Expression of Insulin Receptor and c-MET Is Associated with Clinicopathologic Characteristics and Molecular Subtypes in Premenopausal Breast Cancer Patients

: Receptor Tyrosine Kinases (RTKs) represent a class of transmembrane receptors known to play an important role in cancer development and progression. In this study, the expression of insulin receptor (IR) and the hepatocyte growth factor receptor (c-MET) was examined in breast cancer patients. Immunohistochemistry for IR and c-MET was performed on 71 cases of invasive breast cancer and expression scores were correlated with clinicopathologic characteristics and molecular subtypes and further stratiﬁed based on a menopausal status. Expression of IR was signiﬁcantly associated with the tumor grade ( p = 0.017) and estrogen receptor (ER) expression ( p = 0.015). There was a signiﬁcant positive correlation between IR and c-MET expression scores (rho = 0.458, p < 0.001). Among premenopausal cases, IR scores were signiﬁcantly higher in patients with grade I / II disease ( p = 0.025), ER-positive ( p = 0.030), and progesterone receptor (PR)-positive carcinoma ( p = 0.015). c-MET expression scores were signiﬁcantly higher among premenopausal patients with ER-positive ( p = 0.007) and PR-positive carcinoma ( p = 0.024). IR expression scores were signiﬁcantly di ﬀ erent among molecular subtypes for all patients ( p = 0.006) and among premenopausal cases ( p = 0.035). c-MET expression was statistically di ﬀ erent among molecular subtypes for premenopausal patients ( p = 0.019). Survival analysis revealed that the expression status of IR and c-MET was not associated with overall survival. Our ﬁndings support a favorable prognostic value for IR and c-MET expression in premenopausal breast cancer patients.


Introduction
Receptor tyrosine kinases (RTKs) are transmembrane receptors known to bind growth factors mediating the integration and interpretation of diverse extracellular stimuli and functions [1]. Historically, RTKs define the prototypical class of oncogenes involved in most forms of human malignancies [2]. Deregulations of RTKs in terms of mutations or aberrant expression enhance cancer cell proliferation, survival, and invasion, which ultimately drive the development of aggressive tumor phenotypes [3]. records and the electronic database. Body mass index (BMI) was calculated using the standard method as weight in kilograms divided by the height squared in meters. Based on World Health Organization (WHO) definition for obesity, patients were classified as underweight (BMI < 18.5 kg/m 2 ), normal .99 kg/m 2 ), overweight (BMI 25.0-29.99 kg/m 2 ), and obese (BMI ≥ 30.0 kg/m 2 ) [25].
Pathological data were retrieved from pathology reports issued by the Pathology Department at KAUH at the time of diagnosis. Pathological data included details of tumor size, grade, histological subtype, ipsilateral axillary lymph node status, and presence/absence of lymphovascular invasion (LVI). In addition, pathology reports provided details for the expression status of estrogen receptor (ER), progesterone receptor (PR), and Human Epidermal Growth Factor Receptor 2 (HER2/neu). For HER2, immunohistochemistry (IHC) scores of 0 or +1 were considered negative while a score of +3 indicated positive staining. For equivocal results of +2 score, fluorescence in situ hybridization (FISH) analysis positive for gene amplification was positive for HER2 expression. ER-positive patients received hormonal treatment and patients overexpressing HER2 received trastuzumab for one year.
The clinicopathological stage of breast cancer was obtained according to the tumor-node-metastasis (TNM) cancer staging systems of the American Joint Committee on Cancer (AJCC) [26]. The grade of carcinoma was determined based on the Nottingham Combined Histologic Grade system [27]. Accordingly, tumors were classified into grade I (well-differentiated/low grade), grade II (moderately-differentiated/intermediate grade), and grade III (poorly-differentiated/high grade) carcinomas. Breast cancer was further classified into subtypes based on the expression of hormone receptors and HER2 status [28,29] including luminal (ER+ and/or PR+, HER2+/-), HER2-positive (ER-, PR-, HER2+), and triple-negative (ER-, PR-, HER2-).
The Institutional Review Board (IRB) committee of JUST and KAUH (Research number 33/80/2014) approved the study.

Immunohistochemistry
Immunohistochemistry (IHC) was performed using a Dako autostainer-Plus (Dako, Denmark) following a standard protocol. Paraffin-embedded tissues for each tumor were obtained from the Pathology Department and sections were cut at a thickness of 4 µm and prepared for IHC. The sections were dewaxed in two changes in xylene, rehydrated through a descending graded series of ethanol (100%, 90%, 80%, 70%, and 50%), and, lastly, rehydrated in distilled water. Antigen retrieval was performed in a PT-link Instrument (Dako, Glostrup, Denmark) using high pH citrate buffer for 20 min. Endogenous peroxidase activity was blocked by incubation with 2.5% hydrogen peroxide for 10 min. Sections were next washed with phosphate buffered saline (PBS) twice and incubated with primary antibodies for IR (Insulin RB H-70, Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1:100 and c-MET (Met C-12: sc-10, Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1:200 for 45 min at room temperature. These primary antibodies were polyclonal and unconjugated according to the manufacturer's specification sheets. Slides were then washed twice with PBS and signal detection was carried out using The Envision Dual Link System HRP Kit, K8000 (Dako, Glostrub, Denmark) in which slides were incubated with HRP-conjugated secondary antibodies for 30 min at room temperature. Then, slides were washed and incubated with DAB (3,3 -Diaminobenzine) chromogen for 3-4 min. Lastly, slides were washed with distilled water and counterstained with Mayer's Haematoxylin.

Statistical Analysis
Data analysis was performed using IBM SPSS statistical package (IBM Corp. Version 21.0. Armonk, NY, USA). Continuous variables are expressed as mean ± standard deviation or median and interquartile range (IQR). Categorical variables are presented as frequency and percentages (n, %). A non-parametric Mann-Whitney U test was applied to compare two independent groups. Kruskal-Wallis analysis of variance was used to compare multiple independent groups. To compare categorical variables between groups, Pearson's χ 2 test of independence was used. To assess correlations between continuous variables, Spearman's correlation test was applied. All p-values were two-sided and statistical significance is indicated at p < 0.05.
Dichotomization of some categorical variables was considered for correlation analysis of some study variables. This dichotomization was based on sample size and was performed in advance of conducting statistical analysis in order to avoid small sample size upon further stratification of data by menopausal status [28]. Therefore, BMI was categorized as non-obese (<30.00 kg/m 2 ) and obese (≥30.0 kg/m 2 ). The TNM stage of breast cancer was dichotomized as early (I/II) and advanced (III/IV). Histologic grade was further categorized into grades (I/II) and grade (III) tumors. Categories of these tumor variables were selected based on cut points previously published by other researchers [32,33].
Receptor expression was further categorized as either low or high based on IRS scoring. IRS scores of 7-12 were considered high, and scores below 7 were considered low for expression of target receptors [34,35].
Kaplan-Meier survival curves were generated for breast cancer patients based on the expression status of IR and c-MET using GraphPad Prism 8.0.1 software (GraphPad Software, San Diego, CA, USA). Cox proportional hazards models were fitted with overall survival (OS) as the outcome.

Patient Characteristics
Demographic and clinicopathologic features of breast cancer patients are presented in Table 1
The association between clinicopathologic characteristics of breast cancer patients and expression of IR is shown in Table 2. Fifty-three patients (74.6%) were determined to have low expression of IR. IR expression was significantly associated with tumor grade (p = 0.017). Among patients with high IR expression, 42.1% presented with low-to-intermediate grade (I/II) compared to 14.9% of patients who had high grade carcinoma and high IR expression. IR expression was also significantly associated with ER expression (p = 0.015). Low IR expression was mostly found among patients with an ER-negative status (86.7%). Alternatively, high IR expression was associated with an ER-positive status (39.1%). No significant associations were found between IR expression and patient age, BMI, menopausal status, tumor stage, lymph node involvement, PR, HER2, or LVI ( Table 2).
Expression of c-MET was detected in all cases and was mainly observed as cytoplasmic staining of breast cancer cells ( Figure 1E-G). c-MET expression was determined to be strongly positive in 15 (21.1%) cases, moderate in 48 (67.6%), and mild in 8 (11.3%) patients ( Figure 1E-G). Mean immunoreactive score (IRS) for c-MET expression was 7.8 ± 0.3 with a range from 2-12 (median 8.0, IQR 6.0-8.0).
The association between c-MET expression and clinicopathologic characteristics of patients is presented in Table 3. Low c-MET expression was found in 19 (26.8%) patients, while 52 (73.2%) patients had high expression. No significant associations were found between c-MET expression and all demographic and clinicopathologic parameters assessed among the entire sample of breast cancer patients (Table 3).
Correlation between receptor expression scores and clinicopathologic characteristics associated with outcomes in breast cancer patients is shown in Table 4. There was a significant positive correlation between IR and c-MET expression scores among the entire cohort (rho = 0.458, p < 0.001). There were no significant correlations between IR or c-MET scores and age at diagnosis, BMI, tumor size, or number of lymph nodes among breast cancer patients (Table 4).

Correlation between IR and c-MET Expression with Clinicopathologic Characteristics Based on Menopausal Status
Correlations between IR and c-MET scores and clinicopathologic factors associated with outcomes in breast cancer patients stratified according to menopausal status are shown in Tables 5 and 6, respectively. Among premenopausal cases, IR expression scores were significantly higher in patients with grade I/II disease (p = 0.025), ER-positive (p = 0.030), and PR-positive carcinoma (p = 0.015) ( Table 5). There were no significant correlations between IR scores with tumor stage, lymph node status, HER2, and LVI among premenopausal breast cancer patients. Among postmenopausal cases, IR immunostaining scores lacked significant correlation with all clinicopathologic variables tested.
c-MET expression scores were significantly higher among premenopausal patients with ER-positive (p = 0.007) and PR-positive carcinoma (p = 0.024) ( Table 6). There were no significant correlations between c-MET scores with the other tumor characteristics among premenopausal cases. Similar to findings with IR, c-MET expression lacked significant correlations with all clinicopathologic variables tested among postmenopausal breast cancer patients in this study (Table 6).   Figure 2 represents IR and c-MET immunostaining scores among breast cancer molecular subtypes for all patients including premenopausal and postmenopausal cases. IR score was significantly different among molecular subtypes for all patients (p = 0.006) and premenopausal cases (p = 0.035) (Figure 2A,B). IR scores were significantly higher among patients with luminal subtype compared to both HER2-positive and triple-negative disease. No significant difference in IR score was found among molecular subtypes in postmenopausal patients ( Figure 2C). Expression of c-MET was statistically different among molecular subtypes for premenopausal patients (p = 0.019) with higher expression scores for the luminal subtype compared with HER2-positive and triple-negative cases ( Figure 2E).   showed that IR was overexpressed in breast tumors when compared to normal breast tissue [10,37].

292
In cancer cells, IR has been found to drive proliferative and mitogenic effects, which are distinct from 293 the regular metabolic activity mediated by the receptor [5,18]. Activation of IR phosphorylates IR 294 substrate (IRS) which, in turn, activates multiple downstream signal transduction effectors such as 295 phosphoinositide 3-kinase (PI3K)/AKT and ERK pathways [38,39]. As a result, IR regulates several 296 fundamental cellular processes such as cell proliferation, differentiation, cell cycle, and migration 297 [17]. Insulin-like growth factor 1 receptor (IGF-1R) is a common dimerization partner of IR [5,7]. IGF-

298
1R/IR pathway mediated resistance to endocrine treatment and is further required for the growth of 299 hormone-independent breast cancer [7,38,40].

300
A limited number of studies described the expression of IR in breast cancer tissues and its 301 association with disease presentation and prognosis [41]. Results from this study demonstrated that 302 high IR expression was significantly associated with low tumor grade and expression of hormone 303 receptors, which are classified as favorable prognostic factors [42]. When stratified by menopausal 304 status, these associations were found among premenopausal but not postmenopausal patients.

311
OS among breast cancer patients compared to low expression status [43]. In addition, Björner et al.

312
described the expression of IR in a population-based cohort of 1,026 primary invasive breast cancer 313 patients in Sweden [38]. Moderate-strong staining of IR was detected in 82% of breast cancer patients 314 [38]. Strong IR expression was significantly associated with ER-positivity, while negative-weak

Discussion
Insulin is an established mitogen known to promote breast cancer cell growth [10,36]. The mitogenic actions of insulin in breast cancer cells are mediated by activation of IR. Earlier studies showed that IR was overexpressed in breast tumors when compared to normal breast tissue [10,37]. In cancer cells, IR has been found to drive proliferative and mitogenic effects, which are distinct from the regular metabolic activity mediated by the receptor [5,18]. Activation of IR phosphorylates IR substrate (IRS) which, in turn, activates multiple downstream signal transduction effectors such as phosphoinositide 3-kinase (PI3K)/AKT and ERK pathways [38,39]. As a result, IR regulates several fundamental cellular processes such as cell proliferation, differentiation, cell cycle, and migration [17]. Insulin-like growth factor 1 receptor (IGF-1R) is a common dimerization partner of IR [5,7]. IGF-1R/IR pathway mediated resistance to endocrine treatment and is further required for the growth of hormone-independent breast cancer [7,38,40].
A limited number of studies described the expression of IR in breast cancer tissues and its association with disease presentation and prognosis [41]. Results from this study demonstrated that high IR expression was significantly associated with low tumor grade and expression of hormone receptors, which are classified as favorable prognostic factors [42]. When stratified by menopausal status, these associations were found among premenopausal but not postmenopausal patients. Our findings showed no significant associations between IR expression with patient age, BMI, tumor size, or nodal status. In addition, our findings showed that patient survival was not affected by the IR expression status. In agreement with our findings, Mulligan et al. found that high IR expression was significantly associated with low tumor grade, lymph node negativity, and positive expression of PR in a series of breast cancer patients who underwent surgical treatment at three University of Toronto hospitals [43]. High IR expression was associated with improved distant disease-free survival and OS among breast cancer patients compared to low expression status [43]. In addition, Björner et al. described the expression of IR in a population-based cohort of 1,026 primary invasive breast cancer patients in Sweden [38]. Moderate-strong staining of IR was detected in 82% of breast cancer patients [38]. Strong IR expression was significantly associated with ER-positivity, while negative-weak expression of IR was associated with a triple-negative molecular status. However, IR expression lacked prognostic impact regarding event-free survival or distant metastasis among this cohort of Sweden breast cancer patients [38]. Alternatively, Papa et al. examined the expression of IR in 159 specimens for Italian breast cancer patients and compared the expression to the content of 42 non-malignant breast tissues [10]. The expression of IR in breast cancer tissues was more than six-fold higher than its expression in non-malignant breast tissue. In addition, Papa et al. showed that IR expression was positively correlated with tumor size, histological grade, and ER content [10]. No correlations were found between IR levels and age, body weight, menopausal status, and nodal involvement of breast cancer patients [10]. Aljada et al. showed that mRNA levels of IR were significantly higher in cancer patients presented with advanced clinical stage of the disease compared to early stages [44]. As our findings demonstrate, a favorable prognostic impact for IR expression in breast cancer tissues and IR expression may be perceived as a marker of differentiation [43].
c-MET is a high-affinity receptor for HGF [45]. c-MET is classified as the prototype member of the RTK subfamily with Ron and Sea based on shared structural homology of c-MET [46]. HGF binding induces c-MET dimerization and subsequent activation of several downstream transducers and adaptors including PI3K, Src, Grb2, Shc, and signal transducer and activator of transcription (STATs) [45]. Dysregulation of c-MET signaling upregulates diverse tumor cell functions including cell proliferation, survival, cell scattering and motility, epithelial-to-mesenchymal transition, angiogenesis, invasion, and metastasis [47]. In addition, high c-MET expression is critical for acquiring resistance toward chemotherapy and targeted therapies [48,49]. Findings from this study revealed that c-MET is strongly expressed in breast cancer tissues and its levels are associated with hormone-receptor expression and luminal subtype among premenopausal breast cancer patients. However, no associations were found between c-MET expression and other tumor characteristics or patient survival in this analysis. In agreement with our results, several studies revealed a lack of association between c-MET expression and prognostic factors such as patient age, tumor size, and lymph node status [50]. Wang et al. demonstrated a lack of correlation between c-MET expression and clinicopathologic characteristics or prognosis among a cohort of patients diagnosed with triple-negative breast cancer [51]. Lengyel et al. evaluated the expression of c-MET in 40 primary breast cancer patients with positive axillary lymph nodes in Germany and revealed a lack of association between receptor levels and age, lymph node involvement, ER, PR, tumor size, or grade [52]. Nakopoulou et al. showed that c-MET immunoreactivity was associated with low histological grade and favorable prognostic and predictive factors such as ER and PR positivity and negative HER2 expression [53]. Nevertheless, c-MET expression was not significantly associated with other prognostic factors such as tumor size, lymph node status, and stage [53]. Baccelli et al. showed that expression of c-MET in luminal tumors was associated with lymph node metastasis and reduced OS suggesting that c-MET expression could be utilized as a prognostic marker for risk stratification and treatment of patients with hormone receptor-positive breast cancer [54]. Jia et al. also showed that increased c-MET expression in breast cancer tissues was associated with increased tumor size and stage, and reduced OS [48]. c-MET expression was significantly associated with the tumor grade in a series of patients diagnosed with triple-negative breast cancer from Greece and Austria [55]. Results from a meta-analysis of 6010 breast cancer cases indicated that c-MET overexpression correlated with poor relapse-free survival (RFS) and OS in Western patients. However, such an association was not observed among Asian patients [56]. Inconsistencies in findings correlating IR and c-MET expression with the clinicopathologic characteristics of breast cancer or the prognostic impact could be attributed to different approaches applied to assess immunostaining of target receptors and differences in populations examined.
Our findings revealed that breast cancer tissues showed a significant positive correlation for the expression of IR and c-MET. A limited number of studies addressed the potential crosstalk between IR and c-MET. Fafalios et al. reported that c-MET is essential for an optimal response to insulin in hepatocytes in which c-MET directly engage with IR in a hybrid complex of both receptors [24]. Though this finding has been shown in hepatic tissue, it is possible that IR and c-MET receptor hybrids are formed in cancerous tissues as well. This is particularly important since high expression of both receptors was found among luminal disease and, thus, mitogenic signaling in these tumors could be, in part, dependent on IR and c-MET RTKs, which could enable luminal tumors to escape the suppressive effects of endocrine treatments.
Expression of IR and c-MET in breast cancer tissues could make them sensitive to inhibitors targeted for both receptors. Traditionally, IR pathway has been the focus of antidiabetic therapies [5]. IR had received less attention as a therapeutic target in cancer because of the potential impact on glucose homeostasis [7]. Nevertheless, recent evidence from in vitro and in vivo studies had indicated that inhibitors of IR have potent antitumor activity [17,[57][58][59][60]. Sharma et al. showed that IR antagonist S961 suppressed growth and downregulated IR expression in breast cancer cells [17]. KW-2450 (Kyowa Kirin Pharmaceutical Development, Inc., Princeton, NJ, USA) is an orally active multi-targeted tyrosine kinase inhibitor with both IGF-1R and IR inhibitory activity [58]. KW-2450 exerts synergistic antitumor effects in combination with lapatinib and letrozole. This drug is underway in early phase clinical trials to evaluate tolerability and efficacy in combination with lapatinib plus letrozole in patients with HER2-positive advanced or metastatic breast cancer [58]. The dual IGF-1R/IR tyrosine kinase inhibitor BMS-754807 (Bristol-Myers Squibb, New York, NY, USA) resulted in synergistic antitumor effects in combination with letrozole or tamoxifen in preclinical breast cancer models [59]. OSI-906 (OSI Pharmaceuticals, Melville, NY, USA) is another dual tyrosine kinase inhibitor of IGF-1R/IR that has been shown to effectively suppress the growth of hormone-independent breast tumors in the animal model alone, and in combination with the ER down regulator fulvestrant [7]. Several c-MET inhibitors are commercially available, and some are approved for treatment of human cancers [47]. Currently, c-MET inhibitors are being investigated in breast cancer clinical trials as monotherapy or in combination with targeted treatments [61,62].
The main limitation of this study is the comparatively small sample size of our breast cancer patients. This may explain the lack of statistical significance for the association and correlation analysis performed in this study. However, patients in our study are included from a single center with unified diagnostic and management protocols. It is highly recommended to extend this analysis to involve a larger number of breast cancer patients to better elucidate the impact of receptor expression on clinicopathologic and prognostic factors.

Conclusions
To the best of our knowledge, this is the first study to characterize the expression of IR and c-MET in Jordanian breast cancer patients and describe associations with clinicopathologic and prognostic factors. Collectively, IR and c-MET were expressed in different molecular subtypes of breast cancer. The majority of patients had high tumor expression of c-MET. Expression of IR was associated with hormone receptors and tumor grade, which could imply a greater prognostic value for IR compared to c-MET. c-MET expression was associated with hormone receptors but lacked significant associations with other prognostic factors. When stratified based on menopausal status, these associations were confined to premenopausal cases. Despite their association with favorable prognostic factors in this study, ongoing research is being directed toward developing inhibitors of IR and c-MET, which could benefit a specific population of breast cancer patients. Future research will be needed to understand the potential crosstalk between IR and c-MET in breast cancer.