Expression of some selected cytokeratins and Ki67 protein in prostatic tumor: can these be used as tumor markers

Introduction Diagnosis of prostatic diseases with Immunohistochemistry still faces challenges because of the peculiar histology of the prostate and difference(s) in reactivity of Monoclonal antibodies (MoAb) to benign and malignant changes. Methods Thirty (30) archived paraffin embedded tissue samples from primary prostate tumors (15 Benign Prostatic Hyperplasia (BPH) and 15 Cancer of the prostate (CaP)) were sectioned at thickness of 5µm and confirmed as BPH or CaP. Sections from each sample were stained by Immunohistochemistry using the Streptavidin-biotin method and using CK5/6, CK7, CK8,CK20 and Ki67 antibodies (Zymed Antibody products). Appropriate positive and negative controls for each antibody were setup alongside the test slides. Results BPH samples were reactive to Ck5/6 (93.3%), Ck7 (80%) and Ck8 (100%). Only 13.3% of BPH samples were reactive to Ki67. The reactivity of Ck5/6, 7, 8 in CaP is a contrast with only 3(20%) of samples positive with Ck5/6, 2(13.3%) positive with Ck7 and 14(93.3%) with Ck8. While reactivity of Ck 8 is similar in BPH and CaP, no reaction was recorded in Ck 20 in both BPH and CaP. Ki67 was only reactive in 2(13.3) of BPH samples and 15(100%) of CaP. Only Ck 8 was expressed in both BPH and CaP. There was co-expression of Ck5/6, 7,8 and Ki67 in13.3%; Ck7and Ki67 in 13.3% in both BPH and CaP. Conclusion The various cytokeratins are individually expressed in both BPH and CaP. Ck5/6 and Ck7 are co-expressed and may be used in the diagnosis of BPH, Ck5/6,7,8 and Ki67 are co-expressed in Prostatic adenocarcinoma and squamous cell carcinoma of the prostate while Ck8 and Ki67 are co-expressed and may be used for diagnosis of Prostatic adenocarcinoma alone.


Introduction
The prostate is the site of two of the most common diseases in elderly men, Benign Prostatic Hyperplasia (BPH) and Prostate cancer (CaP). Both of these conditions are disorders of cell differentiation and cell proliferation [1]. Prostate cancer is the second most frequently diagnosed cancer as well as the sixth leading cause of death in males worldwide [2]. Little is known about the basic biology of cell phenotypes in either normal prostate or prostatic cancers. Phenotypes that are intermediate between those of basal and luminal cells in the heterogeneous prostate epithelial cells have been reported [3]. The prostate gland contains epithelial cells expressing two major phenotypes: luminal and basal cells separated by basement membrane from the stroma. If appropriate makers can be used, characterization of the phenotypes of cells will thus represent a major advantage [4].
The variety of morphologic patterns of different entities of the genitourinary tract can represent a diagnostic dilemma for the pathologist. This is especially true in cases that mimic cancer, in cancer of unknown primary or poorly differentiated tumors in which it is hard to assign histiogenesis needed to plan the correct therapy for the patient [5]. It has also observed that routine staining of prostatic lesions sometimes cause diagnostic dilemma especially in premalignant lesions like atypical adenomatous hyperplasia and prostatic intraepithelial neoplasia (PIN) [6].
The use of Immunohistochemistry in diagnosis is presently widespread because of its sensitivity and specificity. However, diagnosis of prostatic diseases with Immunohistochemistry is still facing challenges because of peculiar histology of the prostate and difference(s) in reactivity to Monoclonal antibodies (MoAb) by benign and malignant changes.
The expression of cytokeratins in prostatic epithelium is varied and is well recognized [7]. It has been reported that Immunohistochemistry offers a better capacity than Haematoxylin and eosin staining alone and its addition to the diagnostic armamentarium for genitourinary pathologic diagnosis has increased the sensitivity and specificity of diagnosis and aided in the selection of optional therapeutic regimens in selected cases [5]. The first study utilizing anti keratin monoclonal antibodies was reported [8].
Others reported immunoreactivity of keratin in basal cell of normal and hyperplastic prostatic epithelium with no staining of adenocarcinoma [9,10]. Conversely, keratin was identified in one case of prostatic adenocarcinoma using polyclonal anti serum raised against bovine muzzle pre-keratin [11]. It was further reported that keratin immunoreactivity differs in the two epithelial cells of the prostate, probably due to expression of different keratin proteins [12].
Several biological markers have been reported in literature to be good objective markers of progression of different neoplasms. The Ki67 antigen is a useful proliferation marker [13]. The potential of Ki67 antibody for diagnostic purpose was investigated, assessing the proliferation activity in normal prostate tissue, Prostatic Intraepithelial Neoplasia and Prostatic Cancer patients [14]. Cell proliferation is a fundamental aspect of a number of prostatic diseases ranging from hyperplasia to neoplasia [14]. It was reported that cellular proliferation can be studied using Immunohistochemical marker directed against nuclear antigen Ki67 [15].
This study therefore examined the individual expression and the then co-expression of some selected cytokeratins and Ki67 in benign and malignant prostatic biopsies with a view to find out differential, combinations of individual or co-expressed proteins that could make diagnosis more specific and precise.

Methods
Thirty (30) archived paraffin embedded tissue samples from primary prostate tumors comprising 15 benign prostatic hyperplasia (BPH) and 15 cancer of the prostate (CAP) were sectioned at thickness of 5µm. One section from each sample was stained by the Haematoxylin and eosin method to reconfirm diagnosis of BPH and CaP. Subsequently 2µm sections from each sample were stained by Immunohistochemistry using the Streptavidin-biotin method using the following antibodies CK5/6, CK7, CK8, CK20 and Ki67 (Zymed Antibody products). Negative and positive controls for each antibody were set up alongside the test slides.The positive controls for Ck5/6, 7 and 8 was the normal ureter tissue which are known to express the antigens, while the control for Ck 20 was malignant colonic tissue which expresses the antigen. The control for Ki 67 was a reactive lymphoid hyperplastic tissue. For the negative controls, test tissue sections were used but the antibodies were replaced at the appropriate stage by Tris buffered saline to confirm that all antibodies are reactive and there was no cross reactivity. All Page number not for citation purposes 3 sections were dewaxed and hydrated. They were subsequently heated at 1210C for 10minutes in 10mmol/L Citrate buffer (pH 6.0) and the antibodies applied for 45minutes after passing through peroxidise and protein blocks. The secondary biotin-labeled antibody was incubated for 15min at room temperature. The streptavidin labelled streptavidin-biotin amplification method was carried out for 30minutes followed by 3,3'-Diaminobenzidine (DAB) and haematoxylin counter stain were subsequently applied in sequence.
Reactivity was observed as brown coloration. Photomicrographs of reactive sections were taken.
The necessary institutional ethical clearance was sought and obtained from Meena Histopathology laboratory.

Results
The percentage of cases that expressed each of the monoclonal antibodies (MoAb) in BPH and CaP are presented in Table 1

Discussion
The co-expression and individual expression of CK5/6, Ck7, Ck8, Ck20 and Ki67 in BPH and CaP was examined. This information intends to make distinction between benign and malignant diseases of the prostate clearer. Our study shows that majority of BPH samples were reactive to Ck5/6 (93.3%), Ck7 (80%) and Ck8 (100%) ( Table 1) and is agreeable with previous research [16].
Only 13.3% of samples were reactive to Ki67. This is understandable because BPH is benign and Ki 67 is known to be a proliferative MoAb and is also consistent with previous reports [10,17]. The observation of cytokeratins (Ck5/6and7) in a few CaP samples was noted. It was however, observed that this occurred only in poorly differentiated squamous cell carcinoma. The only case that was not reactive with the cytokeratins is the well differentiated squamous cell carcinoma. Other researchers [18,19] similarly reported that cytokeratins are differentially expressed in normal as well as abnormal prostate epithelia. This is significant when coexpression of monoclonal antibodies is used for differential diagnosis. has been reported [20] to be absent in the prostate. The presence of CK8 in both BPH and CaP is said to be normal to the prostate stem cells therefore cannot be used as a diagnostic biomarker [21] Ki67 was only reactive in 2(13.3) of BPH samples and 15(100%) of CaP. The distinction between the two is that reactivity in BPH was weak while in CaP reactions were strong. This finding is consistent with that of previous studies and has been reported as a proliferation marker with demonstrated usefulness in many neoplasms including the prostate. It is used in determining their progression [22][23][24]. Figure   2 and Ck5/6, 7,8 and Ki67 in13.3%; Ck7and Ki67 in 13.3% that cuts across BPH and CaP. It was also, observed that these coincided with samples that had a diagnosis of squamous cell carcinoma. This is significant in the differential diagnosis of squamous cell carcinoma of the prostate and Prostatic Adenocarcinoma.

Conclusion
Definitive diagnosis of prostatic tumors using individual expression and co-expression of monoclonal antibodies is critical to the selection of options for therapeutic regimens and therefore the information obtained can be used although to a limited extent to make diagnosis more accurate. Consequently, the pathologist needs to be knowledgeable in the use of co-expressed antibodies to make decisions in diagnosis.

Competing interests
The authors declare no competing interest.