Distribution and Prognostic Significance of Estrogen Receptor α (ERα), Estrogen Receptor β (ERβ), and Human Epidermal Growth Factor Receptor 2 (HER-2) in Thyroid Carcinoma

Purpose The primary aim of this study was to determine the incidence of estrogen receptor α (ERα), estrogen receptor β (ERβ), and human epidermal growth factor receptor 2 (HER-2) expression in various subtypes of thyroid carcinoma (TC) of follicular origin and the secondary aim was to correlate the expression with various clinicopathologic prognostic factors. Methods Immunohistochemistry analysis was performed on archival paraffin-embedded tissue sections (1991–2016). ERα, ERβ, and HER-2 expressions were correlated with clinicopathologic prognostic factors, disease recurrence, and overall survival (OS). Results A total of 264 TC patients were included in the study. Incidences of ERα, ERβ, and HER-2 were 8.1 vs 16.3 vs 13.9% (p=0.15), 26.6 vs 11.5 vs 36.1% (p=0.002), and 12.9 vs 2.9 vs 0% (p=0.003) in papillary thyroid carcinoma (PTC), follicular thyroid carcinoma (FTC), and poorly differentiated thyroid carcinoma (PDTC), respectively. Overall ERα had significant correlation with distant metastases (0.038) and in case of PDTC with multicentricity (p=0.037). ERβ had significant correlation with lymph node metastases (p=0.023) in FTC. HER-2 correlated with tumor size (p=0.027) only on univariate analysis. OS did not correlate with expression of any receptor. Conclusion ERα, ERβ, and HER-2 have differential expression and prognostic implications in different TC subtypes.


Introduction
ere is undeniable association of thyroid carcinoma (TC) with female gender. But only few studies have examined the role of estrogen receptors (ER) in TC [1]. ere has been recent rekindling of interest in the subject and experimental studies are trying to look into the mechanisms by which the female sex hormone works on TC cells [2,3]. Estrogens play a critical role in endocrine tumors, including those of the breast, prostate, and thyroid [4]. As is true for the other malignancies, both isoforms of ER, alpha (ERα) and beta (ERβ), are expressed in TC. ERα is linked with estrogendependent proliferation and ERβ with apoptosis and other suppressive functions in thyroid tumors [5]. Presence of ER is routinely looked for in the breast cancer, which is the most common malignancy among women. Another receptor routinely examined in breast cancer is human epidermal growth factor receptor 2 (HER-2) [6]. ere is evidence of cross-talks between ER and HER-2 pathways [1]. Presence of ER and HER-2 is of prognostic and therapeutic value in breast cancer, and drugs are available to target these receptors [7]. ere is ever growing need to look for alternative therapy in cases of radioiodine refractory TC. Currently, only a few targeted therapies are available, but their efficacy is limited and these are associated with high incidences of debilitating side effects. ER and HER-2 are potential targets, which could be exploited, but there is no enough data on incidence of ER and HER-2 in TC. erefore, we do not know for sure what proportion of patients would benefit from anti-ER and anti-HER-2 therapy [3,5,[8][9][10]. e primary aim of this study was to determine the incidence of ERα, ERβ, and HER-2 expression in various subtypes of TC of follicular origin and the secondary aim was to correlate the expression with various clinicopathologic prognostic factors.

Materials and Methods
is is a retrospective study .
e Institute Research Committee and the Ethics Committee approved the study (2014-187-IMP-EXP). One hundred and twentyfour papillary thyroid carcinoma (PTC), 104 follicular thyroid carcinoma (FTC), and 36 poorly differentiated thyroid carcinoma (PDTC) patients were included in the study. Immunohistochemistry analysis was performed on archival paraffin-embedded tissue sections. Patients with at least 2 years of follow-up were included and those with insufficient data and nonavailable or poorly preserved specimens were excluded from the study. e clinicopathologic profile and follow-up findings were entered in a predesignated proforma.

Immunohistochemistry (IHC).
yroid tissue sections were obtained from archives of pathology department of our institute. Paraffin-embedded 4 μm tissue sections fixed in formalin were used. All histology slides were meticulously reviewed by 2 pathologists experienced in thyroid pathology. e diagnosis of thyroid cancer was reconfirmed. PDTC was diagnosed as per Turin criteria [11]. Slides were deparaffinized, rehydrated, and subjected to microwave heat antigen retrieval in 10 mM citrate buffer (pH 6.0) or Tris EDTA (pH 9.0) for 20-25 min. After blocking endogenous peroxidase activity, sections were incubated with primary antibodies against ERα, ERβ, and HER-2 for 2 hours at room temperature. After three washes with PBS, slides were incubated with universal secondary antibody for 30 minutes at room temperature. Immunoreactivity was visualized using the chromogen 3,3′-diamino-benzidine (DAB) and counterstained with hematoxylin.
Cases of breast cancer with positive staining of all the three primary antibodies were used as positive controls. Sections incubated without primary antibodies served as negative controls. Ten cases each of follicular adenoma (FA) and multinodular goiter (MNG) were also included as controls. Two pathologists independently examined the slides. Nuclear positivity of ERα and ERβ was taken as positive stain. Immunostained slides were scored using the Allred scoring system [12]. A combined score of three or higher was defined as positive staining. HER-2 staining was scored according to the updated ASCO-CAP system [13]. 2+ and 3+ scores of complete membranous staining for HER-2/neu were considered positive. Expression was correlated with subtypes of TC and with various prognostic factors of age, sex, tumor size, lymph nodal status, extrathyroidal invasion, and presence of distant metastases and also with disease recurrence and OS. Data analysis was performed on SPSS software. Descriptive statistics was used according to the distribution of variables. e chi square test or Fisher's exact test was used for the comparison of the frequency or proportions of single variables. Univariate and multivariate analyses were performed to correlate the expression of various IHC markers with the different prognostic factors. Multivariate analysis (step-wise regression with backward selection method) was performed for all the tumors together and each histology type separately. All the prognostic factors selected for the univariate analysis were included and then excluded in the step-wise fashion. e Kaplan-Meier method was used for survival analysis. p values less than 0.05 were considered statistically significant for all the tests.

Results
A total of 264 patients were included in the study. ese consisted of 124 PTC, 104 FTC, and 36 PDTC patients. Mean age of the whole cohort was 46.1 ± 15.5 years and the male to female ratio was 1 : 1.9. e mean tumor size was 4.7 ± 2.8 cm. Tumor multicentricity, extrathyroidal invasion; lymph node metastases, and distant metastases were noted in 26.5, 24.6, 38.6 and 31.4% of the patients, respectively. Seventeen percent of the PTC and 66% of FTC specimens showed presence of poorly differentiated areas. Mean follow-up duration was 73 months. Eleven percent patients developed locoregional recurrence, 7% patients developed metachronous distant metastases, and 17% patients died during the follow-up. Ten percent patients were lost to follow-up and were excluded from survival analysis.
ere was no correlation among ERα, ERβ, and HER-2 expression (p � 0.64, 0.49, and 1.00). Both ERα and ERβ receptor positivity was observed in 6 tumors (PTC � 3, FTC � 2, and PDTC � 1). Among HER-2 expressing tumors, ERα and ERβ expression was observed in one and four patients each of FTC and PTC, respectively, and remaining all had exclusive HER-2 expression. None of the specimen showed all the three receptor expression together. Expression did not correlate with tumor subtypes, i.e., minimal invasive versus widely invasive FTC and PTC variants. e ERα, ERβ, and HER-2 expression pattern was similar in differentiated and poorly differentiated areas of a tumor. Figure 1 shows the photomicrographs depicting immunostaining for ERα, ERβ, and HER-2 in different TC.
Clinicopathologic correlation of ERα, ERβ, and HER-2 is summarized in Tables 1-3, respectively. On univariate analysis including all TC, ERα expression was significantly associated with distant metastases (p � 0.040) and ERβ with lymph node metastases (p � 0.014) and with presence of poorly differentiated areas (p � 0.026) within a tumor. HER-2 expression had significant positive correlation with tumor size (p � 0.027). On analyzing the individual tumor types separately, ERα expression significantly correlated with male gender (p � 0.043) in PTC and multicentricity in PDTC (p � 0.029), whereas ERβ with female gender in PDTC (p � 0.014) and HER-2 with tumor size in FTC (p � 0.014). On multivariate analysis (Table 4) including all the TC, ERα expression was associated with distant metastases (0.038). On separate multivariate analysis for each histology type, no significant correlation was found in case of PTC, whereas significant negative correlation was seen between ERβ and lymph node metastases (p � 0.023) in FTC and positive correlation of ERα and multicentricity in PDTC (p � 0.037). None of the receptor expression had any significant correlation with disease recurrence and OS ( Figure 2).

Discussion
e current study shows that ERα, ERβ, and HER-2 are expressed differentially in TC with the least overlap in expressions. ERα expression was significantly associated with distant metastases in TC in general and multicentricity in PDTC.
Incidence of ER expression in the current study was at par with other published reports, which ranged from 17-54% [1,3,5,[14][15][16][17]. ERα expression has been correlated with tumor virulence, large tumors, extrathyroidal invasion, distant metastases, and reduced DFS [4][5][6][7][8][9][10][11][12][13][14][15][16][17]. But on the contrary, there are reports, which showed that ERα and cosecretary proteins expression are positively associated with well-differentiated tumors and inversely with disease recurrence [1,3]. In the current study, a significant correlation of ERα expression was noted with distant metastases and multicentricity, but not with tumor recurrence and OS. On the other hand, ERβ expression is generally observed in incidental or less virulent tumor [18,19]. ere are reports of differential expression of ERα and ERβ, but the comparison has generally been available between PTC and PDTC and/or anaplastic carcinoma, and FTC has not been studied much [5]. In the current study, ERβ expression was significantly more in PTC and PDTC than in FTC.
ere are limited but conflicting reports on HER-2 expression in TC. HER-2 expression is generally not seen in normal thyroid tissue and a few studies did not find any significant variation in expression among benign or malignant thyroid tumors [10,20]. e reported incidence of HER-2 expression varies from 7-18% in PTC [1,6,[8][9][10].
ere are a few reports showing that the incidence of HER-2 expression is higher in FTC (44-50%), but these had limited number of such cases included in the study [8,9]. On the contrary, one study reported no HER-2 expression in FTC and an incidence of 6.9% in PTC [21]. Our findings seem more in line with the latter. Expression of the HER-2 has been positively linked with capsular invasion and disease recurrence and inversely with well-differentiated tumors [8][9][10]. In our study, the expression correlated with tumor size. e variability in the results of ER and HER-2 expression can partly be explained by the different techniques of detection employed in the various studies. But it is possible that expression of other genes and coregulators might modulate the expression and function of the ER resulting in confounding observations [1,3]. ere is evidence that estrogen may exert a direct effect on thyroid tumorigenesis by ER-independent mechanisms through modulation of the genes involved in cell proliferation independently of its classic nuclear receptor [1,[22][23][24][25][26][27][28]. e proliferative effects of estrogen in thyroid cancer are mainly mediated through the regulation of bcl-2, Bax, and c-fos genes [29,30]. Further, central to the functioning of ER are the coregulators such as steroid receptor coactivator (SRC-1) and nuclear receptor corepressor (NCoR). SRC-1 is significantly associated with the expression of ERα but inversely associated with capsular invasion and positively with well-differentiated tumors [1]. In an experimental study, treatment with estrogen increased protein expression of SRC-1 and its target gene cyclin D1 in the FTC-133 cell line and decreased expression of ERβ, but no alterations in ERα, NCoR, or HER-2 protein expression was detected.
ese effects seem to be contradictory in nature and have partly been explained by the opposing roles of SRC-1 and NCoR [1]. e probable explanation of the contradiction observed in the current study is that no expression of HER-2 in PDTC and yet positive correlation of ERβ with occurrence of PDTC in females could also be hidden in the presence/absence of such coregulators of ER. Since we did not analyze this aspect in the current study, we cannot be absolutely sure and should wait for the future research to throw light on this subject.

Journal of yroid Research
As is true about ER expression, other genetic and epigenetic factors might be responsible for variability in the results of HER-2 expression in TC and SRC-1 is one of them [31]. Another fact worth mentioning is cytoplasmic staining for HER-2, which was noted in 8% of PDTC in the current study. Cytoplasmic staining is usually considered nonsignificant, but in colon cancer, it has been found to have prognostic significance [32]. e exact molecular mechanism behind cytoplasmic expression of HER-2 is yet to be elucidated, but one of the theories is that upregulation of the promoter-binding proteins leads to increase in HER-2 production, which in addition to cell membrane is also expressed in the cytoplasm. Currently, anti-HER-2 molecule targeting membranous (e.g., trastuzumab) as well as cytoplasmic HER-2 (lapatinib) are available for clinical use [33]. e potential clinical application of ER and HER-2 is therapeutic. Estrogen receptor modulators and targeted therapy against HER-2 are an integral part of already wellestablished treatment protocols in breast and prostate cancer [6,7]. A selective estrogen receptor modulator has been used in certain benign thyroid condition, e.g., Reidel's yroiditis, and there are some anecdotal reports of the beneficial role of Tamoxifen in multidrug-resistant thyroid  Recurrence, n (%) 2 (6.2) 23 (9.9) 1.00 1.000 1.000 1.    Journal of yroid Research 5 cancers [34]. But, currently, there is no robust data to support the use of these medicines in TC. eir utility needs to be explored further in cases of radioiodine-resistant TC. e strength of the study is that it included large number of cases with various tumor types including the largest number of FTC patients. e limitation is that it was done on archival tissue and expression of various receptors could be under represented. We also did not employ FISH to confirm HER-2 expression in tumors exhibiting a 2+ staining pattern.

Conclusion
ERα, ERβ, and HER-2 expressions seem to have differential expression and prognostic significance in various thyroid carcinoma subtypes. eir therapeutic implications need to be explored.

Data Availability
e data are property of the institute and would be made available if specific request is made.

Ethical Approval
is article does not contain any studies with human participants or animals performed by any of the authors. is study was approved by the Institute Ethics Committee (2014-187-IMP-EXP).

Conflicts of Interest
e authors declare no conflicts of interest.

Authors' Contributions
Anjali Mishra and Niraj Kumari have equally contributed to this study.