The difference between cytokeratin 20 expression in high- and low-grade urothelial bladder carcinomas: A cross-sectional study

Abstract

Background:

Bladder cancer is one of the most common cancers worldwide. Expression of cytokeratin 20 (CK 20) could be used as a biomarker in different epithelia to determine malignancy, especially in gastrointestinal, urinary tract, and Merkel cells. CK 20 could be detected in several urothelial carcinomas and was associated with bladder cancer recurrence. The study aimed to assess the utility of CK 20 expression for bladder cancer grading.

Materials and Methods:

This was a retrospective study assessing CK 20 expression in 73 bladder cancer patients who had transurethral resection of bladder tumor or cystectomy. The data were then collected and analyzed with SPSS Statistics version 20.0.

Results:

Fifty-six (76.7%) cases of high- and 17 (23.3%) cases of low-grade urothelial bladder cancer were examined for CK 20 expression. Positive expression was present in 57 (78.1%) samples. A significant difference (P = 0.034) in CK 20 expression was observed between low-grade and high-grade urothelial carcinomas. Positive expression was seen in 44 (77.2%) high-grade cases and only 13 (22.8%) low-grade cases.

Conclusion:

The difference in the CK 20 expression was found to be statistically significant among different grades of bladder cancer but not to metastatic bladder cancer. Further, studies are required to establish CK 20 as a diagnostic tool. We suggest a combination with several markers to compare which is superior.

INTRODUCTION

Bladder cancer is the 10^th most diagnosed cancer worldwide, with approximately 573,000 new cases and 213,000 deaths.[1] Clinically, pathological staging of bladder cancer can be performed using the conventional tumor, lymph node, and metastasis (TNM) cancer staging. However, this classification method is inaccurate in predicting the evolution and progression of bladder cancer whether it will progress aggressively as in high-grade tumors or slowly as in low-grade tumors.[2] As a result, molecular marker becoming increasingly popular as a reliable tool for differentiating cancer grade and predicting prognosis. This is possible as these markers are able to identify genetic mutations correlated with high-grade tumors.[3] In bladder cancer, many genetic mutations, including tumor protein 53 and fibroblast growth factor receptor 3 mutations, have been associated with tumorigenesis and epithelial–mesenchymal transition of tumor cells, triggering their progression toward becoming invasive tumors.[4,5] Among these, one of the more specific markers is cytokeratin 20 (CK 20).[4-6] The low molecular weight of CK 20 was specifically expressed in the superficial epithelial and intermediate cells of the normal urothelium.[7] Expression of CK 20 could be used as a biomarker to determine malignancy, especially in gastrointestinal and urothelial carcinomas, and was associated with bladder cancer recurrence.[8-11] Thus, we intended to investigate and assess the diagnostic value of CK 20 as a tool for differentiating between low-grade and high-grade bladder cancers and to predict distant metastases in bladder cancer patients, especially among the Indonesian population.

MATERIALS AND METHODS

Study design and population

This retrospective study used consecutive sampling and involved 73 bladder urothelial carcinoma patients from our center, H. Adam Malik General Hospital and Universitas Sumatera Utara Hospital during 2013–2015, who had undergone transurethral resection of the bladder tumor (n = 66 patients) and cystectomy (n = 7 patients). Data collection includes clinical and demographical information, which includes age, gender, grade, stage, history of smoking, and postoperative chemotherapy of the patients. The 2017 American Joint Committee on Cancer TNM system was used for pathologic staging. Lymph node (N) metastases were assessed from intravenous-contrast abdominal computed tomography (CT), and the metastatic stage was evaluated from the chest X-ray, liver ultrasonography, bone scan, and brain CT. Pathology reports and formalin-fixed paraffin-embedded specimens were requested from the treating physician/pathology laboratories. They were reviewed by one pathologist, and an immunohistochemical examination was carried out for CK 20.

The authors were fully responsible for all actions across the work and ensured that the questions presented were appropriately investigated and resolved. The study was done as per the Declaration of Helsinki (as revised in 2013) and was reported according to the STROBE guidelines. The study was approved by the Institutional Ethics Board of Universitas Sumatera Utara (NO.:500/TGL/KEPK FK USU-RSUP HAM/2017) and individual consent for this retrospective analysis was waived.

Methods

CK 20 antibodies were used for the immunohistochemical analysis. Three to four micrometers of paraffin sections were placed on poly-L-lysine-coated glass slides and kept at 37° C for 24 h. This was followed by a dehydration process at 60° C for 45 to 60 min and deparaffinized in xylene and then rehydrated using graded alcohol. The slides were then rinsed with water. Tris ethylenediaminetetraacetic acid buffer was used to treat the slides in pH 9.0 and microwaved with high temperature to retrieve the antigen. This was repeated for 10 × 3 times. The slides were incubated at room temperature for 10–15 min before they are washed in distilled water. The endogenous peroxidase was blocked for 5 min with blocking solution. These steps were then followed by treatment with tris buffer in pH 7.6 and CK 20 antibodies for 30 min in room temperature. The slides were then washed by tris buffer and polymer. Diaminobenzidine was used to stain the slides for 5–7 min, then washed under water, and counterstained in hematoxylin. Finally, they were dehydrated with graded alcohol, and xylene phenol and xylene were used to clear the slides.[12]

Statistical analysis

The Statistical analysis for the prevalence of CK 20 expression by patient demographics, characteristics, tumor histology, and stage was performed using the SPSS software for Windows version 20.0 (IBM Corp. Released 2011. Version 20.0. Armonk, NY, USA: IBM Corp) was used to analyze the prevalence of CK 20 expression by patient demographic characteristics, tumor histology, and stage. The Chi-square test, using contingency tables, was applied to analyze the data.

RESULTS

The study involved a total of 73 patients with 58 (79.5%) males and 15 (20.5%) females. The mean age was 57.8 years. Among the 73 patients, 52 (71.2%) and 21 (28.8%) patients had high- and low-grade carcinomas, respectively. The CK 20 immunohistochemical results showed 16 (21.9%) negative and 57 (78.1%) positive results. Among 73 samples, 60 (82.2%) showed metastatic deposits in the lymph node, while 5 samples showed distant metastasis [Table 1].

Table 1.

Patients’ characteristics

Variable	Total	CK 20 expression
		Negative, n (%)	Positive, n (%)
Number of patients	73	16 (21.9)	57 (78.1)
Age, mean±SD	57.80±11.1	55.40±10.7	58.10±12.2
Sex
Men	58 (79.5)	10 (62.5)	48 (84.2)
Women	15 (20.5)	6 (37.5)	9 (15.8)
Histopathological grading
High grade	52 (71.2)	8 (50)	44 (77.2)
Low grade	21 (28.8)	8 (50)	13 (22.8)
Nodal metastasis
N0	13 (17.8)	9 (56.3)	4 (7.0)
N1–N3	60 (82.2)	7 (43.7)	53 (93.0)
Distant metastasis
M0	68 (93.2)	13 (100.0)	52 (91.2)
M1	5 (6.8)	0	5 (8.8)
Management
Radical cystectomy	7 (9.6)	4 (25.0)	3 (5.3)
Chemoradiation	66 (90.4)	12 (75.0)	54 (94.7)

CK 20: Cytokeratin 20, N: Node, M: Metastasis, SD: Standard deviation

Table 2 shows the correlation between histopathological grading and CK 20 expression. Histopathological grading was divided into high grade and low grade, while CK 20 was categorized into negative and positive. The results showed that there were 44 (77.2%) patients with both positive and high-grade tumors, and 13 (22.8%) patients with both positive and low-grade tumors. The difference in expression between low-grade and high-grade urothelial carcinomas was statistically significant (P = 0.034). Five patients showed distant metastases, which were all positive for CK 20 expression. There was no statistical significance difference between CK 20 expression and distant metastasis (P = 0.57) showed in Table 3.

Table 2.

Correlation of cytokeratin 20 with tumor grading

Tumor grade	CK 20, frequency (%)		P
	Negative	Positive
High grade	8 (50.0)	44 (77.2)	0.034*
Low grade	8 (50.0)	13 (22.8)
Total	16 (100)	57 (100)

*Statistical analysis was carried out using the Chi-square test where P is statistically significant at <0.05 level. CK 20: Cytokeratin 20

Table 3.

Correlation of cytokeratin 20 with distant metastasis

Distant metastasis	CK 20, frequency (%)		P
	Negative	Positive
M0	16 (100)	52 (91.2)	0.57
M1	0	5 (8.8)
Total	16 (100)	57 (100)

*Statistical analysis was carried out using Fisher’s exact test where P is statistically significant at <0.05 level. CK 20: Cytokeratin 20, M: Metastasis

DISCUSSION

Cytokeratins are intermediate filament proteins. There are 20 different cytokeratin isotopes, of which 2–10 isotopes were expressed by epithelial cells that can be used to reflect its type and differentiation status and hence used in tumor diagnosis.[13,14] CK 20, a type of cytokeratin, has a low molecular weight initially and is identified as a protein in the intestinal epithelial.[15] CK 20 exhibits several interesting features such as lower acidity compared to other cytokeratins and its limited expression in normal tissues as it is only expressed in epithelium within the gastrointestinal or urinary tract and Merkel cells.[16] Transitional epithelium or urothelium, found in the urinary tract, showed changes in cytokeratin expression that was associated with stratification and differentiation.[17] A CK 20-positive pattern was expressed as a collection of more than five cells in the intermediate and/or the deep layers of urothelium stained with moderate or strong cytoplasmic stain reaction.[14] A positive result also includes cases with a strong diffusion of cytoplasmic response in well-differentiated cell cancer including high-grade tumors. A negative CK 20 pattern was expressed as a total absence or weak-moderate stain reaction on the occasional cells of the superficial layer (umbrella cells layer) [Figure 1].[14]

Figure 1.

The result from pathology anatomy examination of cytokeratin 20, (a) Negative cytokeratin 20 expression of transitional cell carcinoma of the bladder, (b-d) Positive of cytokeratin 20 expression of transitional cell carcinoma of the bladder

Prior studies had shown that CK 20 could be used as an accurate marker of urothelial bladder cancer differentiation degree and was suggested as a tool in diagnosing noninvasive papillary urothelial bladder cancer. The dysregulation of CK 20 expression observed in primary urothelial bladder cancer showed urothelial differentiation and signified the initial occurrence in tumor dysgenesis. A strong correlation between abnormal CK 20 expression and high-grade tumor was shown. Nonmuscle-invasive urothelial bladder cancer has also demonstrated an association between CK 20 expression and tumor recurrence and disease-specific survival.[12,15,17]

Several studies have compared CK 20 expression in urothelial tumors of different grades with various parameters and new immunohistochemical markers. A study by Desai et al. showed consistency with our results that CK 20 positivity was associated with grading and staging of tumor. The prevalence of high-grade tumors which showed CK 20 positivity was 69.4% compared to 45% of the low-grade tumor group.[18] In line with Desai et al., Mumtaz et al. reported a significant difference in expression of CK 20 between low- and high-grade urothelial carcinomas, with a total of 48 (50.5%) cases of low-grade tumor and 47 (49.5%) cases of high-grade urothelial tumor. Thirty-three (68.8%) patients of high-grade and 19 (40.4%) low-grade urothelial bladder cancer were reported with diffuse CK 20-positive expression.[19] Furthermore, our study supported the previous study by Desai et al. which reported an association between higher CK 20 positivity and higher tumor grade and stage of urothelial carcinoma.[18]

In contrast with our findings, Al Shaikhly et al. reported that the difference in expression between low-grade and high-grade urothelial carcinomas was not statistically significant (P = 0.5). Positive expression was observed in 16 high-grade cases, while 17 low-grade cases showed such positivity.[14] Similarly, Harnden et al. reported in their study of 51 urothelial bladder cancer patients that there was no significant correlation between CK 20 and tumor grade (P = 0.065).[10] The contrast in the results could be attributed to the different study designs used, number of samples in each study, as well as the different methods used to detect CK 20 expression.

Our research additionally explored into the correlation between CK 20 expression and distant metastases. Five patients were diagnosed with distant metastases, and all had positive CK 20 expressions. Statistical analysis showed no significant differences (P = 0.57) between the two groups. This was consistent with a study by Ghaini et al. in evaluation of CK 20 expression of 53 bladder transitional cell carcinoma patients. The study reported an inverted relationship between CK 20 and stage.[13] In addition, Sikic et al. in their study of 222 patients with upper tract urothelial carcinomas stated no association between CK 20 expression and the presence of metastasis (P = 0.28).[20]

We acknowledged that there were limitations in the number of samples included in our study. There was also a discrepancy between the number of metastatic cases and nonmetastatic cases, which would have made a negative impact when evaluating CK 20 as a metastatic marker. We urge future studies to include more cases and to include equal number of cases between metastatic and nonmetastatic cases.

CONCLUSION

The expression of CK 20 was statistically significant in the bladder cancer grading group but not in the metastatic bladder cancer group. Future studies are required to establish CK 20 as a diagnostic. The use of several other markers could be used to assess which markers were superior and which ones can be combined.

Financial support and sponsorship

This study was supported by the Research Institute of Universitas Sumatera Utara through TALENTA Universitas Sumatera Utara 2017 Grant, with contract number 5338/UN5.1.R/PPM/2017.

Conflicts of interest

There are no conflicts of interest.