Breast cancer is the second leading cause of cancer death in women worldwide, although overall survival (OS) of breast cancer patients has been gradually increased over the last several decades, with advances in cancer therapeutics. Approximately two-thirds of breast cancer are estrogen-dependent cancers, which express estrogen receptor (ER)/progesterone receptor (PR). It is well-established that anti-estrogen therapy, with tamoxifen or aromatase inhibitors (AIs), is effective to reduce recurrence and mortality rate of estrogen-dependent breast cancer. The National Surgical Adjuvant Breast and Bowel Project implemented the B-14 trial, the first randomized clinical trial to assess the efficacy of adjuvant tamoxifen treatment in patients with ER-positive breast cancer and negative axillary lymph nodes. At 5-year follow-up, tamoxifen treatment was found to have a highly statistically significant benefit in disease-free survival (DFS) of all ages [1,2]. Since the release of the first results of the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial, for adjuvant treatment of postmenopausal women with early breast cancer, AIs have widened the field of endocrine therapy in breast cancer and AIs show an enhanced antitumor effect compared to tamoxifen [3,4]. Despite a gradual decrease in death rate from breast cancer with anti-estrogen therapy, resistance to anti- estrogen therapy often develops in some estrogen-dependent tumors [5]. Since the first appearance of tamoxifen and AIs, endocrine- resistant breast cancer remains an active area of research [6,7].

In our previous study, it was demonstrated that an ER/PR-decreased subgroup, among luminal B human epidermal growth factor receptor 2 (HER2)-negative breast cancer, was associated with a higher recurrence rate and shorter survival time than an ER/PR-intact subgroup [8]. The current study attempted to investigate the clinical implication of the decreased expression of ER/PR, in a more homogeneous breast cancer subgroup, identified as HER2-negative, low proliferative, with a Ki-67 below 20%.

Medical records were reviewed, of breast cancer patients who underwent breast surgery and treatment between September 2000 and October 2014 at Sanggye Paik Hospital. Two hundred ninety-seven breast cancer patients, with a low Ki-67 proliferation index (20%) and HER2-negative from final pathologic reports, were identified. Data from 37 patients with stage 0 were excluded and triple-negative breast cancer was excluded, due to the small sample size (n=8). The remaining 252 breast cancer data were divided into two subgroups: strong luminal and weak luminal, according to the strength of hormonal receptor expression. In both ER and PR immunohistochemical (IHC) staining, an Allred score of 7 or 8, was considered as belonging to the strong luminal subgroup and Allred scores between 2 and 6 were assigned to the weak luminal subgroup. Outcome measures included age at diagnosis, tumor size, node metastasis, breast and axilla operative methods, tumor type, tumor-node-metastasis (TNM) stage, histological grade, nuclear grade, tumor number, ER, PR, Ki-67 and bcl-2 expression, recurrent or metastatic characteristics, DFS and OS, of each subgroup. For the total recurrence rate, we compared the number of patients who were diagnosed as any site of metastasis. The patient with multiple metastasis was counted for each site of metastasis. The study was approved by the Institutional Review Board of Inje University Sanggye Paik Hospital (2017-01-012-002) and the informed consent was waived.

Among a total of 252 breast cancer patients, 174 patients (69.0%) were classified as belonging to the strong luminal subgroup and 78 patients (31.0%) to the weak luminal subgroup.

For a patient with newly diagnosed breast cancer, the classification of the patient into either the high- or low-risk population of relapse and the need for adjuvant chemo-, radiation, anti-estrogen, or target therapy are controversial. Although node-negative breast cancer patients are considered as low-risk population for relapse, approximately up to 10% to 30% of these patients relapse within 10 years after initial therapy and the majority of recurrent breast cancer patients eventually die from the disease [14]. Accordingly, abundant clinical studies have been conducted, to identify the prognostic and predictive factors that enable clinicians to estimate the risk of relapse and the drug response of individual breast cancer patients. Risk assessment for relapse, is considered an essential factor in deciding the optimal therapeutic strategy and predicting the prognosis of breast cancer.

Over the last two decades, molecular profiling or multigene assays or IHC staining of ER, PR, HER2, Ki-67 and histologic grade, to identify at least four subtypes of breast cancer, such as luminal A, luminal B, HER2-enriched, and basal-like, have been investigated and performed, to improve the prognostication and risk stratification in breast cancer [15]. Ki-67, a nuclear marker of cell proliferation, is one of the most extensively studied markers in breast cancer [16].

A new cutoff value of Allred score 7 was defined for the strength of ER and PR expression in this research. High levels of ER expression (Allred scores of 7 and 8) ensured the response rate of neoadjuvant endocrine therapy for breast cancer patients in previous studies [17]. From this perspective, we investigated the significance of this value in determining the prognosis of ER/PR positive low proliferative HER2-negative breast cancer.

The Ki-67 cutoff value, to differentiate two luminal A and B categories, has been changing over time. The 2011 Saint Gallen Consensus Meeting, defined low-risk tumors as having a Ki-67 cutoff of <14%. In the 2013 Saint Gallen Conference, the majority of panelists voted that a threshold of 20% was indicative of high Ki-67 status. For the last several years, many investigators proposed that a 20% Ki-67 cutoff, is the best to stratify high-risk patients in luminal breast cancer [12]. In this context, the current study investigated the characteristics of low proliferative HER2-negative breast cancer patients, with a Ki-67 cutoff of 20%.

In many studies including meta-analyses, the expression of Bcl-2 was confirmed as a favorable prognostic factor for breast cancer due to its correlation with ER status [18]. Recently, Dawson et al. [19] reported that Bcl-2 is an independent parameter of a better prognosis for all types of early-stage breast cancer. In multivariable analysis, Ki-67 was also concluded as an independent prognostic indicator, in terms of both DFS and OS for breast cancer [20]. When we analyzed tumor characteristics, according to the strength of ER and PR expression, the weak luminal subgroup was associated with frequent neural invasion, low Bcl-2 expression and a relatively higher Ki-67 (14%–20%) than the strong luminal subgroup. Therefore, the weak luminal subgroup is related to worse prognostic factors relative to the strong luminal subgroup.

Regarding distant metastatic recurrences and locoregional recurrences, we observed a more frequent distant metastasis, particularly to the lung, liver and brain, in the weak luminal subgroup than the strong luminal subgroup. We also observed a survival gap between the strong and weak luminal subgroups. These findings indicate that decreased ER/PR expression, has clinical prognostic implication in the HER2-negative luminal subgroup. Meaningfully, the weak luminal subgroup in this study, also includes cases with an Allred score of 5 and 6, expressing sufficient ER/PR to respond to endocrine therapy. This is the reason we carefully presume that decreased expression of ER/PR itself, might be related to drug resistance or metastatic potency.

The study by Balduzzi et al. [21] concluded tumors presenting low ER/PR expression, had similar outcomes when compared with triple-negative breast cancer. Their result showed no clear effect on survival outcomes with the addition of endocrine therapy, for patients with ER/PR expression of 1% to 10% [21]. The poor response to hormonal therapy, can be considered as a factor in explaining the worse prognosis of the weak compared to the strong luminal subgroup. In contrast, a meta-analysis of randomized trials concluded that even weakly positive ER, substantially beneted from tamoxifen but the proportional ect at a much higher ER was only slightly better [22].

Ono et al. [23] recently reported that PR expression and tumor size, were independent prognostic factors in the luminal A-like subtype and, also, the luminal A/PR-negative subtype had a higher risk of recurrence than the luminal A/PR-positive subtype. Given that the luminal A/PR-negative subtype largely overlapped with the weak luminal subgroup in our study, we carefully suggest that decreased ER/PR expression might be considered a meaningful prognostic factor in low proliferative HER2-negative breast cancer.

The weak luminal subgroup showed a worse prognosis than the strong luminal subgroup, among HER2-negative low proliferative breast cancer patients. Decreased expression of ER/PR itself, could be a clinical prognostic factor in HER2-negative low proliferative breast cancer. Current molecular classification of breast cancer represented as luminal type, is not sufficient for determining the high-risk group and additional gene analysis should be considered for the patients with relatively poor prognostic factors, like low ER/PR expression.