Their data also indicated that 61 (51.3%) of 119 IHC-luminal A cancers with PgR expression of ≤ 20% were within the intrinsic luminal B cancer classification, and 25 (30.9%) of 81 IHC-determined luminal B cancers in the PgR > 20% group were classified as intrinsic luminal A cancers. These findings seem to suggest that the clinically used Ki-67–based IHC classification of luminal subtypes is inaccurate and that semiquantitative measurement of PgR expression can improve the IHC-based classification. Although the prognostic significance of PgR has already been identified in the case of IHC-detected luminal A cancers, it has yet to be determined for IHC-detected luminal B cancers. In addition, because ER+ breast cancers are affected by the estrogen milieu, the significance of PgR expression in pre- and postmenopausal women should be investigated separately.

The aim of the present study was to evaluate the prognostic significance of semiquantitative measurement of PgR expression in the IHC-based luminal subtype of breast cancers, defined in terms of Ki-67 expression, taking menopausal status into consideration.

Section snippets

Patient Eligibility

For the present study, 327 consecutive patients with ER+ and HER2 invasive breast cancer treated with mastectomy or breast-conserving surgery at the Hyogo College of Medicine from January 2006 to June 2013 were recruited. All patients had been pathologically diagnosed with breast cancer, and noninvasive carcinoma or patients who had received chemotherapy or endocrine therapy before surgery were not included. The histologic classification and nuclear grade were decided according to the Japanese

Relationship Between Clinicopathologic Characteristics and Ki-67 and PgR Expression

We examined 175 (53.5%) breast cancer cases with low and 152 (46.5%) with high Ki-67 expression and 128 (39.1%) with low PgR and 199 (60.9%) with high PgR expression in the present study. A comparison of the clinicopathologic characteristics of breast cancer in the low and high Ki-67 and low and high PgR subsets is presented in Table 1. Large cancers (> 2 cm) were frequently identified in both high Ki-67 and low PgR subsets (P = .026 and P = .027, respectively). Significantly fewer cases of

Discussion

The findings of our study showed that Ki-67 and PgR expression were independently associated with the prognosis of patients with ER+ and HER2 breast cancer. Consistent with the report by Prat et al,11 patients with low PgR expression (≤ 20%) had significantly worse DRFS than those with high PgR expression (> 20%). Because a similar poorer prognosis for those with low PgR cancer has apparently been established for those with low Ki-67 (≤ 15%, IHC luminal A) and high Ki-67 (> 15%, IHC luminal B)

Conclusion

We were able to demonstrate that low PgR expression is associated with a poorer prognosis for ER+ and HER2 breast cancer. Because PgR was identified as a prognostic factor independent of Ki-67, the determination of PgR expression, combined with that of Ki-67, might thus improve the accuracy of the IHC-based classification of luminal A and luminal B breast cancer, especially for postmenopausal patients.

Disclosure

The authors have stated that they have no conflicts of interest.

Acknowledgments

The present study was supported by the grant of Hyogo College of Medicine.

References (24)

  • C. Creighton et al.

    Molecular profiles of progesterone receptor loss in human breast tumors

    Breast Cancer Res Treat

    (2009)
  • A. Prat et al.

    Prognostic significance of progesterone receptor-positive tumor cells within immunohistochemically defined luminal A breast cancer

    J Clin Oncol

    (2013)
  • Cited by (0)

    View full text