Research ArticleClinical Studies

Expression of Cytokeratin 20 Indicates Invasive Histological Phenotype in Poorly Differentiated Colorectal Adenocarcinoma

YASUO IMAI, HIDETSUGU YAMAGISHI, KAZUNORI FUKUDA, TAKUYA OKAMURA, YUKO ONO, SHINICHI BAN, TOHRU INOUE and YOSHIHIKO UEDA
Anticancer Research January 2014, 34 (1) 159-167;

Abstract

Background/Aim: Cytokeratin (CK) 20 expression is an independent prognostic factor of poorly differentiated adenocarcinoma (PDA) of the colon and rectum. We aimed to investigate the mechanism of its involvement through a clinicopathological study. Patients and Methods: We analyzed 156 surgically resected PDAs, which were sub-classified as solid type (Por1) showing expansive growth, or non-solid type (Por2) showing infiltrative growth. Associations of CK20 expression with morphological features and molecular markers were analyzed. Results: CK20+ PDA (n=91) was associated with more advanced disease stage and unfavorable prognosis compared with CK20− PDA (n=65). Pathologically, CK20+ PDA was significantly associated with p53 overexpression, Por2, abundant fibrous stroma, and stepwise de-differentiation, while CK20− PDA was significantly associated with mismatch repair deficiency, Por1, sparse fibrous stroma, and de novo histogenesis. Conclusion: CK20 expression in PDA is closely associated with invasive histological features, providing prognostic significance, and may also point to a specific histogenetic pathway.

To date, more than 20 subtypes of cytokeratin (CK) have been classified and numbered based on molecular weight and isoelectric pH (1, 2). CK7 is expressed in numerous ductal and glandular epithelia, including those of lung, breast, ovary, and endometrium, while CK20 is expressed in the gastrointestinal epithelium, urothelium and in Merkel cells. Coordinate CK7/CK20 expression profiles have been extensively studied in a variety of primary and metastatic carcinomas (3-8). The CK7−/CK20+ profile is known to be highly characteristic of colorectal cancer (CRC); however, not all CRCs exhibit this profile. A substantial proportion of CRC was reported to be either CK7+ or CK20−, particularly in right-sided and high-grade CRCs, consisting of poorly differentiated adenocarcinoma (PDA) and undifferentiated carcinoma, among others (9). CRC may be classified into adenocarcinoma, mucinous adenocarcinoma, signet-ring cell carcinoma, squamous cell carcinoma, adenosquamous carcinoma, medullary carcinoma, undifferentiated carcinoma, in addition to other rare variants described in the World Health Organization (WHO) classification of colorectal cancer (10). Previously, we investigated CK7/CK20 profiles of various histological CRC subtypes. Normal colonic mucosa usually expresses CK20, and we expected that loss of CK20 might suggest marked de-differentiation and disease progression, and accordingly, an unfavorable prognosis. CK20 expression was retained in 61 out of 63 (96.8%) well- and moderately differentiated adenocarcinomas (WMDAs) and 47 out of 91 (51.7%) PDAs with less than 50% glandular structural components (11). Surprisingly, survival analysis revealed that loss of CK20 expression was associated with significantly favorable prognosis in PDA (11). In the present study, we performed immunohistochemical analyses of mismatch repair (MMR) deficiency and p53 overexpression. In addition, we histologically-classified PDA as solid type (Por1) or non-solid type (Por2), semi-quantified amounts of fibrous stroma, and searched for histological evidence of disease progression from WMDA to PDA. We also investigated associations between these histological features and CK20 expression status to explore the mechanism of an unfavorable prognosis of PDA with CK20 expression.

Patients and Methods

Patients and tumor samples. PDAs were selected from all CRCs surgically resected at the Dokkyo Medical University Koshigaya Hospital (DMUKH) between 1990 and 2011 (tumors 1-53), Tokyo Kosei Nenkin Hospital (TKNH) between 1991 and 2010 (tumors 54-90), Saiseikai Kawaguchi General Hospital (SKGH) between 2000 and 2011 (tumors 91-141), and the International University of Health and Welfare Shioya Hospital (IUHWSH) between 2000 and 2011 (tumors 142-156) solely on the basis of tumor histology, and included 155 patients with a total of 156 resected tumors. Solid-type PDAs (tumors 67 and 68) were derived from one patient with hereditary non-polyposis colorectal cancer. For comparison, WMDAs surgically-resected at TKNH between April 1998 and March 2000 were also included, and consisted of 63 tumors from 63 patients (tumors 201-263). All tumors in this study were previously diagnosed as primary CRC.

Figure 1.
Figure 1.

Sub-classification of poorly differentiated adenocarcinoma (PDA). A: Solid type (Por1) tumor. Tumor 144, showing pushing margin with med fibrous stroma [hematoxylin and eosin (H&E) staining, ×20]. B: Non-solid type (Por2) tumor. Tumor 154, showing infiltrative margin with int fibrous stroma (H&E staining, ×20). C: Por1 with int fibrous stroma. Tumor 82 (H&E staining, ×20). D: Por2 with med fibrous stroma. Tumor 140 (H&E staining, ×40).

Archival tissue blocks were obtained from the pathological departments of the involved hospitals. For survival analysis, patients with complete medical records were included and followed-up for five years, while individuals with invasive cancer originating from other sites were excluded. Clinicopathological classifications and stage groupings were performed based on the WHO classification of colorectal cancer and the tumor, node, metastases (TNM) staging system by the American Joint Committee on Cancer (10, 12). PDA was sub-classified as either Por1 or Por2. Por1 consisted of tumor cells with solid growth, typically accompanied by sparsely distributed stromal cells (case no. 144: Figure 1A). Tumor nests revealed mostly pushing margins. In contrast, Por2 consisted of tumor cells forming a trabecular structure or a small cluster of tumor cells, and was typically rich in fibrous stroma (case no. 154: Figure 1B). Tumor nests displayed an infiltrative margin. In cases exhibiting intermingling of Por1 and Por2 subtypes, classification was based on the predominant type. In cases displaying features of both Por1 and Por2 subtypes (for example, solid growth with abundant stroma (case no. 82: Figure 1C) or a trabecular/small clustering structure with sparse stroma (case no. 140: Figure 1D), classification was performed on the basis of cell growth pattern. The quantity of fibrous stroma was semi-quantified as either med (sparse fibrous stroma), int (intermediate between med and sci), or sci (abundant fibrous stroma). These classifications are consistent with those proposed by the Japanese Classification of Colorectal Carcinoma with modifications (13).

Histogenetic pattern was analyzed by investigating the boundary between non-neoplastic mucosa and cancer. PDA was then subclassified as either de novo type (de novo PDA) when PDA appeared abruptly next to the non-neoplastic mucosa (case no. 144 and 154: Figure 2A and B) or sequentially dedifferentiated type (PDA in the WMDA-PDA sequence or in the stepwise dedifferentiation) when sequential disease progression from non-neoplastic mucosa to adenoma and WMDA and from WMDA to PDA was observed (case no. 149 and 129: Figure 2C and D). This study protocol was approved by the ethical review boards of the participating hospitals: DMUKH, Koshigaya 23008; TKNH, 30/11/2011; SKGH, no. 24-5: IUHWSH, FK-94.

Figure 2.
Figure 2.

Histogenetic analysis by investigating the boundary between non-neoplastic mucosa and cancer. A: De novo histogenesis of solid type (Por1) tumors. Tumor 144 [hematoxylin and eosin (H&E) staining, ×4]. B: De novo histogenesis of non-solid type (Por2) tumors. Tumor 154 (H&E staining, ×4). C: Stepwise dedifferentiation of Por1. Tumor 149 (H&E staining, ×4). D: Stepwise dedifferentiation of Por2. Tumor 129 (H&E staining, ×10).

Immunohistochemistry. Tumor specimens were fixed in 10% neutral-buffered formalin for 48 h, embedded in paraffin, and cut into 4-μm-thick sections. Representative tissue blocks with the least differentiated carcinoma component were used.

For analysis of CK expression, antigen retrieval was performed using microwave irradiation for 5 min. Sections were incubated with primary antibodies detecting CK7 (mouse monoclonal, OV-TC, 1:100; Dako, Glostrup, Denmark) and CK20 (mouse monoclonal, Ks20.8, 1:100; Dako), for 30 min at room temperature. For analysis of human mutL homolog 1 (hMLH1), human mutS homolog 2 (hMSH2), and p53 overexpression, antigen retrieval was performed using microwave irradiation for 10 min, autoclaving (121°C) for 5 min, and by heating at 95°C for 20 min in a water bath, respectively. Primary antibodies detecting MLH1 (rabbit monoclonal, EPR3894, 1:400; GeneTex, San Antonio, TX, USA), MSH2 (rabbit polyclonal, 15520-1-AP, 1:200; Proteintec, Chicago, IL, USA), and p53 (mouse monoclonal, clone DO-7, 1:200; Dako) were used. Samples were treated overnight with primary antibodies at 4°C. Immunostaining was performed using an N-Histofine Simple Stain MAX-PO kit (Nichirei, Tokyo, Japan). On the basis of expression levels in normal colonic mucosa, positive staining of ≥25% of tumor cells for CK20 and ≥1% of tumor cells for CK7 was evaluated as positive, respectively. The immunostaining results for MMR proteins were either completely negative (negative) or nearly 100% positive (positive). Tumors that failed to express either of the two proteins were considered MMR-deficient. For p53, strong positive staining of ≥25% of tumor cell nuclei was evaluated as positive for overexpression, which significantly matched the results of p53 mutational analysis in the WMDA cases (data not shown).

Statistics. Specific parameters between two patient cohorts were compared using the chi-square test with or without Yates' correction, or the Fisher's exact test. Age was compared using the Mann–Whitney U-test, and survival curves were analyzed using Kaplan–Meier and log-rank tests. Univariate analyses were performed using Cox regression analysis for investigating associations between parameters and prognosis. Multivariate analyses were performed using a step wise variable elimination method with an entry limit of p<0.10 and a removal limit of p>0.05. A value of p<0.05 was considered significant. Statistical analyses were performed using IBM SPSS Statistics 20 (IBM, Armonk, NY, USA).

Results

Clinicopathological characteristics of PDA and comparison with WMDA. Clinicopathological characteristics of PDA were analyzed and compared to those of WMDA. A total of 156 PDA tumors and 63 control WMDAs were analyzed. PDA tended to be located in the proximal colon, and tumor stage of progression in terms of depth of invasion, lymphatic invasion, nodal metastasis, and TNM stage, was significantly more advanced than that of WMDA. In accordance with these results, patients with PDA were significantly more likely to undergo chemotherapy compared to patients with WMDA. PDA was slightly more frequent in female than in male patients compared with WMDA. MMR deficiency was significantly more frequent in PDA than in WMDA. These characteristics of PDA are in accordance with those of previous studies (14-16) (Table I).

View this table:
Table I.

Clinicopathological characteristics of well- and moderately differentiated adenocarcinoma (WMDA) and poorly differentiated adenocarcinoma (PDA).

Clinicopathological characteristics of PDA in association with CK20 expression status. We previously reported that CK20 expression is a significant prognostic predictor of PDA (11). In this study, we combined the first and the second PDA cohorts in our previous study and analyzed them as a single cohort. Of the 156 PDAs, 50 (32.1%) were positive and 106 (67.9%) were negative for CK7, and 91 (58.3%) were positive and 65 (41.7%) were negative for CK20. No significant differences in age, gender, family history of CRC, depth of tumor invasion, venous invasion, lymphatic invasion, and irradiation were observed between CK20+ and CK20− PDAs. In contrast, we observed a significant difference in tumor location, nodal metastasis, chemotherapy, and TNM stage between CK20+ and CK20− PDAs. CK20− PDA was observed in the proximal colon significantly more often than in the distal colon. Nodal metastasis was significantly more frequent in CK20+ PDA than in CK20− PDA. Patients with CK20+ PDA underwent chemotherapy significantly more frequently than patients with CK20− PDA. CK20+ PDAs exhibited more advanced TNM stage compared with CK20− PDAs. MMR deficiency and p53 overexpression were observed in 35 (22.4%) and 90 (57.7%) PDAs, respectively. MMR deficiency was significantly more frequent in CK20− PDA, while p53 overexpression was significantly more frequent in CK20+ PDA. Similar trends in patients' gender, tumor location, and MMR proficiency were observed in CK20+ PDA and WMDA. In contrast, depth of tumor invasion, venous and lymphatic invasion, nodal metastasis, chemotherapy, and p53 overexpression were significantly more frequent in CK20+ PDA than in WMDA. This is in contrast with comparison between CK20− PDA and WMDA. These results are summarized in Table I.

Cox regression analysis revealed that CK20 expression and TNM stage were the only significant prognostic predictors of disease-free survival and overall survival (Table II). The disease-free survival was significantly higher in CK20− PDA than in CK20+ PDA, and comparable to that of WMDA. Overall survival was also significantly higher in CK20− PDA compared with CK20+ PDA, although inferior to that of WMDA (Figure 3).

Histological sub-classification and histogenetic analysis of PDA. PDA was sub-classified according to the histological structure of tumor cell clusters. PDA comprised of 62 Por1 and 94 Por2 cases, with a stromal tissue component of 23 med, 106 int, and 27 sci. The stromal tissue component of Por1 was med in 22 and int in 40, while those of Por2 were med in one, int in 66, and sci in 27. The amount of stromal tissue in Por1 was significantly greater than that in Por2. As for histogenesis, Por1 consisted of 34 de novo cancers and 28 cancers in stepwise de-differentiation, while Por2 consisted of 29 de novo cancers and 65 cancers in stepwise dedifferentiation. Thus, we observed a significant tendency in Por1 to favor de novo occurrence and Por2 to favor stepwise de-differentiation (Table III).

Figure 3.
Figure 3.

The disease-free (A) and overall (B) survival curves of well- and moderately differentiated adenocarcinoma (WMDA), cytokeratin (CK) 20− and CK20+ poorly differentiated adenocarcinoma (PDA) shown by the Kaplan-Meier method.

Histological sub-classification of PDA in association with CK20 expression status. We next investigated the associations between CK20 status and histological subclassification, MMR deficiency, and p53 overexpression in PDA. CK20− PDA consisted of 35 Por1 and 30 Por2, with a stromal tissue component of 17 med, 40 int, and 8 sci. CK20+ PDA consisted of 27 Por1 and 64 Por2 with a stromal tissue component of 6 med, 66 int, and 19 sci. These results suggest that CK20− PDA is associated with a higher percentage of Por1 and a smaller amount of stromal tissue component compared with CK20+ PDA. As for histogenesis, CK20− PDA consisted of 39 de novo cancers and 26 cancers in stepwise de-differentiation, while CK20+ PDA consisted of 24 de novo cancers and 67 cancers in stepwise de-differentiation. Thus, CK20− PDA was significantly associated with de novo occurrence while CK20+ PDA was significantly associated with stepwise de-differentiation.

View this table:
Table II.

Cox regression analysis of prognostic parameters in poorly differentiated adenocarcinoma (PDA).

In summary, CK20− PDA significantly favored Por1, de novo histogenesis, and a smaller amount of stromal tissue. Conversely, CK20+ PDA significantly favored Por2, histogenesis in stepwise de-differentiation, and a higher levels of stromal tissue. These results are summarized in Table IV.

View this table:
Table III.

Histological sub-classification and histogenetic analysis of poorly differentiated adenocarcinoma (PDA).

View this table:
Table IV.

Histological sub-classification of poorly differentiated adenocarcinoma (PDA) in association with cytokeratin (CK) 20 expression status.

Discussion

While CK20 expression is characteristic of CRC, not all cases of CRC express CK20. We recently reported that nearly half of PDAs do not express CK20 and that CK20 expression status can be a marker of patient prognosis in terms of disease-free and overall survival (11). This was initially observed following the analysis of 91 PDAs (the first cohort) and subsequently validated by analyzing 66 PDAs (the second cohort). We proposed several mechanisms, such as promoter methylation of transcription factors regulating the CK20 gene associated with MMR deficiency and CpG island methylation phenotype (CIMP), and CK20-mediated cell motility. In this study, we further explored this mechanism by performing detailed histological analyses in a large patient cohort of 156 PDAs. One PDA was excluded from the present study because of unavailability of further material.

Compared to WMDA, PDA was significantly associated with proximal tumor location, more advanced depth of tumor invasion, lymphatic invasion, nodal metastasis, high TNM stage, performance of chemotherapy, and MMR deficiency, highlighting the possible association with high levels of microsatellite instability (MSI-H). Comparative analysis between CK20− and CK20+ PDAs revealed that CK20− PDA was significantly associated with proximal tumor location, absence of nodal metastasis, low TNM stage, absence of chemotherapy, MMR deficiency, and absence of p53 overexpression, suggesting a possible association with MSI-H (14-18). Conversely, CK20+ PDA was significantly associated with distal tumor location, nodal metastasis, high TNM stage, performance of chemotherapy, MMR proficiency, and p53 overexpression. It is well-known that mutation of the p53 gene plays a critical role in the malignant transformation of CRC in the adenoma–carcinoma sequence of the multistep carcinogenetic model of CRC (19). p53 gene mutations and p53 protein overexpression have been reported to be inversely associated with MSI-H (16, 18). This suggests that the majority of CK20+ PDA may arise through pathogenetic pathways distinct from MMR deficiency. In fact, CK20+ PDA revealed similarities with WMDA in terms of patient gender, tumor location, and MMR proficiency in contrast to CK20− PDA. However, depth of tumor invasion, venous and lymphatic invasion, nodal metastasis, chemotherapy, and p53 overexpression were significantly more frequent in CK20+ PDA than in WMDA in contrast to comparison between CK20− PDA and WMDA. These data suggest that CK20+ PDA may actually represent a more progressed form of WMDA.

The histopathological features of PDA in association with CK20 expression status were also investigated in detail. PDAs were sub-classified into Por1 and Por2, and the amount of fibrous stroma was also classified into med, int and sci, as proposed by the Japanese Society for Cancer of the Colon and Rectum (13). In this study, histological diagnosis of PDA was performed based on the least differentiated component of cancer according to the WHO classification (10). However, histological sub-classification of intermingling Por1 and Por2 cases was performed according to the most predominant feature, as proposed by the Japanese classification, because Por1 and Por2 cannot be classified on the basis of differentiation. The CK20− phenotype was significantly associated with Por1, de novo histogenesis, less fibrous stroma, and MMR deficiency, while the CK20+ phenotype was significantly associated with Por2, stepwise de-differentiation, more abundant fibrous stroma, and p53 overexpression. These results suggest that many CK20− PDAs may arise as de novo cancer through mechanisms related to MMR deficiency or CIMP, while CK20+ PDAs may arise via the stepwise de-differentiation pathway from WMDA to PDA in multistep carcinogenesis. Therefore, the majority of CK20+ PDAs may not completely lose CK20 expression and bear p53 mutations. It is well-known that p53 mutations are a marker of unfavorable prognosis (20-22).

Generally, Por1 tumors grow expansively with a pushing margin and Por2 tumors grow infiltratively with a jagged margin, suggesting that Por2 might be more invasive than Por1. Indeed, Komori et al. investigated 78 cases of colorectal PDA, consisting of 29 Por1 and 49 Por2, and found that frequencies of nodal and organ metastases in the Por2 group were significantly higher than those in the Por1 group (23). The survival rate of the Por2 group was lower than that of the Por1 group, although this was not statistically significant. They concluded that prognosis of patients with Por2 tended to be unfavorable compared with these with Por1. In accordance with this study, we observed that Por2 demonstrated worse prognosis than Por1, although this was not statistically significant (data not shown). Our present data indicate that CK20 expression status is clearly a more powerful prognostic predictor of PDA than histological subclassification into Por1 or Por2. Growth of fibrous stroma around tumor cells, termed desmoplastic reaction, is the result of tumor cell infiltration and the amount of fibrous stroma is closely associated with the mode of invasion. Thus, it is reasonable that Por1 is associated with sparse stroma and Por2 with abundant stroma. Accordingly, we speculate that CK20+ PDA may appear in the late stages of the stepwise de-differentiation pathway with a highly invasive growth pattern. Conversely, CK20− PDA may arise de novo with a malignant phenotype that is less aggressive than CK20+ PDA.

In our previous study, we discussed whether down-regulation of CK20 may result from down-regulation of caudal type homeobox 1 (CDX1), which plays a role in intestinal epithelial cell differentiation (24). CDX1 is down-regulated by promoter hypermethylation in a number of CRC-derived cell lines, as well as in patient samples (25-28). CDX1 overexpression led to induction of CK20, while CDX1-knockdown resulted in down-regulation of CK20 in colon cancer cell lines. Promoter methylation is closely associated with CIMP, which underlies but does not completely coincide with MSI-H in CRC (29, 30). CIMP+ CRCs exhibit characteristic clinical and pathological features that include origin in the proximal colon, higher frequency in females and poorly differentiated, mucinous histology (29, 31, 32). We therefore speculated that CIMP may be the underlying mechanism, leading to down-regulation of CDX1, resulting in decreased CK20 expression. Methylation of several CpG islands in the normal colonic mucosa has been shown to increase with age (33-35). Age-related methylation and subsequent inactivation of certain genes, such as p16 (INK4) and hMLH1, has been suggested as a predisposing factor for increased risk of cancer with age (36, 37), and this may partly account for de novo carcinogenesis of Por1 tumors with MMR deficiency from normal colonic mucosa.

Figure 4.
Figure 4.

Scheme of the two histogenetic pathways of poorly differentiated adenocarcinoma (PDA). Densities of cytokeratin (CK) 20 bars represent CK20 expression levels from positive, indicated by solid bars, to negative, indicated by open bars. WMDA, well- and moderately differentiated adenocarcinoma. Por1, solid type. Por2, non-solid type. CIMP, CpG island methylation phenotype. MMR, Mismatch repair.

Why preservation of CK20 expression is associated with a more invasive PDA phenotype remains unknown. In our previous study, we discussed whether CK20 expression might be directly associated with metastasizing potential. This is based on the study showing that metastasizing CRCs typically express CK20 and previous reports have shown that CK20 expression regulated by peroxisome proliferator-activated receptor-γ enhanced invasiveness of tamoxifen-resistant MCF-7 breast cancer cells (3, 5, 6, 9, 38). In the present study, Por1 and Por2 comprised 53.8% (35/65) and 46.2% (30/65) of CK20− PDAs, respectively, and 29.7% (27/91) and 70.3% (64/91) of 91 CK20+ PDAs, respectively. These data indicate that approximately half of CK20− PDAs grow infiltratively and that about a third of CK20+ PDAs grow expansively. Therefore, we infer that down-regulation of certain cell adhesion molecules or up-regulation of cell motility-associated molecules may occur more frequently in Por2 tumors as a result of accumulated gene mutations during multistep carcinogenesis. Our hypothesis representing two distinct histogenetic pathways of CK20− and CK20+ PDAs in the most typical form is summarized in Figure 4.

In conclusion, we performed a detailed investigation of the histological phenotype of CK20− and CK20+ PDAs to gain deeper insight into CK20 expression and prognosis. CK20+ PDA was more closely associated with a highly invasive histological phenotype and likely to occur following multistep de-differentiation, while CK20− PDA was more closely associated with a less invasive phenotype and likely to occur de novo. CK20 expression may be a marker of invasive PDA phenotype and may serve as a useful prognostic marker.

Footnotes

  • This article is freely accessible online.

  • Received November 9, 2013.
  • Revision received November 29, 2013.
  • Accepted December 2, 2013.

References

  1. ↵
    1. Moll R,
    2. Franke WW,
    3. Schiller DL,
    4. Geiger B,
    5. Krepler R
    : The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31(1): 11-24, 1982.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Moll R,
    2. Zimbelmann R,
    3. Goldschmidt MD,
    4. Keith M,
    5. Laufer J,
    6. Kasper M,
    7. Koch PJ,
    8. Franke WW
    : The human gene encoding cytokeratin 20 and its expression during fetal development and in gastrointestinal carcinomas. Differentiation 53(2): 75-93, 1993.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Varadhachary GR,
    2. Abbruzzese JL,
    3. Lenzi R
    : Diagnostic strategies for unknown primary cancer. Cancer 100(9): 1776-1785, 2004.
    OpenUrlCrossRefPubMed
    1. Moll R,
    2. Löwe A,
    3. Laufer J,
    4. Franke WW
    : Cytokeratin 20 in human carcinomas. A new histodiagnostic marker detected by monoclonal antibodies. Am J Pathol 140(2): 427-447, 1992.
    OpenUrlPubMed
  5. ↵
    1. Loy TS,
    2. Calaluce RD
    : Utility of cytokeratin immunostaining in separating pulmonary adenocarcinomas from colonic adenocarcinomas. Am J Clin Pathol 102(6): 764-767, 1994.
    OpenUrlPubMed
  6. ↵
    1. Wauters CC,
    2. Smedts F,
    3. Gerrits LG,
    4. Bosman FT,
    5. Ramaekers FC
    : Keratins 7 and 20 as diagnostic markers of carcinomas metastatic to the ovary. Hum Pathol 26(8): 852-855, 1995.
    OpenUrlCrossRefPubMed
    1. Chu P,
    2. Wu E,
    3. Weiss LM
    : Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: A survey of 435 cases. Mod Pathol 13(9): 962-972, 2000.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Miettinen M
    : Keratin 20: immunohistochemical marker for gastrointestinal, urothelial, and Merkel cell carcinomas. Mod Pathol 8(4): 384-388, 1995.
    OpenUrlPubMed
  9. ↵
    1. Park SY,
    2. Kim HS,
    3. Hong EK,
    4. Kim WH
    : Expression of cytokeratins 7 and 20 in primary carcinomas of the stomach and colorectum and their value in the differential diagnosis of metastatic carcinomas to the ovary. Hum Pathol 33(11): 1078-1085, 2002.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Bozman FT,
    2. Carneiro F,
    3. Hruban RH,
    4. Theise ND
    1. Hamilton SR,
    2. Bosman FT,
    3. Boffetta P,
    4. Ilyas M,
    5. Morreau H,
    6. Nakamura SI,
    7. Quirke P,
    8. Riboli E,
    9. Sobin LH
    : Carcinoma of the colon and rectum. In: WHO classification of tumours of the digestive system, Fourth Edition. Bozman FT, Carneiro F, Hruban RH, Theise ND, (eds). Lyon: IARC Press, pp. 134-146, 2010.
  11. ↵
    1. Yamagishi H,
    2. Imai Y,
    3. Okamura T,
    4. Fukuda K,
    5. Ono Y,
    6. Ban S,
    7. Inoue T,
    8. Ueda Y
    : Aberrant cytokeratin expression as a possible prognostic predictor in poorly differentiated colorectal carcinoma. J Gastroenterol Hepatol 28(12): 1815-1822, 2013.
    OpenUrlPubMed
  12. ↵
    1. American Joint Committee on Cancer
    . Colon and Rectum Cancer Staging. Seventh Edition. 2009. Available from URL: http://www.cancerstaging.org/staging/index.html.
  13. ↵
    The General Rules of Japanese Classification of Colorectal Carcinoma, Eighth Edition. The Japanese Society for Cancer of the Colon and Rectum, Kanehara, Tokyo, Japan, 2013.
  14. ↵
    1. Kim H,
    2. Jen J,
    3. Vogelstein B,
    4. Hamilton SR
    : Clinical and pathological characteristics of sporadic colorectal carcinomas with DNA replication errors in microsatellite sequences. Am J Pathol 145(1): 148-156, 1994.
    OpenUrlPubMed
    1. Jass JR,
    2. Do KA,
    3. Simms LA,
    4. Iino H,
    5. Wynter C,
    6. Pillay SP,
    7. Searle J,
    8. Radford-Smith G,
    9. Young J,
    10. Leggett B
    : Morphology of sporadic colorectal cancer with DNA replication errors. Gut 42(5): 673-679, 1998.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Forster S,
    2. Sattler HP,
    3. Hack M,
    4. Romanakis K,
    5. Rohde V,
    6. Seitz G,
    7. Wullich B
    : Microsatellite instability in sporadic carcinomas of the proximal colon: association with diploid DNA content, negative protein expression of p53, and distinct histomorphologic features. Surgery 123(1): 13-18, 1998.
    OpenUrlPubMed
    1. Samowitz WS,
    2. Curtin K,
    3. Ma KN,
    4. Schaffer D,
    5. Coleman LW,
    6. Leppert M,
    7. Slattery ML
    : Microsatellite instability in sporadic colon cancer is associated with an improved prognosis at the population level. Cancer Epidemiol Biomarkers Prev 10(9): 917-923, 2001.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Samowitz WS,
    2. Holden JA,
    3. Curtin K,
    4. Edwards SL,
    5. Walker AR,
    6. Lin HA,
    7. Robertson MA,
    8. Nichols MF,
    9. Gruenthal KM,
    10. Lynch BJ,
    11. Leppert MF,
    12. Slattery ML
    : Inverse relationship between microsatellite instability and K-ras and p53 gene alterations in colon cancer. Am J Pathol 158(4): 1517-1524, 2001.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Kinzler KW,
    2. Vogelstein B
    : Lessons from hereditary colorectal cancer. Cell 87(2): 159-170, 1996.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Hamelin R,
    2. Laurent-Puig P,
    3. Olschwang S,
    4. Jego N,
    5. Asselain B,
    6. Remvikos Y,
    7. Girodet J,
    8. Salmon RJ,
    9. Thomas G
    : Association of p53 mutations with short survival in colorectal cancer. Gastroenterology 106(1): 42-48, 1994.
    OpenUrlPubMed
    1. Zeng ZS,
    2. Sarkis AS,
    3. Zhang ZF,
    4. Klimstra DS,
    5. Charytonowicz E,
    6. Guillem JG,
    7. Cordon-Cardo C,
    8. Cohen AM
    : p53 nuclear overexpression: An independent predictor of survival in lymph node-positive colorectal cancer patients. J Clin Oncol 12(10): 2043-2050, 1994.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Westra JL,
    2. Schaapveld M,
    3. Hollema H,
    4. de Boer JP,
    5. Kraak MM,
    6. de Jong D,
    7. ter Elst A,
    8. Mulder NH,
    9. Buys CH,
    10. Hofstra RM,
    11. Plukker JT
    : Determination of TP53 mutation is more relevant than microsatellite instability status for the prediction of disease-free survival in adjuvant-treated stage III colon cancer patients. J Clin Oncol 23(24): 5635-5643, 2005.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Komori K,
    2. Kanemitsu Y,
    3. Ishiguro S,
    4. Shimizu Y,
    5. Sano T,
    6. Ito S,
    7. Abe T,
    8. Senda Y,
    9. Misawa K,
    10. Ito Y,
    11. Uemura N,
    12. Kato T
    : Clinicopathological study of poorly differentiated colorectal adenocarcinomas: comparison between solid-type and non-solid-type adenocarcinomas. Anticancer Res 31(10): 3463-3467, 2011.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Soubeyran P,
    2. André F,
    3. Lissitzky JC,
    4. Mallo GV,
    5. Moucadel V,
    6. Roccabianca M,
    7. Rechreche H,
    8. Marvaldi J,
    9. Dikic I,
    10. Dagorn JC,
    11. Iovanna JL
    : Cdx1 promotes differentiation in a rat intestinal epithelial cell line. Gastroenterology 117(6): 1326-1338, 1999.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Pilozzi E,
    2. Onelli MR,
    3. Ziparo V,
    4. Mercantini P,
    5. Ruco L
    : CDX1 expression is reduced in colorectal carcinoma and is associated with promoter hypermethylation. J Pathol 204(3): 289-295, 2004.
    OpenUrlCrossRefPubMed
    1. Suh ER,
    2. Ha CS,
    3. Rankin EB,
    4. Toyota M,
    5. Traber PG
    : DNA methylation down-regulates CDX1 gene expression in colorectal cancer cell lines. J Biol Chem 277(39): 35795-35800, 2002.
    OpenUrlAbstract/FREE Full Text
    1. Wong NA,
    2. Britton MP,
    3. Choi GS,
    4. Stanton TK,
    5. Bicknell DC,
    6. Wilding JL,
    7. Bodmer WF
    : Loss of CDX1 expression in colorectal carcinoma: Promoter methylation, mutation, and loss of heterozygosity analyses of 37 cell lines. Proc Natl Acad Sci USA 101(2): 574-579, 2004.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Chan CW,
    2. Wong NA,
    3. Liu Y,
    4. Bicknell D,
    5. Turley H,
    6. Hollins L,
    7. Miller CJ,
    8. Wilding JL,
    9. Bodmer WF
    : Gastrointestinal differentiation marker Cytokeratin 20 is regulated by homeobox gene CDX1. Proc Natl Acad Sci USA 106(6): 1936-1941, 2009.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Toyota M,
    2. Ahuja N,
    3. Ohe-Toyota M,
    4. Herman JG,
    5. Baylin SB,
    6. Issa JP
    : CpG island methylator phenotype in colorectal cancer. Proc Natl Acad Sci USA 96(15): 8681-8686, 1999.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Weisenberger DJ,
    2. Siegmund KD,
    3. Campan M,
    4. Young J,
    5. Long TI,
    6. Faasse MA,
    7. Kang GH,
    8. Widschwendter M,
    9. Weener D,
    10. Buchanan D,
    11. Koh H,
    12. Simms L,
    13. Barker M,
    14. Leggett B,
    15. Levine J,
    16. Kim M,
    17. French AJ,
    18. Thibodeau SN,
    19. Jass J,
    20. Haile R,
    21. Laird PW
    : CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nat Genet 38(7): 787-793, 2006.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Hawkins N,
    2. Norrie M,
    3. Cheong K,
    4. Mokany E,
    5. Ku SL,
    6. Meagher A,
    7. O'Connor T,
    8. Ward R
    : CpG island methylation in sporadic colorectal cancers and its relationship to microsatellite instability. Gastroenterology 122(5): 1376-1387, 2002.
    OpenUrlCrossRefPubMed
  32. ↵
    1. van Rijnsoever M,
    2. Grieu F,
    3. Elsaleh H,
    4. Joseph D,
    5. Iacopetta B
    : Characterisation of colorectal cancers showing hypermethylation at multiple CpG islands. Gut 51(6): 797-802, 2002.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Issa JP,
    2. Ottaviano YL,
    3. Celano P,
    4. Hamilton SR,
    5. Davidson NE,
    6. Baylin SB
    : Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat Genet 7(4): 536-540, 1994.
    OpenUrlCrossRefPubMed
    1. Ahuja N,
    2. Li Q,
    3. Mohan AL,
    4. Baylin SB,
    5. Issa JP
    : Aging and DNA methylation in colorectal mucosa and cancer. Cancer Res 58(23): 5489-5494, 1998.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Nakagawa H,
    2. Chadwick RB,
    3. Peltomaki P,
    4. Plass C,
    5. Nakamura Y,
    6. de La Chapelle A
    : Loss of imprinting of the insulin-like growth factor II gene occurs by biallelic methylation in a core region of H19-associated CTCF-binding sites in colorectal cancer. Proc Natl Acad Sci USA 98(2): 591-596, 2001.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Issa JP,
    2. Ahuja N,
    3. Toyota M,
    4. Bronner MP,
    5. Brentnall TA
    : Accelerated age-related CpG island methylation in ulcerative colitis. Cancer Res 61(9): 3573-3577, 2001.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Ye C,
    2. Shrubsole MJ,
    3. Cai Q,
    4. Ness R,
    5. Grady WM,
    6. Smalley W,
    7. Cai H,
    8. Washington K,
    9. Zheng W
    : Promoter methylation status of the MGMT, hMLH1, and CDKN2A/p16 genes in non-neoplastic mucosa of patients with and without colorectal adenomas. Oncol Rep 16(2): 429-435, 2006.
    OpenUrlPubMed
  38. ↵
    1. Min YS,
    2. Yi EH,
    3. Lee JK,
    4. Choi JW,
    5. Sim JH,
    6. Kang JS,
    7. Kim YN,
    8. Juhnn YS,
    9. Kim HR,
    10. Ye SK
    : CK20 expression enhances the invasiveness of tamoxifen-resistant MCF-7 cells. Anticancer Res 32(4): 1221-1228, 2012.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Reprints and Permissions
Expression of Cytokeratin 20 Indicates Invasive Histological Phenotype in Poorly Differentiated Colorectal Adenocarcinoma
YASUO IMAI, HIDETSUGU YAMAGISHI, KAZUNORI FUKUDA, TAKUYA OKAMURA, YUKO ONO, SHINICHI BAN, TOHRU INOUE, YOSHIHIKO UEDA
Anticancer Research Jan 2014, 34 (1) 159-167;
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords