A Comprehensive Expression Analysis of Mucins in Appendiceal Carcinoma in a Multicenter Study: MUC3 Is a Novel Prognostic Factor

Background Mucins are implicated in survival in various cancers, but there have been no report addressed on survival in appendiceal carcinoma, an uncommon disease with different clinical and pathological features from those of other colon cancers. We aimed to investigate the clinical implications of expression of mucins in appendiceal carcinoma. Methods Expression profiles of MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC6, MUC16 and MUC17 in cancer tissue were examined by immunohistochemistry in 108 cases of surgically resected appendiceal carcinoma. Results The following relationships of mucins with clinicopathologic factors were identified: MUC1 with positive lymphatic invasion (p = 0.036); MUC2 with histological type (mucinous carcinoma, p<0.001), superficial invasion depth (p = 0.007), negative venous invasion (p = 0.003), and curative resection (p = 0.019); MUC3 with non-curative resection (p = 0.017); MUC5AC with histological type (mucinous carcinoma, p = 0.002), negative lymphatic invasion (p = 0.021), and negative venous invasion (p = 0.022); and MUC16 with positive lymph node metastasis (p = 0.035), positive venous invasion (p<0.05), and non-curative resection (p = 0.035). A poor prognosis was related to positive lymph node metastasis (p = 0.04), positive lymphatic invasion (p = 0.02), positive venous invasion (p<0.001), non-curative resection (p<0.001), and positive expression of MUC3 (p = 0.004). In multivariate analysis, positive venous invasion (HR: 6.93, 95% CI: 1.93–24.96, p = 0.003), non-curative resection (HR: 10.19, 95% CI: 3.05–34.07, p<0.001) and positive MUC3 expression (HR: 3.37, 95% CI: 1.13–10.03, p = 0.03) were identified as significant independent prognostic factors in patients with appendiceal carcinoma. Conclusions Expression of MUC3 in appendiceal carcinoma is an independent factor for poor prognosis and a useful predictor of outcome in patients with appendiceal carcinoma after surgery.


Introduction
Appendiceal cancer is rare in the United States, with an age-adjusted incidence of 0.12 cases per 1,000,000 people per year [1], and a rate among intestinal cancers of 0.7%, compared to 1.5% for small bowel carcinoma and 97.8% for colon carcinoma in the Surveillance, Epidemiology and End Results (SEER) registry [2]. A similar rarity of appendiceal carcinoma is also found in Japan, with incidences of 0.2% in the Japanese Society for Cancer of the Colon and Rectum Registry and 0.08% in the Japanese Autopsy Annual Database of Colorectal Cancer [3]. The disease differs from cancers at other sites in the colon, with clinical presentation of acute abdominal symptoms suggestive of appendicitis [4,5] and peritoneal mucinous carcinomatosis. The 5-year survival rate for appendiceal carcinoma after surgery is 46-64% [5][6][7]. Curative surgical resection is required for improving survival, and the pathological characteristics of the tumor affect prognosis. Among histological types, patients with non-mucinous carcinoma have poorer survival than those with mucinous carcinoma [5,8], and those with signet ring cell carcinoma also have poor survival [1]. Cases with a high histological grade have poorer survival than low grade cases [6,7]. Thus, prognostic factors in appendiceal carcinoma have included curative resection [6], primary tumor status [6], histological type [1,5,6,8], and histological grade [6,7,9].
The aim of this study was to investigate whether expression of mucins (MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC6, MUC16 and MUC17) has prognostic significance in patients with appendiceal carcinoma using surgical specimens collected from multiple centers.

Patients and Tissue Specimens
Between 1991 and 2013, 108 resected specimens of appendiceal carcinoma were collected from 23  Hospital, and Mie Prefectural General Medical Center (no specified number in these eleven hospitals) waived the need for written informed consent from the participants, and gave us their approval for use of the resected specimens, under the strict condition of privacy protection of the personal information of the patients.
Primary appendiceal carcinomas that were clinically and pathologically diagnosed by surgeons and pathologists were included in the study. Possible cecum cancers with invasion of the appendix, metastatic cancer to the appendix, or carcinoid of the appendix were excluded. Samples were collected from 55 males and 53 females with a mean age of 65 years (range 23-95). The surgical procedures are shown as Table 1. Mucinous peritonitis was found in laparotomy in 14 cases. Of the 108 patients, 34 died, and the causes of death were the primary disease in 30, another disease in 3, and an unknown cause in 1. All specimens were fixed in formalin, embedded in paraffin and cut into 4-mm -thick sections for immunohistochemistry (IHC), in addition to hematoxylin and eosin (HE) staining.

Evaluation of Staining
The results were evaluated based on the percentage of positively stained carcinoma cells. Staining of the following components was evaluated: membrane and cytoplasm for MUC1 and MUC16; supranuclear area for MUC2; membrane for MUC3, cytoplasm for MUC4, MUC5AC and MUC6; and supranuclear area, cytoplasm and membrane for MUC17. Carcinoma cells are considered to be stained positively when at least one of the components was positive. A tumor was considered positive if more than 5% of carcinoma cells were stained, based on our previous use of 5% as the cutoff for mucin expression [17,[28][29][30][31][32][33].

Statistical Analysis
Associations between mucin expression profiles and clinicopathological factors were examined by chi-square test. Postoperative survival was calculated using the Kaplan-Meier method. Differences in survival curves were compared by log-rank test. A Cox proportional hazard analysis was used to estimate hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) in multivariate analysis. P,0.05 was considered significant.

MUC expression in carcinomas
In the 108 cases, the positive expression rates (more than 5% of carcinoma cells stained) of each mucin antigen were MUC1, 47

Relationship of Clinicopathological Factors and Mucin Expression with Survival
The 5-year overall survival rate and median survival period were 62.4% and 2.1 years, respectively. Log-rank tests showed that positive lymph node metastasis (p50.04), positive lymphatic invasion (p50.02), positive venous invasion (p,0.001), and non-curative resection (p,0.001) were significantly related to a worse prognosis (Table 3). Positive expression of MUC3 (p50.004) was also significantly related to a worse prognosis (

Discussion
In this study, the rates of positive expression were MUC1, 47 [37]; and MUC16, 64% [18]. The MUC17 expression rate in colon cancer is unknown, but is lower than that in  normal epithelium [38]. In appendiceal carcinoma, MUC3, MUC6 and MUC16 had lower expression, MUC2 expression was markedly higher, and MUC1 expression was higher than the respective rates in colorectal carcinoma. These differences indicate the distinct characteristics of appendiceal carcinoma compared to other colorectal cancers. We also previously examined mucin expression in small intestinal carcinoma, and found positive expression rates of MUC1, 51.7%; MUC2, 26.7%; MUC3, 55.0%; MUC4, 51.7%; MUC5AC, 33.3%; MUC6, 10.0%; and MUC16, 8.3% (MUC17 expression was not examined) [17]. Appendiceal carcinoma was MUC1-positive in about half of the cases, similarly to small intestinal carcinoma, but other mucin profiles were different. Thus, with regard to mucin expression, appendiceal carcinoma may have a different   carcinogenesis mechanism compared with other colorectal or small intestinal carcinomas.
MUC3 maps to a mucin cluster on chromosome 7q22 and is a membranebound mucin with tandem repeats of 17 amino acids (HSTPSFTS-SITTTETTS) [10]. IHC of MUC3 (mMUC3-1) in formalin-fixed paraffin-embedded specimens has been developed as a specific method for epitope retrieval [13,17,27]. We found that clear linear staining of the surface of villi in the normal mucosa of the small intestine is a good positive control for MUC3 staining [17,27]. Other studies have used different antibodies, including 1143/B7 [34,39,44,45] and M3P [40], and some have evaluated both membranous and cytoplasmic expression [34,44,45]. Using the 1143/B7 antibody, Aloysius et al [39] showed that MUC3 membranous expression is an independent prognostic factor in periampullary cancer. The use of different antibodies and evaluation of different expression patterns might give different results for MUC3, and the association of tumor behavior with results from each MUC3 antibody will be an interesting area for future study.
MUC3 is associated with a poor prognosis in appendiceal carcinoma, but the molecular mechanism of MUC3 in carcinogenesis is uncertain. Epigenetically, expression of MUC3A is contributed by promoter hypomethylation [46]. Cysteine-rich domains of MUC3 promote cell migration and inhibit apoptosis  [47], and the MUC3 C-terminal domain undergoes autoproteolysis at its SEA module, which maintains its availability for potentiation of signaling modulated by HER/ErbB2 phosphorylation to promote migration and invasion [48]. Enhanced MUC3 expression by a tetrameric branched peptide with a conserved TFLK motif inhibits bacteria adherence [49], and expression of MUC3 is altered in inflammatory bowel disease and correlated with disease activity and the extent of inflammation [50]. Thus, MUC3 has several potential roles in malignant and inflammatory cells and these effects might be implicated in the poor prognosis of MUC3-positive patients with appendiceal carcinoma. Expression of MUC1, MUC2, MUC4, MUC5AC, MUC6, MUC16 and MUC17 was not related to survival in appendiceal carcinoma. MUC1 expression is related to a poor prognosis of various human neoplasms and plays an important role in tumor invasion and metastasis [10,12], but in our cases MUC1 expression was only related to positive lymphatic invasion. Mucinous carcinoma has high MUC2 expression compared to other adenocarcinomas in the pancreas, bile duct, ovary, breast [10,12] and colorectum [51]. MUC2 expression is also related to a better prognosis of neoplasms in the stomach, pancreas and bile duct [10,12]. The role of MUC2 in mucinous carcinoma suggests that production of this type of mucin may act as a barrier to cancerous extension, resulting in the indolent nature of many tumors [10]. In the current study, MUC2 expression was associated with mucinous carcinoma, consistent with a previous report [51], and with superficial invasion depth, negative venous invasion, and curative resection, but not with a better prognosis in appendiceal carcinoma.
Shanmugam et al. [35] found that MUC4 expression ($ 75%) is a poor prognostic factor in colorectal cancer. However, high MUC4 expression ($ 75%) in appendiceal carcinoma was not significantly related to survival (data not shown). A cut-off value of more than 5% for MUC4 expression detected with antibody 8G7 is significantly related to survival in many tumors [30][31][32]52]. Kocer et al. [36] found that MUC5AC expression is associated with a better prognosis in colorectal carcinoma, and we also found that MUC5AC expression was related to favorable clinicopathological factors such as negative lymphatic invasion and negative venous invasion. MUC6 expression is a useful marker of pancreatobiliary neoplasms [53][54][55], but has no relationship with clinicopathological factors or survival. MUC16 expression is a poor prognostic factor in cholangiocarcinoma and small intestinal cancer [15,17], and was related to positive lymph node metastasis, positive venous invasion and non-curative resection in appendiceal carcinoma in the current study.
MUC17 expression is related to tumor progression in pancreatic cancer [19], but was not related to clinicopathological factors or survival in appendiceal carcinoma. MUC17 and MUC3 are similarly expressed on the apical surface of intestinal epithelia, are both present in glycocalyx, and are both located on chromosome 7q22 [56,57]. MUC17 and MUC3A both have promoter methylation sites, but those are different (2179 to +52 in MUC17 and 2345 to 275 in MUC3A) [58]. Regarding histone modification, histone H3-K9 is more highly acetylated in MUC17-positive cells, whereas H3-K9 does not play a critical role in MUC3A regulation [58]. In the molecular structures, MUC17 and MUC3 both have an N-terminal large mucin domain, a SEA domain, a transmembrane domain, a cytoplasmic tail, and PDZ-binding motifs [59,60], but their molecular function is different. In enterocytes, in response to carbachol, MUC17 is relocated from the apical membrane to an intracellular vesicular pool distinct from classical endosomes; this behavior is specific for MUC17, and does not occur for MUC3 [60,61]. The current study showed a different IHC staining pattern, with MUC17 in the supranuclear area and MUC3 in the membrane, and different clinical significance. The differences in biological behavior between MUC17 and MUC3 may be due to differences in promoters and regulators, or in the structure and domains. Further studies are needed to determine the differences in the roles of these mucins in carcinogenesis.
We emphasize that this study is base on a large collection (n5108) of a very rare appendiceal carcinoma. Furthermore, we prove that MUC3 only affected the survival, while other mucins with prognostic potentials in many malignancies have little importance. These new data would have a significant clinical impact. The patients with positive MUC3 expression of appendiceal carcinoma, should be followed-up carefully after surgery.
In conclusion, we found that expression of MUC3 in appendiceal carcinoma is an independent poor prognostic factor. MUC3 is a useful predictor of outcome in patients after surgery, and the key mucin for tumor progression in this rare tumor.