Breast Cancer Molecular Subtypes Defined by ER/PR and HER2 Status: Association with Clinicopathologic Parameters in Ivorian Patients.

BACKGROUND
Breast cancer is a heterogeneous disease that represents a major public health problem. The immunohistochemical determination of breast cancer subtypes with regard to estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor (HER2) status can contribute to improved selection of therapy and patientcare. The purpose of this study was to determine the prevalence of the molecular breast cancer subtypes and to assess their associations with classical clinicopathologic parameters for better therapeutic decisions in women with breast cancer in the Ivory Coast.


MATERIALS AND METHODS
Formalin- fixed and paraffin-embedded blocks of patients diagnosed with primary breast carcinoma were subjected to immunohistochemical assay for the assessment of ER/RP and HER2 expression. The one-way analysis of variance evaluated the difference between breast cancer subtypes and mean age of patients. The Chi-square Test was used to compare standard clinicopathologic prognostic parameters with tumor subtypes.


RESULTS
Among 302 patients, 57% were premenopausal and 43% were postmenopausal. The invasive ductal carcinoma not otherwise specified (IDC NOS) (82.8%) was the most frequent histological type, and the tumor grade 2 (56%) was predominant followed by grade 3 (20.9%). The proportion of positivity of ER, PR, and HER2 was 56%, 49%, and 15.6%, respectively. Half of patients of this study (51.6%) had luminal A breast tumor type followed by TN (32.1%). Other subtypes were luminal B (10.1% ) and non-luminal HER2+ (6.3%).


CONCLUSIONS
The findings of the present study are in line with the literature and should assist in management of breast cancer in our country.


Introduction
Breast cancer is the most common malignant tumor in women, and its mortality declines gradually due to advanced diagnostic and therapeutic methods in the developed countries. Nevertheless, this cancer-related mortality remains very high in the developing countries because of late diagnosis associated with a delay in treatment (Ferlay et al., 2013;Stewart and Wild, 2014). Breast cancer is a heterogeneous disease displaying clinical, pathological, and molecular varieties with various prognostic and therapeutic implications (Ronnov-Jessen et al., 1996;Perou et al., 2000;Sorlie et al., 2001;Geyer et al., 2009).

Breast Cancer Molecular Subtypes Defined by ER/PR and HER2 Status: Association with Clinicopathologic Parameters in Ivorian Patients
Ahoua B Effi 1 , N'guiessan A Aman 1 *, Baumaney S Koui 2 , Kouadio D Koffi 1 , Zie C Traore 1 , Mohamed Kouyate 2 Genetic testing of breast cancer markers, such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) is not only very expensive but also exhibits technical difficulties. Therefore, the genomic classification cannot be applied in routine clinical practice. However, several studies have shown that the molecular classification based on gene expression of ER, PR, and HER2 is markedly correlated with immunohistochemical testing of these markers (Perou et al., 2000;Sorlie et al., 2001;Carey et al., 2006;Parise et al., 2009;Bauer et al., 2010;Paris and Caggiano, 2014). Despite the advanced genetic diagnostic methods, the management of breast cancer still depends on standard clinicopathologic prognostic factors associated with immunohistochemical (IHC) expression of ER/PR and HER2 (Geyer et al., 2009).
In Ivory Coast, breast cancer (34.2%) represents the first cancer in women followed by cervical cancer (Ferlay et al., 2013). However, to the best of our knowledge, there are no data on the molecular breast cancer subtypes. Initially, the decision of hormone therapy was based on the breast cancer patients' menopausal status. Indeed, it was hypothesized that postmenopausal subjects would likely be expressed the ER/PR , and therefore, would benefit from systematic adjuvant tamoxifen, which remained ineffective without knowing the status of ER/PR and HER2 of these patients. Since 2013, the routine ER/PR and HER2 immunohistochemical testing helps improve clinical, prognostic, and therapeutic outcomes in women with breast cancer in Ivory Coast. Thus, it appears urgent to carry out the present study to determine the prevalence of breast cancer molecular subtypes and their associations with the classical clinicopathologic parameters in the Ivorian patients.

Patients
The prospective study was conducted between November 2013 and June 2015, including 302 patients diagnosed with primitive and invasive breast carcinomas at the Central Laboratory in Abidjan, Ivory Coast. The histological diagnosis was performed upon formalinfixed and paraffin-embedded breast tissue blocks sampled from biopsies and mastectomies. On each sample, the histological type and the Nottingham grade of the tumor were determined according to the criteria of Elston and Ellis (Elston and Ellis, 1991). The parameters of the study were standard clinicopathologic prognostic parameters (age, menopausal status, histological type, and tumor grade) and the status of ER/PR and HER2. Formalinfixed and paraffin-embedded blocks of tissue breast were subjected to the IHC assay.

Immunohistochemical analysis
Formalin-fixed and paraffin-embedded specimens were cut into 3-µm sections and mounted on positively charged slides. The slides were deparaffinized and heated in the drying oven BINDER ® (BINDER Company, Tuttlingen, Germany) for at least 12 hours at 60 O C to unmask the antigenic sites. The sections were stained using the Ventana Benchmark ® GX in automatic mode (Ventana Medical Systems Inc., Tucson, AZ, USA) for the assessment of ER/PR and HER2 expression. The antibody clones were monoclonal, developed in rats, consisted of SP1 for the ER, 1E2 for the PR, and 4B5 for the HER2, and manufactured by Ventana Medical Systems, Inc.

ER/PR and HER2 staining assessment
The visual analysis through the optic microscope allowed to quantify the percentage of tumor cells showing a nuclear immunostaining for ER and PR (range: 0-100%). Breast tissue sections were considered positive for ER and PR if ≥ 1% of tumor cells showed a positive nuclear staining in accordance with the recommendations of the American Society of Clinical Oncology/ College of American Pathologists (Hammond et al., 2010). The HER2 scoring is based on the intensity of tumor cell membrane staining (on a scale of 0 to 3+) and the percentage of positive tumor cells. Briefly, HER2 is negative for score 0 or 1+, intermediate for score 2+, and positive for score 3+ (Wolff et al., 2013). The HER2 intermediate cases were not examined because the fluorescent in situ hybridization was not available in our laboratory. The percentages of ER/ PR and the score of HER2 were reviewed independently by two pathologists. In this study, the expression of Ki-67, cytokeratin 5/6 (CK 5/6) and human epidermal growth factor receptor 1 (HER1) were not examined. The triple negative consists of the basal-like type (ER-/PR-/ HER2-/ CK 5/6+ / HER1+) and normal-like type (all five markers are negative) (Nielsen et al., 2004). Herein, the breast cancer subtypes were defined according to the IHC expression of ER/PR and HER2 as follows (Carey et al., 2006;Parise et al., 2009;Bauer et al., 2010): Luminal

Statistical analysis
Data were collected in an Excel database from Windows 8 (Microsoft Corporation, Redmond, WA, USA). The difference between the breast cancer subtypes defined by ER/PR and HER2 expression and the mean age were evaluated by the one-way analysis of variance. The Chi-Square Test was used to analyze associations of standard clinicopathologic prognostic parameters (menopausal status, histological type, and tumor grade) with breast tumor subtypes. A p-value <0.05 was considered statistically significant. The data were reported as frequencies for menopausal status, histological type, tumor grade, ER/PR status, and HER2 expression and as means for patients' age at diagnosis.

Characteristics of patients
The mean age of all patients of this study was 48 ± 10.7 years at the time of diagnosis (extremes: 24-84 years). The Table 1 shows clinicopathologic parameters in women with breast cancer. Among 302 patients, 57% were premenopausal and 43% were postmenopausal. The invasive ductal carcinoma not otherwise specified (IDC NOS) (82.8%) was the most frequent histological type, and the tumor grade 2 (56%) was predominant followed by grade 3 (20.9%). The proportion of positivity of ER, PR, and HER2 was 56%, 49%, and 15.6%, respectively. Half of patients of this study (51.6%) had luminal A breast tumor type followed by TN (32.1%). Other subtypes were luminal B (10.1% )and non-luminal HER2+ (6.3%).

Association of standard clinicopathologic factors with subtypes
The Table 2 summarizes the relationship between classical clinicopathologic features and molecular breast cancer subtypes. The mean age difference was not significantly associated with various tumor subtypes (p=0.52). The proportion of the molecular subtypes was linked to the menopausal status (p=0.76), although premenopausal patients displayed more luminal A and triple negative tumor types. The histological type was independent of the tumor molecular subtypes (p = 0.59).
Breast cancer categories defined by ER/PR and HER2 expression were closely correlated with the Nottingham grade (p < 0.00001). The luminal A breast cancer patients had a relatively low grade with 19.6% of grade 1 and 58.1% of grade 2 compared to those of triple negative presenting high tumor grade with 43.5% of grade 2 and 37% of grade 3.

Discussion
This manuscript represents the first study conducted in Ivory Coast, which determines the molecular groups of breast cancer based on the IHC expression of ER/PR and HER2 to guide the best therapeutic strategy for patients and to improve their survival.
In the present study, 56% of women with breast cancer expressed ER. This result is in accordance with the literature data (Adebamowo et al., 2008;Lin et al., 2009;Lowery et al., 2009;Ontilio et al., 2009;Joensuu et al., 2013;Inwald et al., 2013), but is higher than the finding reported in Ivory Coast (40% in small sample size of 22 patients) (Coulibaly et al., 2008), in Ghana (32,1%) (Seshie et al., 2015), in Nigeria/Senegal (24%) (Huo et al., 2009). The high rate of ER positivity in our study confirms that the ER expression is undoubtedly the most significant biomarker in breast cancer. In fact, several studies have demonstrated that breast cancer is a hormone-related disease caused by the mitogenic effects of estrogen and progesterone (Knight et al., 1980;Weigel and Dowsett, 2010). Thus, the positivity of ER provides the index for sensitivity to hormone therapy (Byar et al., 1979;Weigel and Dowsett, 2010).
The geographical variations of breast cancer subtypes proportions could be related to technical issues (Werner et al., 2000), environmental factors (Kwan et al., 2009 ;Trivers et al., 2009;Stark et al., 2010;Kurian et al., 2010), and genetic factors (Lakhani et al., 2002;Malone et al., 2006;Pal et al., 2015). More importantly, the significant rate of the TN in our study requires a particular attention and may result from a defect in preanalytical factors, particularly the duration and the quality of fixation, such as under fixation or prolonged fixation, and thus, causing the epitope loss, as investigated by Werner et al. (2000). Since the practice of IHC is recent in Ivory Coast, an effective collaboration among oncologists, radiologists, and pathologists is needed to improve the breast tissue fixation within recommended time for the therapeutic benefit of patients. Additionally, numerous studies have identified that breast cancer-related risk factors , including poverty, multiparity, early age of first childbirth, short duration of breastfeeding, and black race are strongly associated with the elevated incidence of TN (Carey et al., 2006;Milikan et al., 2008;Kwan et al., 2009;Phipps et al., 2011;Ambrosone et al., 2014). The overrepresentation of TN among Ivorian women could be also illustrated by genetic factors. The predominance of this breast tumor type among the black American women is significantly associated with BRCA1 gene mutations that predispose them to develop this aggressive and resistant molecular subtype to hormone therapy (Lakhani et al., 2002;Malone et al., 2006;Pal et al., 2015). Lakhani et al. have found that breast cancer women associated with BRCA1 mutations are not more likely to express the ER, PR, and HER2 (Lakhani et al., 2002). The remarkable finding of Lakhani et al. underscores the need to conduct an in-depth study not only to assess the association of risk factors with triple negative patients but also to investigate the BRCA1 mutation status among triple negative women in Ivory Coast.
To determine the impact of standard clinicopathologic factors on the molecular breast cancer subtypes in this study, the relationship between these parameters and different tumor groups have been examined in 302 breast carcinomas. The mean age and the menopausal status were independent from molecular subtypes. This lack of correlation was found by several authors (Huo et al., 2009;Tesfamariam and Roy, 2013;Seshie et al., 2015) and could be clarified by the young age of our study population because of a relative short female life expectancy (54 years) in Ivory Coast. Before the advent of the IHC in Ivory Coast, adjuvant hormone therapy, tamoxifen, was systematic in postmenopausal women diagnosed with breast carcinoma. Nevertheless, our present study points out that the choice of tamoxifen should be given regardless the menopausal status for adequate medical care of patients in Ivory Coast. In addition, in our study, histological type was not related to various tumor subtypes, although they contain the majority of the IDC NOS. Our finding differs from result of Weigelt et al. (2008) and Huo et al. (2009). The absence of association of histologic type with cancer subtypes would probably be due to the sample size of our population study.
In this study, the tumor grading was closely correlated to the mammary subtypes. Interestingly, the TN patients displayed a relatively high grade compared to those of luminal groups. Our results are consistent with findings of other investigations (Carey et al., 2006;Huo et al., 2009;Lin et al., 2009;Ontilio et al., 2009;Paris and Caggiano, 2014), and thus, highlighting distinct clinical and therapeutic responses observed in breast cancer subtypes (Pusztai et al., 2003;Ontilio et al., 2009). The most common occurrence of TN in premenopausal women with high tumor grade may substantially explain the increased mortality rate due to breast cancer in the developing countries, particularly in Ivory Coast, where patients have limited access to ill-equipped health infrastructures. Our remarkable observations (high rate of TN in young women under 50 years exhibiting high tumor grade) corroborate with findings of other studies in the black American and the African women, confirming that the TN is strongly associated with unfavorable clinicopathologic prognostic parameters, and therefore, showing the aggressive feature of this subtype (Nielsen et al., 2004;Carey et al., 2006;Trivers et al., 2009).
Our study has some limitations. First, the unavailability of basal-like markers in our routine IHC testing did not allow us to distinguish the TN into basal-like and normallike categories. The differentiation of TN provides specific treatment strategies for normal-like type which has a favorable prognosis compared to basal-like group (Nielsen et al., 2004). Second, since the tumor size and the lymph node status were not considered in our study because of a large number of core needle biopsy breast samples, we were unable to evaluate the correlation of these significant clinicopathologic parameters with molecular breast cancer subtypes.
In conclusion, Breast cancer occurs in the young woman with high tumor grade. The most frequent molecular subtypes are the luminal A and the triple negative. The tumor grade of patients is significantly associated with tumor subtypes. The current study helps reevaluate the systematic use of hormone therapy and guide the therapeutic strategy choice for patients with breast carcinoma in Ivory Coast. The evaluation of basal-like markers (CK 5/6 and HER1) is recommended to distinguish triple negative for an effective therapy. Further studies should be carried out to investigate BRCA1 mutation status in triple negative breast cancer women in Ivory Coast.