sign in sign up
<<
Home , Prostate cancer

ANTICANCER RESEARCH 28 : 4149-4154 (2008)

Immunohistochemical Detection of - specific Antigen Expression in Primary Urothelial of the JUNG-CHIN CHEN 1,2 , CHUNG-LIANG HO 1,3 , HUNG-WEN TSAI 1,3 , TZONG-SHIN TZAI 4, HSIAO-SHENG LIU 5, NAN-HAW CHOW 1,3 , WEN-HORNG YANG 4 and HONG-LIN CHENG 4

1Department of , National Cheng Kung University Hospital, Tainan 704; Departments of 2Medical Laboratory Science and Biotechnology, 3Pathology, 4Urology, and 5Microbiology and Immunology, College of Medicine, National Cheng Kung University, Tainan 701, Taiwan, R.O.C.

Abstract. Background: Prostate-specific antigen (PSA) is only serum biomarker recommended by the American a 33-kDa serine protease that is secreted by prostatic Society for use in the of (2). epithelium and non-prostate tissues. Ectopic expression of In addition, immunohistochemical analysis of PSA PSA has also been reported in a variety of non-prostatic expression in tumor tissue is generally successfully applied epithelial tumors. Patients and Methods: Expression of PSA in the identification of the prostatic origin of metastatic and its clinical implication were examined in a total of 422 carcinoma (3) and the distinction between poorly cases of primary urothelial carcinoma of the bladder. differentiated prostatic cancer and urothelial carcinoma in the Results: Cytoplasmic reactivity to PSA was observed in 6 bladder neck region (3-5). cases (1.4%), with the staining being to a low degree in 3 Despite fact that PSA may be a specific marker for and a high degree in the remaining 3 cases. Expression of prostatic differentiation (3-5), ectopic expression of PSA has PSA was positively related to multiplicity, large tumors (≥3 been reported in a variety of non-prostatic tumors (6-16), cm) and muscle invasion (≥pT2) (p=0.0008, 0.004 and 0.03, most commonly in (6-9) and salivary gland respectively). Patients with tumors of low PSA expression (10-12). Other non-prostatic tumors reported to had a better survival than those with tumors of high PSA express PSA include of the urinary bladder expression (p=0.03). Conclusion: Focal expression of PSA (13), melanomas (14), acinar cell carcinoma of the pancreas can occasionally be detected in some large, invasive (15) and lung cancer (17, 18). Interestingly, Mai et al. urothelial cancer cells. This phenotypic change should be demonstrated a phenotypic alteration in the urothelial considered in the of poorly carcinoma of the bladder when it spreads into the prostate. differentiated carcinoma of the lower urinary tract. All ten primary urothelial tumors of urinary bladder were negative for PSA, whereas six tumors invasive to the prostate Prostate-specific antigen (PSA) is a 33-kDa serine protease were focally positive (19). Urothelial cancer cells thus might that is secreted by prostatic epithelium and non-prostate exhibit PSA expression through mesenchymal−epithelial tissues, such as epithelial lining of the periurethral and interaction in the prostatic stroma. However, small amounts perianal glands, urachus remnant and cowper’s glands (1). It of PSA mRNA were also detected in the normal human functions in the liquefaction of seminal coagulum to allow bladder and in the cancerous tissue by reverse transcriptase the release of spermatozoa. Because of its role in the polymerase chain reaction (RT-PCR) (20). We thus detection and prognosis of prostate cancer, PSA remains the performed this cohort study to examine the expression of PSA and its potential clinical implication in conventional urothelial carcinoma of the bladder.

Correspondence to: Dr. Hong-Lin Cheng, Department of , Patients and Methods National Cheng Kung University Hospital, 138 Sheng-Li Road, Tainan 70428, Taiwan, R.O.C. Tel: +886 6 2353535 ext. 5271, Fax: Cell lines and culture. Immortalized E6 human uroepithelial cells and +886 6 2383678, e-mail: [email protected] five human bladder cancer cell lines (RT4, TSGH-8301, TCC-SUP, J82, and T24) were propagated as described elsewhere (21). The Key Words: Prostate-specific antigen, urothelial carcinoma, invasion, localized established cell lines UB09, UB37, UB40 and UB47 were prognosis, PSA. derived from transitional cell bladder cancer of grade II pT2, grade III

0250-7005/2008 $2.00+.40 4149 ANTICANCER RESEARCH 28 : 4149-4154 (2008) pT3, grade III pT1 and grade III pT3, respectively. They were therapy or by a radical or partial cystectomy if there was disease maintained in Dulbecco’s modified Eagle’s medium (Gibco BRL, progression. Disease progression at recurrence was defined as being Grand Island, NY, USA) supplemented with 10% fetal calf serum at a higher stage than the previous result. Fifty-one patients with (FCS; Gibco BRL), L-glutamine (2 mmol/ l), sodium pyruvate muscle-invasive tumors were given a radical cystectomy. Forty-five (110 mg/ l), penicillin (100 U/m l), and streptomycin (50 g/m l) at 37˚C patients received systemic with methotrexate, cisplatin, in a 5% CO 2-humidified atmosphere. The -sensitive, PSA- doxorubicin, and vinblastine. Patient survival status was determined producing human prostate cancer cell line, LNCaP, was chosen as by checking outpatient clinic records, interviewing patients’ families, positive control for experiments in vitro . Cells were grown in RPMI- or both. Clinical follow-up ranged from 8 to 17 years (median, 10 1640 supplemented with L-glutamine, non-essential amino acids, 10% years). FCS (Life Technologies), and (200 U/ml penicillin and 100 mg/ml streptomycin). (IHC) of PSA expression. To preclude the possibility of sampling bias, histological sectioning, rather than RNA preparation and RT-PCR. The RT-PCR assay and the primer tissue array, of surgical specimens was chosen for this investigation. sets for PSA have been described previously (22). Briefly, the Immunostaining procedures were previously described (21). Briefly, upstream primer sequence was: 5’-GATGACTCCAGCCACGACCT- tissue sections were incubated at room temperature for 2 h with 3’; and the downstream sequence was: 5’-CACAGACACCC monoclonal anti-PSA antibody (Dako Corp., Carpenteria, CA, CATCCTATC-3’. The PCR reaction conditions included 2 μM USA) (19). The optimal dilution (1/25) was determined using dig-11-dUTP and 0.4 μM PSA primers. The PCR was performed for normal human prostate as a control. The StrAviGen Super Sensitive 35 cycles consisting of the following steps: denaturation at 94˚C for MultiLink kit (BioGenex Laboratories, Inc., San Ramon, CA, USA) 1 min; annealing at 66˚C for 1 min; and extension at 72˚C for 2 min. was used to detect the resulting immune complex. Peroxidase Gel analysis was then examined for PCR product spanning 710 bp, activity was visualized by an aminoethyl carbazole substrate kit and the glyceraldehyde-3-phosphate dehydrogenase housekeeping (Zymed Laboratories, Inc., San Francisco, CA, USA). Sections was used to assess the mRNA integrity. LNCaP was used to were finally counterstained with hematoxylin. For negative control, evaluate the sensitivity of PCR experiments. Internal negative control non-immune mouse immunoglobulin was substituted for primary for RT-PCR was performed using all of the reagents in the absence antibody in the incubation. of cDNA. To evaluate the PSA staining, both reviewers (N-H.C. and H-L.C.) were blinded to the clinical outcomes. Since there was no apparent difference of staining intensity, a three-category scoring Clinicopathologic al characteristics. As approved by the institutional system was modified from previous reports (19, 23). "High review board, patients who underwent transurethral resection of expression (++)" indicates that more than 10% of the tumor cells bladder tumor or radical cystectomy for urothelial carcinoma of the exhibited PSA staining; "low expression (+)", between 0% and bladder at National Cheng Kung University Hospital, Tanian, 10% reactivity; and "negative (−)", no immunoreactivity for PSA Taiwan, ROC were identified using a database. Paraffin-embedded was detected. For negative control, different types of non-prostatic specimens between 1990 and 1999 were retrieved from the tissue were tested. There was no staining in positive controls when pathology archives. Those patients with primary carcinoma in situ , treated with non-immune sera in place of the PSA primary antibody. urothelial crcinoma with glandular lumen or prostatic involvement, primary urothelial carcinoma of the prostate, and mixed carcinoma Statistics. were excluded. A total of 422 patients with conventional urothelial Associations of PSA expression and clinicopathological indicators of bladder cancer were examined using Fisher’s exact or carcinoma of the bladder were recruited for this investigation. There 2 were 256 men and 166 women, ranging in age from 23 to 93 years χ test. The difference of survival in relation to PSA expression was t- old (mean age, 64.2±10.4 years). All cases were reviewed for examined by test. All measures of association were calculated by p histological according to the World Health Organization two -t ail test. Only those variables with a -value <0.05 were classification (1973). Clinical staging was determined according to considered significant. the Tumor-Node- System of the Union International Contre Cancer (2002) using a survey of the clinical details, image Results studies and pathological data. The treatment and follow-up of patients were conducted according Non-neoplastic urothelium with/without dysplastic change to the standard strategy previously described in detail (21). Briefly, all (n=62) did not reveal PSA staining. For comparison, urothelial patients with superficial bladder cancer (n=371) received transurethral resection and postoperative intravesical chemotherapeutic agent carcinoma of the bladder neck, invasive to the prostate, or instillation with either thiotepa (30 mg in 30 m l of normal saline for 70 adenocarcinoma (n=9) were examined for PSA expression. As patients) or epirubicin (40 mg in 40 m l of normal saline for 34 reported by Mai et al. (19), focal cytoplasmic PSA reactivity patients) weekly for 8 consecutive weeks. Those patients who received was observed in some of cancer cells (data not shown). intravesical bacillus Calmette-Guerin therapy or neoadjuvant A total of 422 cases of primary urothelial carcinoma of the chemotherapy were excluded from this study to preclude the potential bladder were examined for PSA expression. Cytoplasmic bias in prognostic correlation ( e.g. recurrence rate or tumor reactivity to PSA (Figure 1) was observed in some tumor cells progression). The patients were followed up every 3 months for the first 2 years, every 6 months for an additional 2 years, and yearly in 6 cases (1.4%). The staining was low and focal in 3 cases thereafter. Recurrent tumours were confirmed by and the while high expression was exhibited in the remaining 3 cases patient was given a transurethral operation. This was followed by (Table I). In addition, these 6 tumors were located in posterior, another 8-week course of intravesical chemotherapeutic instillation lateral, trigone, or dome areas of the bladder. None of these

4150 Chen et al : PSA in Urothelial Carcinoma

Figure 1. Immunolocalization of PSA in the cytoplasm of urothelial cancer cells. Expression of PSA protein was demonstrated in the cytoplasm of urothelial carcinoma cells highlighted by the arrows ( magnification ×150).

Table I. Clinicopathological characteristics of patients with PSA-positive urothelial carcinoma.

Tumor

Number Age/ Gender Grade Stage Size (cm) Location PSA Status

1 62/M III pT1 3.7 Lateral, trigone + Dth (33)* 2 74/M III pT1 3.5 Posterior + Dth (58) 3 70/F II pT2a 3.0 Right posterior ++ Dth (8) 4 68/F II pT3 3.5 Posterior, dome ++ Dth (20) 5 74/F II pT2a 6.0 Lateral, posterior ++ Dth (22) 6 64/M III pT2a 4.5 Lateral, posterior + Dth (62)

*Death at indicated months after surgery.

PSA-positive tumors were at the bladder neck, in which PSA tumors exhibited stromal invasion ( p= 0.0002). In addition, expression may be induced through mesenchymal –epithelial patients with tumors expressing low PSA had a better interaction in the prostatic stroma (23). survival (mean 51±15.7 months) than those with tumors In terms of clinical correlate, positive PSA expression was expressing high PSA (16.7±7.6 months) ( p= 0.03) (Table I). positively related to multiplicity, large tumors (≥3 cm) and These data show that PSA appears to play a positive role in muscle invasion (≥pT2) of urothelial cancer ( p= 0.0008, the progression of human bladder cancer, although the 0.004 and 0.03, respectively) (Table II). All PSA-positive phenotypic alteration is rare.

4151 ANTICANCER RESEARCH 28 : 4149-4154 (2008)

Table II. Pertinent clinicopathologic al characteristics of study cases. reagents. Mai et al. proposed that mesenchyma−epithelial interaction induced by prostatic stroma may play a role in Factor No. of No. of PSA(+) p-Value cases tumors the unusual PSA immunoreactivity in bladder (19, 23). This interpretation, however, is not valid in our cases, Grade 0.08 since all these PSA-positive urothelial tumors were located I 60 0 (0.0%) in the posterior, lateral, trigone, or dome areas of the bladder. II 288 3 (1.0%) In addition, non-mucinous glandular differentiation in III 74 3 (4.5%) Stage 0.03 urothelial carcinoma was not included in our study. Thus, the pTa 322 0 (0.0%) molecular mechanism underlying PSA expression in pT1 49 3 (6.1%) urothelial carcinoma remains to be elucidated. pT2a-b 28 2 (7.1%) Tremblay et al. showed that PSA (HK3) antiserum may PT3 12 1 (8.3%) cross-react with other members of the kallikrein family, pN1/M+ 11 0 (0.0%) Size 0.004 especially with HK2 (24). This possibility is being tested in <1 cm 63 0 (0.0%) vitro . Alternatively, prostatic differentiation can be detected 1–3 cm 189 0 (0.0%) in the cystitis cystica/glandularis of the bladder mucosa, >3 cm 170 6 (3.5%) whereas adjacent surface uroepithelium or transitional cell Shape 0.40 nests were negative for PSA (16). The phenomenon suggests Papillary 342 4 (1.2%) Non-papillary 80 2 (2.5%) a broad metaplastic potential in various differentiated but Number 0.0008 embryologically related tissues (16). Therefore, prostatic Singe 347 1 (0.2%) differentiation in the urinary tract may not be “ectopic” or as Multiple 75 5 (6.7%) rare as previously proposed (25, 26). A well-known metaplastic potential of urothelial cancer is the production of human chorionic gonadotropin (hCG) (27, 28). Clinical investigation of bladder cancer cohort revealed Since expression of PSA mRNA was demonstrated in the hCG expression in from 10% to 36% of cancer cells (28- normal and cancerous bladder tissue (20), we examined the 30) and the status was positively related to high grade or expression of PSA in a total of ten uroepithelial cell lines in advanced tumor staging (27, 28). The prevalence of this vitro by RT-PCR. Except for LNCaP positive control cells, phenotype in the highly proliferating areas implies that hCG none of the cell lines, including immortalized E6 human may play a positive role in the progression of bladder cancer uroepithelial cells, revealed evidence of PSA transcription (data (28, 31). Taken together with the facts that PSA-positive not shown). Thus, expression of PSA does not exist in most tumors were all invasive and patients with tumors expressing commonly reported human bladder cancer cell lines in vitro . high PSA have a worse prognosis, PSA may play a positive role in the progression of urothelial . Discussion One important issue in the urological diagnostic service is the differential diagnosis of primary urothelial carcinoma of PSA is the most important and clinically useful marker for the bladder from that of poorly differentiated prostatic prostate cancer. In non-prostatic tissues, however, ectopic carcinoma invasive to the bladder neck (19, 32-34). IHC for expression of PSA has been observed in anal glands of male PSA can identify most of the poorly differentiated patients, urethral glands, cystitis cystica/glandularis, and in of the prostate (5, 32-36), although the a minority of nephrogenic adenoma of the prostatic staining decreases with increasing Gleason grade (36-38). (6-16). The non-prostatic neoplasms reported to express However, a small subset of poorly differentiated prostate PSA include breast cancer (6-9), salivary gland neoplasms tumors (between 3% and 15%) does not express PSA (4, 5, (10-12), adenocarcinoma of the urinary bladder (13), 36). The results of our data suggest that a marker profile melanomas (14), acinar cell carcinoma of the pancreas (15) should be included in the IHC analysis of these cases. For and lung cancer (17, 18). those tumors suspicious of prostatic origin, additional markers, In this cohort study, cytoplasmic expression of PSA was such as prostein, prostate-specific membrane antigen, NKX3.1 observed in some large, invasive urothelial tumors. Such (an androgen-related tumor suppressor gene), or proPSA expression, although rare (1.4%), is quite different from that (precursor form of PSA), should be included in the of prior reports showing absolutely negative staining for PSA investigation (5). In contrast, a marker panel containing in the urothelial carcinoma of the bladder (4, 5, 19, 23). cytokeratin 7, cytokeratin 20, 34betaE12, or p63 may be Several explanations may account for the discrepancy, such helpful when urothelial origin is considered (4, 5, 35). as differences in the primary antibody used, the efficiency of In conclusion, focal expression of PSA can occasionally antigen retrieval, or the sensitivity of immunostaining be demonstrated in some large, invasive urothelial cancer

4152 Chen et al : PSA in Urothelial Carcinoma cells in locations other than bladder neck. Th is phenotypic 13 Grignon DJ, Ro JY, Ayala AG, Johnson DE and Ordonez NG: alteration may have a prognostic implication. This fact Primary adenocarcinoma of the urinary bladder. A clinico- should be considered in the differential diagnosis of poorly pathologic analysis of 72 cases. Cancer 67 : 2165-2172, 1991. 14 Bodey B, Bodey B Jr and Kaiser HE: Immunocytochemical differentiated carcinoma of the lower urinary tract. detection of prostate-specific antigen expression in human primary and metastatic melanomas. Anticancer Res 17 : 2343- Acknowledgements 2346, 1997. 15 Kuopio T, Ekfors TO, Nikkanen V and Nevalainen TJ: Acinar This study was supported by grants NSC-96-2621-Z-006-012-MY2 cell carcinoma of the pancreas. Report of three cases. APMIS and NSC-95-2320– B-006-058-MY3 from the National Science 103 : 69-78, 1995. Council, and by grant DOH-TD-B-111-004 from the Department of 16 Cohen RJ, Garrett K, Golding JL, Thomas RB and McNeal JE: Health, Executive Yuan, Taiwan, R.O.C. Epithelial differentiation of the lower urinary tract with recognition of the minor prostatic glands. Hum Pathol 33 : 905-909, 2002. References 17 Bilgrami S, Singh NT, Shafi N and Ciesielski T: Raised prostate- specific antigen in adenocarcinoma of lung. Lancet 344 : 1371- 1 Hernández J and Thompson IM: Prostate-specific antigen: a 1372, 1994. review of the validation of the most commonly used cancer 18 Levesque M, Yu H, D’Costa M, Tadross L and Diamandis EP: biomarker. Cancer 101 : 894-904, 2004. Immunoreactive prostate-specific antigen in lung tumors. J Clin 2 Smith RA, Cokkinides V and Eyre HJ: American Cancer Society Lab Anal 9: 375-379, 1995. Guidelines for the Early Detection of Cancer, 2005. CA Cancer 19 Mai KT, Yazdi HM and Farmer J: Changes of phenotypic J Clin 55 : 31-44, 2005. expression of prostatic antigen in secondary transitional cell 3 Varma M and Jasani B: Diagnostic utility of immunohisto- carcinoma of the prostate: Evidence for induction phenomenon chemistry in morphologically difficult prostate cancer: review of as a mechanism for acquisition of prostatic antigens in prostatic current literature. 47 : 1-16, 2005. transitional cell carcinoma. Prostate 47 : 172-182, 2001. 4 Kunju LP, Mehra R, Snyder M and Shah RB: Prostate-specific 20 Ishikawa T, Kashiwagi H, Iwakami Y, Hirai M, Kawamura T, antigen, high-molecular-weight cytokeratin (clone 34betaE12), Aiyoshi Y, Yashiro T, Ami Y, Uchida K and Miwa M: Expression and/or p63: an optimal immunohistochemical panel to of alpha-fetoprotein and prostate-specific antigen in distinguish poorly differentiated prostate adenocarcinoma from several tissues and detection of mRNAs in normal circulating urothelial carcinoma. Am J Clin Pathol 125 : 675-681, 2006. blood by reverse transcriptase-polymerase chain reaction. Jpn J 5 Chuang AY, DeMarzo AM, Veltri RW, Sharma RB, Bieberich CJ Clin Oncol 28 : 723-728, 1998. and Epstein JI: Immunohistochemical differentiation of high- 21 Cheng HL, Liu HS, Lin YJ, Chen HW, Hsu PY, Chang TY, Ho grade prostate carcinoma from urothelial carcinoma. Am J Surg CL, Tzai TS and Chow NH: Co-expression of RON and MET is Pathol 31 : 1246-1255, 2007. a prognostic indicator for patients with transitional cell 6 Bodey B, Bodey B Jr and Kaiser HE: Immunocytochemical carcinoma of the bladder. Br J Cancer 9: 1906-1914, 2005. detection of prostate-specific antigen expression in human breast 22 Shariat SF, Gottenger E, Nguyen C, Song W, Kattan MW, carcinoma cells. Anticancer Res 17 : 2577-2581, 1997. Andenoro J, Wheeler TM, Spencer DM and Slawin KM: 7 Alanen KA, Kuopio T, Collan YU, Kronqvist P, Juntti L and Preoperative blood reverse transcriptase-PCR assays for Nevalainen TJ: Immunohistochemical labelling for prostate- prostate-specific antigen and human glandular kallikrein for specific antigen in breast . Breast Cancer Res Treat prediction of prostate cancer progression after radical 56 : 169-176, 1999. . Cancer Res 62 : 5974-5979, 2002. 8 Hall RE, Clements JA, Birrell SN and Tilley WD: Prostate- 23 Nowels k, Kent E, Rinsho K and Oyasu R: Prostate-specific specific antigen and gross cystic disease fluid protein-15 are antigen and acid phosphatase-reactive cells in cystitis cystica and coexpressed in -positive breast tumours. Br J glandularis. Arch Pathol Lab Med 112 : 734-737, 1988. Cancer 78 : 360-365, 1998. 24 Tremblay RR, Deperthes D, Tetu B and Dube JY: Immuno- 9 Howarth DJ, Aronson IB and Diamandis EP: Immunohisto- histochemical study suggesting a complementary role of chemical localization of prostate-specific antigen in benign and kallikreins hK2 and hK3 (prostate-specific antigen) in the malignant breast tissues. Br J Cancer 75 : 1646-1651, 1997. functional analysis of human prostate tumors. Am J Pathol 150 : 10 Tazawa K, Kurihara Y, Kamoshida S, Tsukada K and Tsutsumi 455-459, 1997. Y: Localization of prostate-specific antigen-like immuno- 25 Gutierrez J and Nesbit RM: Ectopic prostatic tissue in bladder. J reactivity in human salivary gland and salivary gland tumors. Urol 98 : 474-478, 1967. Pathol Int 49 : 500-505, 1999. 26 Klein HZ and Rosenberg ML: Ectopic prostatic tissue in bladder 11 Fan CY, Wang J and Barnes EL: Expression of androgen trigone: Distinctive cause of . Urology 23 : 81-82, 1984. receptor and prostatic-specific markers in salivary duct 27 Lazar V, Diez SG, Laurent A, Giovangrandi Y, Radvanyi F, Chopin carcinoma: an immunohistochemical analysis of 13 cases and D, Bidart JM, Bellet D and Vidaud M: Expression of human review of the literature. Am J Surg Pathol 24 : 579-586, 2000. chorionic gonadotropin beta subunit genes in superficial and 12 Holmes GF, Eisele DW, Rosenthal D and Westra WH: PSA invasive bladder carcinomas. Cancer Res 55 : 3735-3738, 1995. immunoreactivity in a parotid oncocytoma: a diagnostic 28 Dirnhofer S, Koessler P, Ensinger C, Feichtinger H, Madersbacher pitfall in discriminating primary parotid neoplasms from S and Berger P: Production of trophoblastic by metastatic prostate cancer. Diagn Cytopathol 19 : 221-225, transitional cell carcinoma of the bladder: association to tumor 1998. stage and grade. Hum Pathol 29 : 377-382, 1998.

4153 ANTICANCER RESEARCH 28 : 4149-4154 (2008)

29 Martin JE, Jenkins BJ, Zuk RJ, Oliver RT and Baithun SI: 35 Martínez-Rodríguez M, Ramos D, Soriano P, Subramaniam M, Human chorionic gonadotrophin expression and histological Navarro S and Llombart-Bosch A: Poorly differentiated adeno- findings as predictors of response to radiotherapy in carcinoma carcinomas of prostate versus high-grade urothelial carcinoma of of the bladder. Virchows Arch A 414 : 273-277, 1989. the bladder: a diagnostic dilemma with immuno-histochemical 30 Smith DJ, Evans HJ, Newman J and Chapple CR: Ectopic evaluation of 2 cases. Int J Surg Pathol 15 : 213-218, 2007. human chorionic gonadotrophin (HCG) production: is the 36 Sheridan T, Herawi M, Epstein JI and Illei PB: The role of detection by serum analysis of HCG of clinical relevance in P501S and PSA in the diagnosis of metastatic adenocarcinoma transitional cell carcinoma of the bladder? Br J Urol 73 : 409- of the prostate. Am J Surg Pathol 31 : 1351-1355, 2007. 412, 1994. 37 Keillor JS and Aterman K: The response of poorly differentiated 31 Gillott DJ, Iles RK and Chard T: The effects of beta-human prostatic tumors to staining for prostate-specific antigen in chorionic gonadotrophin on the in vitro growth of bladder cancer prostatic acid phosphatase: a comparative study. J Urol 137 : 894- cell lines. Br J Cancer 73 : 323-326, 1996. 896, 1987. 32 Bassily NH, Vallorosi CJ, Akdas G, Montie JE and Rubin MA: 38 Partin AW, Carter HB, Chan DW, Epstein JI, Oesterling JE, Coordinate expression of cytokeratins 7 and 20 in prostate Rock RC, Weber JP and Walsh PC: Prostate-specific antigen in adenocarcinoma and bladder urothelial carcinoma. Am J Clin the staging of localized prostate cancer: influence of tumor Pathol 113 : 383-388, 2000. differentiation, tumor volume and benign hyperplasia. J Urol 33 Torenbeek R, Lagendijk JH and Van Diest PJ: Value of a panel 143 : 747-752, 1990. of antibodies to identify the primary origin of adenocarcinomas presenting as bladder carcinoma. Histopathology 32 : 20-27, 1998. Received March 20, 2008 34 Anderson GA: Poorly differentiated bladder tumor with Revised June 13, 2008 retroperitoneal adenopathy. Infect Urol 14 : 101-104, 2001. Accepted June 16, 2008

4154