Breast cancer is a complex disease that exhibits variability in clinical manifestation, response to current therapy, and biochemical and histological features among several subgroups. Based on histology, invasive tumors are divided into two categories: histological special types, characterized by specific diagnostic criteria, with invasive lobular carcinoma (ILC) being the most common among them, and invasive carcinoma of no special type (1). The latter type is also known as invasive ductal carcinoma (IDC) and accounts for 70% of cases. It can be described as a breast invasive epithelial neoplasm that fails to fulfill the criteria for any special type, making it a very heterogeneous group of tumors (2).

The fundamental elements for identifying major prognostic markers are the examination of lesion size, axillary lymph node status, nuclear grade, and histological subtype. The lesion's histopathological features reveal many forms of breast tumor biological behavior (3).

The histological categorization of breast cancer is not without flaws, though. Approximately 85% of cases fall into one of the two main categories of IDC or ILC after using such classification, taking into consideration also the subjectivity of the diagnostic criteria. Determining the histologic grade and the immunohistochemical examination of progesterone receptor (PR), human epidermal growth factor 2 (HER-2) and estrogen receptor (ER), and the Ki-67 proliferation index play a crucial role in increasing the diagnostic accuracy among the various types of breast carcinoma because the system is unable to group tumors with an extensive biological range and clinical behavior in the same categories (4).

A better prognosis is linked to the presence of hormone receptors (HR). As a result, patients with progesterone (PR)-positive tumors live longer and are disease-free longer. In the same vein, tumors that are estrogen (ER)-positive are linked to both a better likelihood of responding to hormone therapy and an increased disease-free survival. In contrast, patients who were negative for both ER and PR had a poorer prognosis than those who were negative for only one of the receptors (5). The HER-2 proto-oncogene, which produces a protein that sends signals for the proliferation of epithelial cells and whose expression is frequently elevated in breast cancer, is another significant tumor marker. A more aggressive clinical behavior of the tumor is caused by HER-2 over-expression, and determining the marker status is crucial for identifying cancer types with a poorer prognosis (6,7).

High rates of cell proliferation are typically indicative of highly malignant tumors. Therefore, the assessment of mitotic activity is critical for the diagnosis of breast cancer. In order to achieve this, a monoclonal antibody that recognizes the cell proliferation index Ki-67, a nuclear antigen, is employed. Ki-67 is expressed in cells that are entering the cell cycle and measures the percentage of cell growth, making it possible to identify cancers with a poorer prognosis (8).

The aim of the current study was to determine the histopathological and immunohistochemical characteristics of breast tumors from a reference University Laboratory of Pathology in Athens, Greece.

Study design. This is a retrospective observational study conducted at the First Department of Pathology of National and Kapodistrian University of Athens, Greece. We examined the medical records of the histopathological and immunohistochemical examinations of consecutive breast tumors conducted in the laboratory from January 2014 to December 2018. The study was conducted in line with the Declaration of Helsinki and obtained approval by the Ethics Committee of Medical School of the National and Kapodistrian University of Athens (protocol no. 062/ 14.01.2019).

Inclusion criteria. The study included female patients over 18 years of age, whose medical records regarding both histopathological and immunohistochemical examinations were stored in the archives of the First Department of Pathology of National and Kapodistrian University of Athens, and who agreed to be enrolled in the study by signing of the informed consent form.

Data collection. The following data were collected: age, tumor location, parameters related to histopathological examination, and parameters related to immunohistochemical examination. Regarding histopathological aspects, the following were analyzed: tumor size; histologic grade; number of invaded lymph nodes; presence of positive sentinel lymph node; surgical margins; tumor composition; histopathological classification of the tumor (9) and TNM stage (10). As for immunohistochemical features, the following were examined: presence and intensity of expression of cell proliferation antigen Ki-67; expression of HER-2 oncogene; and intensity of expression and presence of ER and PR; expression of E-cadherin, p53, p63, CK14, CK5/6, androgen receptor (AR) and epidermal growth factor receptor (EGFR).

Immunohistochemical analysis. Immunohistochemistry was carried out using standard procedures in all tested specimens. First, the sections were stained with the following antibodies on a Dako system (Autostainer Link 48, Agilent Dako Pathology Solutions, Santa Clara, CA, USA), according to the manufacturer’s protocol. The sections were stained based on the double-staining protocol of Leica Biosystems in the Bond-III fully automated stainer, with antibodies against estrogen receptor (ER), (clone EP1; Agilent DAKO), progesterone receptor (PR), (clone PgR 636; Agilent DAKO), epidermal growth factor 2 (HER-2) (rabbit polyclonal; Agilent DAKO), cell proliferation index (Ki-67) (clone MIB-1; Agilent DAKO), and E-cadherin, (clone NCH-38; Agilent DAKO). Antigen retrieval was performed at pH 6. The Agilent (Dako) visualization system was used. DAB (3,3-diaminobenzidine) was used as a chromogen and hematoxylin as counterstain.

The histological and immunohistochemical study was performed on tissue pieces fixed in a 10% neutral formalin aqueous solution and embedded in paraffin cubes (Paraplast, Leica Biosystems Inc., Richmond, IL, USA) at 55˚C. The indirect immunoperoxidase method was used in the DAKO EnVision FLEX+ detection system (HRP/DAB, Specificity: Anti-Mouse IgG (H+L) – Agilent DAKO, Cat# K8002).

The Sequenza immunohistochemistry device (Thermoshandon, Runcorn, UK) was used using cover-plates technology, which relies on the capillary effect and ensures staining with the minimum amount of reagent and protection of samples from drying.

All the antibodies for immunohistochemical detection were used according to manufacturer’s protocol.

The use of indirect immunohistochemical methods with peroxidase contributes to strengthening the signal of the immune reaction. Peroxidase activity was visualized by staining with 3,3’, diaminobenzidine tetrahydrochloride chromogen solution (DAB Chromogen, EnVision FLEX, Agilent DAKO), (11).

Histological sections from all samples were observed under a photon microscope. The assessment of immunohistochemical staining was performed independently and blindly by two experienced pathologists.

The staining was evaluated as cytoplasmic and/or membranous, nuclear and/or cytoplasmic, both in the neoplastic tissue and in the adjacent non-neoplastic tissue. Intensity was graded using the following scale: 0 for undetectable staining, 1 for weak intensity, 2 for moderate and 3 for strong. Grading of the percentage of positive cells was performed semi-quantitatively, with the following scale: 0 for immunopositivity in 0-10% of cells, 1 for 10-29%, 2 for 30-59%, and 3 for 60-100%. For each sample, a score was derived from the sum of the intensity and distribution scores, ranging from 0 to 6. Tumors with a final score of 1, 2, 3 were included in the weak expression group, and those with a final score of 4, 5, 6 in the strong expression group.

Determination of positive expression. ER and PR were considered positive when >1% of cells showed positive nuclear expression (12). HER-2 categorization was made according to the proposed criteria of the DAKO protocol (13). p53 was considered positive when >10% of cells showed positive nuclear expression (14). Regarding Ki-67 we considered low expression when ≤15%, moderate when 16-30% and high when >30% of cells showed positive nuclear expression (15). CK14 and CK5/6 were scored positive if any (weak or strong) cytoplasmic and/or membranous invasive carcinoma cell staining was observed (16). E-cadherin expression was determined as positive when scores were ≥2 (17). P63 was scored positive when high intensity staining was present on ≥50% of tumor cells (18). A cutoff point of ≥10% cells with at least weak staining intensity was applied to define protein positivity for EGFR (19). AR expression was determined as positive when scores were ≥2 (20).

Statistical analysis. The Kolmogorov–Smirnov test was used to determine whether the distribution of the variables was normal. All continuous variables had non-normal distribution and are displayed as median (range); categorical variables are presented as absolute numbers (frequency percent). We utilized the Mann–Whitney U-test for non-normally distributed variables with two groups and the Kruskal–Wallis test for those with three groups. Categorical variables were analyzed using Fisher's exact or chi-square tests. The Spearman correlation coefficient was utilized to evaluate associations between continuous variables. p-Values under 0.05 were regarded as significant. The IBM SPSS Statistics version 29.0 (IBM, Armonk, NY, USA) was used for the statistical analysis.

A total of 197 females with a median age of 70 years (range=28-98 years) participated in the study. The median value of maximum tumor size was 2.6 cm (0.4-12 cm). Most tumors were located at the left breast (110, 55.8%). The most frequent histologic type was the ductal carcinoma (70, 35.5%), followed by the lobular carcinoma (49, 24.9%) and the non-specific type (NST) adenocarcinoma (46, 23.4%). Most tumors were histologic grade 2 (106, 53.8%) and most tumors classified at stage ΙΙΑ (65, 33%). Twenty-three patients (11.7%) had in situ carcinoma and 38 patients (19.3%) had triple-negative breast cancer. Three of the 49 lobular carcinomas were in situ carcinomas (6.1%) and 46 of the 49 lobular carcinomas were invasive (93.9%). Nine of the 70 ductal carcinomas were in situ (12.9%) and 61 of the 70 ductal carcinomas were invasive (87.1%). Descriptive characteristics of the study population and the analyzed breast tumor types are summarized in Table I. Regarding the age, we observed a statistically significant difference in the median age of patients between those with lobular and non-lobular breast carcinoma [66 (38-92) years in patients with lobular carcinoma vs. 72 (28-98) years in patients without, p=0.009] (Table II).

Regarding the maximum tumor size, there was a statistically significant difference between the patients with invaded lymph nodes and those without [3.5 (0.9-12) cm vs. 2.2 (0.4-10) cm, respectively, p=0.001], between the patients with positive sentinel lymph nodes and those without [3.5 (0.9-12) cm vs. 2.4 (0.4-12) cm, respectively, p=0.001], between the patients with multifocal tumors and those without [2.9 (0.6-10.5) cm vs. 2 (0.4-12) cm, p=0.007], and between the patients with expression of E-cadherin and those without [2.6 (0.6-7.8) cm vs. 1.9 (0.4-10.5) cm, p=0.015] (Table III).

Additionally, we observed that all metaplastic-squamous carcinomas had histologic grade 3 (p=0.001), and more patients in stage I had histologic grade 1 than in any other stage. Most of the patients with histologic grades 2 and 3 were in stage IIA (p=0.001). Most patients with histologic grade 2 exhibited high expression of ER; most patients with grade 3 exhibited no expression of ER; and all the patients with grade 1 had high ER expression (p=0.001). Most patients with grades 1 and 2 exhibited high expression of PR, whereas most patients with grade 3 had no PR expression (p=0.001). Most patients with grades 1 and 2 had low expression of Ki-67, whereas most patients with grade 3 had high expression of Ki-67 (p=0.001). Moreover, patients with triple-negative cancer presented with grades 2 and 3 at a lower percentage compared to patients without a triple-negative phenotype (p=0.001) (Figure 1, Figure 2, Table IV).

Most patients of all TNM stages had no positive sentinel lymph node (p=0.001), most patients at stages IIA, IIB, and IIIa had multifocal tumors (p=0.024), most patients with papillary carcinoma were at stage I (p=0.029), most patients with metaplastic-squamous carcinoma were at stage IIIA (p=0.003), and most patients at stages I, IIA, IIB, and IIIA had high expression of ER, whereas most patients of stage IIIB had no expression of ER (p=0.001). Most patients at stages I, IIA, and IIB had no expression of E-cadherin, and most patients at stages IIIA and IIIB expressed E-cadherin (p=0.033). Most patients of all stages did not have the triple-negative phenotype, and most of the triple-negative ones were at stage IIA (p=0.016), (Table V).

Most patients with metaplastic-squamous carcinoma had a triple-negative phenotype (p=0.001), and most patients with triple-negative cancer had high expression of Ki-67 (p=0.001) and positive expression of p63 and CK14 (p=0.010 and p=0.001, respectively) (Table VI). In addition, we observed a statistically significant association between the expression of E-cadherin and lymph node involvement (p=0.002).

Moreover, we also found a positive weak correlation between the age and the maximum size of the tumor (Spearman’s rho=0.253, p=0.001, data non-shown). At Table VII, we present the 5-year frequency of breast cancer types in the central laboratory of the First Department of Pathology of National and Kapodistrian University of Athens, from 2014-2018, according to histopathological and immunohistochemical classifications. The frequency of the main types of breast cancer in our study is presented in Figure 3.

The median age of the patients included in our study was 70 years (mean age 69.3±15.19 years). This result differs from other studies, in which the mean age of the patients has been reported to be 53 years, 53.3 years, 57.5 years, and 48.5 years (21-24). These variations might result from the diverse age range of women examined in these studies.

Regarding the histological classification, the most frequent type found in the current study was ductal carcinoma. This finding is in line with findings from other studies (21,23,25). In our study, the frequency of lobular carcinoma was 24.9%. Lobular cancer accounts for 15% of breast cancer cases (26). Our study shares similarities with a study from Brazil in which most patients were diagnosed with ductal breast carcinoma, and the second most frequent type of cancer was the lobular one (27).

In our study, most patients with histologic grade 2 exhibited high expression of ER; most patients with grade 3 exhibited no expression of ER; and all the patients with grade 1 had high ER expression (p=0.001). Most patients in grades 1 and 2 exhibited high expression of PR, while most patients in grade 3 had no PR expression (p=0.001). These results indicate an inverse association between the expression of ER and PR. Similar findings have been demonstrated by Soares et al. (21). Also, these results are comparable to the findings of Dayal et al. (28), who showed that the incidence of histologic grade 3 was greater than 50% in cases where there was no ER expression. However, histologic grade 1 was more common when ER expression was high. A comparable study carried out in Asia (29) found that in 70% of grade 1 carcinomas, 48.2% of grade 2, and 3.5% of grade 3 cases were ER positive (p<0.001). Given that the presence of ER and PR in the tumor tissue is associated positively with the response to hormone therapy and chemotherapy (30), we can therefore conclude that better-differentiated tumors (lower grade) are more probable to be ER and PR positive in addition to having a relatively better prognosis.

In addition, in our research, most patients with grades 1 and 2 had low expression of Ki-67, whereas most patients with grade 3 had high expression of Ki-67 (p=0.001). This demonstrates that significant cell proliferation, which is a hallmark of tumor progression and a worse prognosis, is mostly seen in carcinomas of higher histologic grade. This conclusion is consistent with that of Narbe et al. (31), who likewise observed grade 3 tumors with a mean Ki-67 score of 23.2% and confirmed a substantial positive correlation between Ki-67 and histologic grade (p<0.001). Moreover, Soares et al. found that the increased expression of Ki-67 was linked to a higher incidence of high histologic grade (21).

Furthermore, Table IV shows that, with regard to the histologic grade, HER-2 showed a similar pattern to Ki-67 while not being statistically significant (p=0.051). Arantes Junior observed a similar result; he did not find a statistically significant link but noted that high nuclear grade was associated with HER-2 over-expression (32). Thus, as HER-2 over-expression has no statistically significant correlation with different nuclear grade levels, it appears to be an independent indicator of biological aggressiveness. Its over-expression in individuals with breast cancer indicates a poorer prognosis and a poor response to hormone therapy, such as tamoxifen, which results in lower survival (30).

Concerning tumor size, we did not observe a significant association with ER or PR expression, as shown in Table III. Similarly, Soares et al. found that the mean size of tumors of patients with ER-positive was 3.52 cm, whereas that of patients with ER-negative tumors was 3.73 cm. Furthermore, the mean tumor size in patients with PR-positive tumors was 3.51 cm, whereas in patients with PR-negative tumors, it was 3.72 cm. Nevertheless, there was no discernible relationship between tumor size and ER and PR expression (21). Dayal et al. (28) and Ariga et al. (33) have both reported similar results.

It is well recognized that lymph node status affects breast cancer stage and available treatments. The patient's prognosis in breast cancer is significantly affected by their lymph node status. As the number of positive axillary lymph nodes increases, both the survival rate and the likelihood of recurrence decrease (34). Previous research has shown that HER-2 expression is statistically correlated with lymph node involvement and vascular invasion, while this relationship has not been shown for ER and PR (28,35,36). However, this association was observed neither in our study nor in the study by Soares et al. (21).

The cell adhesion molecule E-cadherin is expressed in healthy breast tissue and can be used as a phenotypic diagnostic for breast cancer. A shorter disease-free interval, a lower overall survival rate, a larger tumor size, a higher histological grade, the development of distant metastases, and ER receptor-negative cancers are all linked to decreased or impaired e-cadherin expression (37). It has been shown that invasive ductal carcinomas and the associated metastatic lymph nodes express E-cadherin aberrantly. Tumor size and the number of metastasized lymph nodes are strongly correlated with E-cadherin expression in the metastasized lymph node (38). Notably, in specimens in which the expression of E-cadherin was examined, we observed a positive correlation between its expression and lymph node involvement. This finding is in line with a recent study indicating that E-cadherin promotes, rather than suppresses, the development of metastasis and invasiveness (39).

Of note, a positive correlation between the age and the maximum size of the tumor was observed. It has been reported in the literature that younger females tend to have more aggressive breast tumors (40,41). It has been shown that a lower age at diagnosis raises the chance of death, and this effect is most noticeable in women under 35 years of age (42). Also, it has been reported that the proportion of tumors with lymphatic invasion decreases progressively with increasing age (43). However, studies from elderly females with breast cancer suggest that there is an association between old age and increased tumor size. These findings indicate that breast cancer might be detected at more advanced stages in the ageing population (44).

Moreover, we would like to note that neoadjuvant anti-HER2 therapy with trastuzumab is highly effective and could be administered to all HER2-positive early breast cancer patients who do not have contraindications. One year of trastuzumab therapy is standard for the vast majority of HER2-positive patients and thus we can realize the clinical significance of immunohistochemical prognostic biomarkers in breast cancer (45).

For the first time, we demonstrate the distribution of molecular breast cancer subtypes and their relation to some clinicopathological characteristics in a large cohort of Greek females. However, our study has some limitations. It is a single-center retrospective study. Also, we do not provide data on the outcome of these patients, but this was out of the scope of the research study.

Our findings highlight the significance of tumor analysis carried out using immunohistochemistry and histopathological analysis. Moreover, our findings indicate that molecular and immunohistochemical analyses can be helpful in advancing biological understanding and enhance the treatment of breast cancer patients.

The Authors declare that they have no competing interests in relation to this study.

PP collected all the data, conceived, and designed the study. VEG statistically analyzed the data. PP and VEG prepared the manuscript. DK, RV, CT contributed to the bibliographic support. FAA, AN, ACL, GCZ, NK, GET interpreted all relevant clinical and laboratory data and provided critical revisions. PP, AN, ACL and GET confirm the authenticity of all the raw data. All Authors contributed to manuscript revision and have read and approved the final version of the manuscript.

The Authors thank all the laboratory and administrative staff of the 1st Department of Pathology, School of Medicine, National and Kapodistrian University of Athens, Greece, for their contribution to the collection of the necessary histopathological and immunohistochemical data for this study.

No funding has been received.