Assessing the prognostic significance of MUC4β in mucoepidermoid carcinoma of the salivary glands: An immunohistochemical study

1. Introduction

Salivary glands are diffusely distributed in the oral and para-oral tissues. Salivary gland neoplasms are rare, accounting for just 3–10% of all head and neck neoplasms (Ansari, 2007). The global incidence of malignant salivary gland neoplasms is 0.5–2 per 100,000 (Parkin et al., 2010). Mucoepidermoid carcinomas (MECs) account for 30%–40% of all salivary gland neoplasms and are known for their clinical, histopathological and genetic diversity (Coca-Pelaz et al., 2015; Honjo et al., 2018).

The aggressive behavior of MEC dictates a grade dependent treatment strategy (To et al., 2012). However, an efficient prognostic histopathological grading system is yet to be established (Qannam and Bello, 2016). Qannam in 2016 compared the commonly used grading systems for Mucoepidermoid carcinomas and reported a very low percentage of agreement across all the grading systems, especially in case of minor salivary gland MECs. Thus, research into molecular markers that can be used as an adjunct to routine histopathology becomes important for prognostication of MECs. MUC4 is known for its divergent, tumor suppressor and oncogenic potential (Khiavi et al., 2012; Honjo et al., 2018). Hence, this study, aimed to evaluate MUC4, as a prognostic marker for salivary gland MECs. The review of literature includes a comprehensive list of prognostic markers and molecular cascades that delineates aggressiveness in MEC.

2. Materials and methods

3. Results

4. Discussion

The World Health Organization defines MEC as ‘a malignant glandular epithelial neoplasm characterized by mucous, intermediate and epidermoid cells, with columnar, clear cell and oncocytoid features’ (Coca-Pelaz et al., 2015). MECs can show diverse histologic morphologies, depending on the predominant cell type and pattern. The clinical behavior of this tumor is usually predicted by its histologic grade. Histologically, a predominantly cystic architecture with numerous mucous cells, minimal cytological atypia and scant mitoses qualifies the tumor as a low-grade (Grade I) MECs. On the contrary, the high grade (Grade III) MECs are predominantly cellular, mainly consisting of intermediate and epidermoid cells, interspersed by few mucous cells. High grade MECs are highly anaplastic with large number of mitotic figures, evidence of tumor necrosis, neural, vascular and osseous invasion with infiltrative margins. However, the histopathological criteria used for grading MECs are controversial. Some histopathologically low-grade MECs have shown an aggressive clinical behavior (Auclair et al., 1992; Goode et al., 1998; Brandwein et al., 2001; Bai et al., 2013).

Mucins are high molecular weight glycoproteins normally expressed by various epithelial cell types, including salivary glands (Hollingsworth and Swanson, 2004). Till date, twenty-one mucin genes have been identified in humans, which are further classified into secretory mucins and membrane bound (transmembrane) mucins (Dhanisha et al., 2018). Mucins are usually perceived as the biomolecules implicated in the protection and lubrication of epithelial surfaces. However, current research indicates that mucins, particularly MUC4, can also function as signaling modulators and affect tumor cell phenotype (Singh et al., 2007).

MUC4 is a high molecular weight heterodimeric transmembrane glycoprotein located on chromosome 3 locus 3q29. This gene was first identified in 1991, from a tracheobronchial cDNA library (Dhanisha et al., 2018). The MUC4 glycoprotein complex consists of MUC4 alpha (MUC4α) subunit tightly bound to a transmembrane subunit, MUC4 beta (MUC4β). The MUC4 beta subunit has two epidermal growth factor (EGF)–like domains that act as a ligand for human epidermal growth receptor 2 (HER2) (also known as ErbB2), suggesting that MUC4 acts as an intramembranous autocrine activator of HER2 receptor (Dhanisha et al., 2018). The anti-MUC4β antibody is a more tumor-specific antibody than anti-MUC4α (Weed et al., 2004). In our study, anti-MUC4β antibody was used for immunostaining of MECs.

During embryogenesis, MUC4 is first expressed during the canalicular stage of salivary gland morphogenesis, at the apical surface of well-developed excretory ducts. The expression of MUC4 precedes cytodifferentiation of salivary gland, thus emphasizing its role in cell differentiation. Fully developed salivary glands show weak or no expression of MUC4 in acinar lobules and marked expression of MUC4 on the luminal surface of the ductal cells, particularly the excretory ducts (Liu et al., 2002; Alos et al., 2005; Teshima et al., 2011).

Aberrant expression of MUC4 has been reported in various inflammatory diseases and carcinomas, highlighting the diverse roles of MUC4 in biological processes, such as, epithelial cell renewal and differentiation, cell signaling, cell adhesion and malignancies (Chaturvedi et al., 2008). MUC4 has been studied as a potential biomarker in the diagnosis and prognosis of various neoplasms. Current research has revealed that MUC4 can serve diverse functions in a context-dependent manner. It has been noted that, high expression of MUC4 in oral squamous cell carcinoma, breast cancer, extra hepatic bile duct cancer, pancreatic cancer, colorectal cancer, cholangiocarcinoma and ovarian cancer, is associated with aggressive behavior and high chances of metastasis of these malignancies (Tamada et al., 2006; Chauhan et al., 2006; Shanmugam et al., 2010; Hamada et al., 2012; Mukhopadhyay et al., 2013; Huang et al., 2015; Li et al., 2016; Gautam et al., 2017). On the contrary, improved patient survival is associated with high MUC4 expression in, squamous cell carcinoma of the lung and mucoepidermoid carcinoma of the salivary glands (Weed et al., 2004; Alos et al., 2005; Handra-Luca et al., 2005; Majhi et al., 2013). Thus, the role of MUC4 as a sole marker of prognostication remains inconclusive.

The demographic characteristics of the patients included in this study were consistent with the results of similar previous studies (Weed et al., 2004; Bai et al., 2013). The MUC4β expression was seen in the cell membrane and the cytoplasm of all the neoplastic cells, indicating that during the process of carcinogenesis, there are post-transcriptional alterations that produce aberrantly glycosylated mucins (Remmers et al., 2013; Padler-Karavani, 2014). Although limited by a small sample size, our study demonstrated some important findings. As reported by Weed et al. (2004); Alos et al. (2005) and Handra-Luca et al. (2005), a statistically significant negative correlation between MUC4β expression and the histopathological grade of MEC was observed in the present study. However, among the eight cases of high grade MEC, two cases revealed high MUC4β expression, which is a novel, hitherto unreported finding. Till date, very few studies have evaluated MUC4 as a marker of prognostication in salivary gland MECs. Literature research suggests that MUC4β can serve context dependent diverse functions in cell signaling (Singh et al., 2007).

MUC4β may help identify a subset of patients with favorable prognosis, in cases where it plays a role in tumor differentiation. On the other hand, loss of MUC4-associated antigens, as seen with increase in the grade of MEC, may represent a later event in the dedifferentiation of these tumors. As seen in two of our cases, high-grade MECs showing high expression of MUC4β may either exhibit clinical behavior more consistent with lower-grade tumors or may follow the alternate pathway of MUC4β mediated phosphorylation of ErbB2, ErbB3 and neuregulin, resulting in tumor progression (Weed et al., 2004; Jepson et al., 2002).

Under physiologic conditions, MUC4 is localized to the apical surface of normal epithelial cells, while the receptor tyrosine kinases (RTKs) are restricted to the basolateral membranes. The ErbB family of transmembrane receptor tyrosine kinases consisting of ErbB1/epidermal growth factor receptor, ErbB2, ErbB3 and ErbB4, have been implicated in both cell differentiation and neoplasias (Alroy and Yarden, 1997). Physiological expression of ErbB2 and MUC4, functions as a regulator of cell differentiation via the MUC4β -ErbB2 complex. In normal epithelial cells, the MUC4β-ErbB2 interaction sequesters ErbB2, thereby preventing it from heterodimerizing with the other ErbB family members and activating ErbB2 signaling cascade.During oncogenic transformation, there is loss of polarity of the cells resulting in repositioning of the MUC4 over the entire cell membrane, thus potentiating the interaction between MUC4β and RTKs (Singh et al., 2004). On the contrary, high expression and activation of the ErbB2 receptor is more likely to be mediated by the activated MUC4β-ErbB2-ErbB3-neuregulin complex, potentiating tumor progression. MUC4 expression could thus be an indicator of tumor cell differentiation or a mediator of tumor growth and progression (Cho et al., 1997; Weed et al., 2004; Carraway et al., 2009) (Fig. 4). In neoplastic cells, MUC4 has been reported to cause an increase in the cell-surface populations of ErbB2 and ErbB3 by inducing relocalization of the ErbB2 and ErbB3 receptors from the intracellular compartments to the plasma membrane, thus potentiating ErbB2/ErbB3 signaling and tumor progression (Funes et al., 2006).

In a study by El-Attar RH (El-Attar and Deraz, 2014) it was concluded that normal salivary glands do not express ErbB2. Hence, aberrant expression of ErbB2 in salivary gland neoplasms, like MEC, indicates a distinctly aggressive behavior of these neoplasms. Immunohistochemical studies on expression of ErbB2 in MEC have shown conflicting data, with reports of negative (0%) to as high as 38% ErbB2 expression. However, all the studies have linked expression of ErbB2 with tumor progression and unfavorable prognosis (Kernohan et al., 1991; Sugano et al., 1992; Press et al., 1994). High expression of EGFR has been reported in 73% of high grade tumors and is associated with a poor prognosis (Ettl et al., 2008; Lujan et al., 2010).

A review of literature has shown that some low grade MECs can show an unpredictable aggressive course, while certain high grade MECs may show different clinical outcomes (better/worse) (Coca-Pelez et al., 2015; Aro et al., 2011; Herd et al., 2012). Therefore, it is crucial to predict which MEC patients are prone to an aggressive behavior i.e. recurrence and metastasis, irrespective of the histopathological grade of the tumor. Till date, several diagnostic and prognostic biomarkers for MEC have been evaluated (Table 3). Earlier studies had proposed that the CRTC1-MAML2 or CRTC3-MAML2 translocation positive MECs showed a better prognosis, irrespective of their histopathological grade (Okabe et al., 2006; Behboudi et al., 2006). However, subsequent studies revealed that the CRTC1-MAML2 or CRTC3-MAML2 translocation had a questionable prognostic significance and can only be used as a diagnostic marker (Anzick et al., 2010; Schwarz et al., 2011; Saade et al., 2016). Interestingly, AREG (ligand for ErbB2) expression is positively correlated with CRTC1-MAML2 positive MECs and could be linked to the ErbB2 signaling in MEC (Shinomiya et al., 2016). Studies on p53 expression in MEC revealed its expression in intermediate and high grade MECs and suggested that p53 mutations might represent a genetic switch from a low-grade to a high grade phenotype (Faur et al., 2015). Therefore, irrespective of whether MECs are fusion positive or not, their prognosis is governed by other molecular markers (summarized in Fig. 5).

We would like to hypothesize that, over expression of MUC4β (as seen in two cases of our study) in high grade (grade III) MECs, if associated with a simultaneous high expression of ErBb2/Neu, can synergistically activate the downstream signaling pathways associated with the MUC4β-ERbB2-ERbB3-Neu complex. Thus contributing to the aggressive behavior of the neoplasm (Lujan et al., 2010; Funes et al., 2006; Kang et al., 2017).

In our opinion, MUC4β alone cannot be considered a reliable prognostic marker. The co-expression of ErbB2 and MUC4β needs to be evaluated further as it can serve as a potential and economical marker of prognostication, irrespective of the histological grade of the tumor.

5. Conclusions

In conclusion, the present study demonstrates that MUC4 expression does show a negative correlation with the histopathological grade of MECs. However, there may be some peculiar cases that do not follow this pattern, as seen in two cases of high grade MECs in our study,a hitherto unreported finding. Hence, although previous studies have supported that MUC4 is an indicator of good prognosis; further research is needed to determine whether the role of MUC4 as a marker of prognostication is governed by its association with ErbB2 in view of its divergent tumor suppressor and oncogenic pathway.

Declarations