The Prostate 31:91–97 (1997) Characterization of Residual Tumor Cells Following Radical Radiation Therapy for Prostatic Adenocarcinoma; Immunohistochemical Expression of Prostate-Specific Antigen, Prostatic Acid Phosphatase, and Cytokeratin 8 Gunilla Ljung,1,5* Mona Norberg,2 Lars Holmberg,3 Christer Busch,4 and Sten Nilsson1 1 Department of Oncology, Akademiska Hospital, University of Uppsala, Uppsala, Sweden 2 Department of Diagnostic Radiology, Akademiska Hospital, University of Uppsala, Uppsala, Sweden 3 Department of Surgery, Akademiska Hospital, University of Uppsala, Uppsala, Sweden 4 Department of Pathology, Akademiska Hospital, University of Uppsala, Uppsala, Sweden 5 Department of Oncology, Mälar Hospital Eskilstuna, Sweden BACKGROUND. Our knowledge about the nature and biological activity of tumor cells residing in the prostate gland after radical radiotherapy (RRT) is limited. METHODS. In the present study, residual tumor in core biopsies taken from 37 patients after an average of 6.8 years follow-up after radiation, were investigated with immunohistochemistry for the biochemical markers prostate-specific antigen (PSA), prostatic acid phosphatase (PAP), and the epithelial marker, cytokeratin 8 (CK8). RESULTS. Tumor cells were cytokeratin-positive in 33 of 34 evaluable specimens (97%). PSA and PAP were expressed in the tumor cells in 94% (34/36) and 81% (30/37) of cases, respectively. No significant correlation was observed between PSA/PAP expression and tumor grade after treatment. Endocrine treatment administered in addition to RRT in 12 of the 37 patients did not affect the expression of PSA or PAP. The expression of both biochemical markers was reduced after radiotherapy in 10 of the 12 cases for which pre- and posttreatment specimens were available. CONCLUSIONS. Tumor cells retain their epithelial characteristics immunohistochemically after radiation, though their morphology sometimes suggests an altered phenotype after treatment. PSA and PAP reactivity was demonstrated in tumor cells nearly 7 years after radiotherapy, which indicates that these cells maintain their biochemical integrity and protein synthesis to a certain extent. Furthermore, endocrine treatment did not abolish PSA or PAP expression in the tumor cells. Whether PSA and PAP immunoexpression provides independent prognostic information needs to be further investigated. Prostate 31:91–97, 1997. © 1997 Wiley-Liss, Inc. KEY WORDS: prostatic adenocarcinoma; radiation therapy; endocrine treatment; core biopsy; PSA; PAP; cytokeratin 8 INTRODUCTION If systematic mapping biopsies are performed, residual local tumor in the prostate gland has been detected in a high percentage of cases after radical © 1997 Wiley-Liss, Inc. *Correspondence to: Dr. Gunilla Ljung, The Mälar Hospital, S-631 88 Eskilstuna, Sweden. Received 6 November 1995; Accepted 1 March 1996 92 Ljung et al. radiation therapy (RRT) [1,2]. Very little is known about those residual tumor cells, but immunohistochemical staining techniques permit the possibility of obtaining further information about these cells. A previous report from our institute demonstrated residual tumor cells in 67% of patients in a long-term follow-up after radical radiation therapy . In 94% and 64% of these cases, respectively, proliferative activity could be demonstrated by the presence of the proliferation markers PCNA and Ki-67 . Morphologically, tumor cells have been observed to lose their epithelial characteristics after radiation and, sometimes come to resemble sarcoma cells . Therefore, a study aimed at investigating whether, and to what degree, residual tumor cells maintained their epithelial characteristics after RRT was constructed; furthermore, it was considered of interest to determine whether, and to what extent, the tumor cells retained prostate-specific protein synthesis after high doses of radiation therapy. Prostate-specific antigen (PSA) is a 33-kDa glycoprotein specific to prostatic duct and acinar epithelial cells [6–8]. It is produced in benign and malignant cells [9–12] and acts as a neutral protease for liquifying the seminal coagulum . Prostatic acid phosphatase (PAP), a glycoprotein of approximately 100 kDa and immunologically different from PSA, is found not only in benign and malignant prostatic epithelial cells but also in other tissues [14–17]. The physiological role of PAP remains obscure . Cytokeratin 8 (CK8), an intermediate filament and epithelial marker, has been identified in ductal and glandular benign and malignant prostatic epithelia [19,20] and constitutes a component of the cytoskeleton in cells, stabilizing the cell structure and junctions between cells [21–23]. The aim of the present study was to investigate some of the functional and structural characteristics of residual tumor cells following RRT. For this purpose, immunohistochemical stainings for the two important prostate markers, PSA and PAP, as well as for the epithelial marker, CK8, were undertaken. MATERIALS AND METHODS Patients and Specimens Fifty-five patients treated with radical external beam irradiation for localized prostatic adenocarcinoma underwent transrectal ultrasonically (TRUS)guided sextant core biopsies to assess the histopathological local cure, after an average of 6.8 years after RRT . In 37 cases (67%), residual tumor was found in one or several biopsies. Tumor stage (UICC 1987)  at initial diagnosis was T1 in 4, T2 in 26, and T3 in 7 of these 37 patients and initial G1, G2, and G3 tumor grades [5,25] were present in 3, 19, and 15 patients, respectively. Endocrine treatment (estrogen treatment in three cases, orchiectomy in nine cases) had been administered in 12 patients, at an average of 38 months (range 8–90) prior to the biopsy investigation. In each of the 37 cases in which residual tumor was found after RRT, one representative core biopsy was examined immunohistochemically for the expression of PSA, PAP, and CK8. To estimate the individual variation of PSA and PAP expression, all biopsies in which residual tumor was found [average three biopsies per patient (range 1–7) and 126 biopsies in total] were investigated. Pretreatment TURP specimens were available in 12 of the 37 cases; an average of three blocks per patient (range 2–4) and 17 chips (range 9–28) per block were examined. Staining Methods Stainings were performed according to the ABC method [26,27]. Specimens were de-paraffinated with xylene and rehydrated with graded ethanols. Endogenous peroxidase activity was blocked with H2O2, and nonspecific protein binding was blocked by bovine serum albumin (BSA). Specimens were then incubated with the primary PSA, PAP, and CK8 antibodies (PSA: A562, 1/200, polyclonal rabbit antisera [Dakopatts, Glostrup, Denmark]; PAP: A627, 1/600, polyclonal rabbit antisera [Dakopatts]; CK8: 6D7, undiluted monoclonal mouse antibody [Immuno Develop Lab. Sweden]; specimens were pretreated with 0.1% protease before incubation with this antibody). In a second step, tissue sections were incubated with the biotinylated second antibody (DAKO, Glostrup, Denmark E353: 1/200, swine antirabbit antisera for PSA/ PAP; DAKO E0354: 1/200, rabbit antimouse antibody for CK8) and, finally, sections were incubated with the ABC-horseradish-peroxidase (HRP) complex in BSA and developed with ethylcarbazole (AEC) in DMSO and H2O2. Slides were counterstained with MayersHtx, rinsed in water, and mounted in glycerin/gelatin. All stainings were performed at room temperature; incubation lasted 30 min. All incubation steps were followed by rinsing in phosphate-buffered saline (PBS). As positive control for PSA and PAP (untreated), TURP specimens with benign prostatic hyperplasia (BPH) and foci of invasive carcinoma were used, while intestinal mucosa was used as positive control for CK8. Appropriate negative controls were included. In addition to the histopathological distinction between invasive carcinoma, PIN and radiationinduced atypia in the post-RRT biopsies, a mouse monoclonal antibody reactive with human cytokeratin of high molecular weight and specific for basal cells was also used (DAKO, 34BE12)  (Fig. 1a,b). PSA, PAP, CK8 in Residual Tumor Cells After Radiation Therapy 93 Fig. 1. a: Immunostaining for high density cytokeratin, labeling basal cells present in nonmalignant epithelial tissue. Most cells in the biopsy lack a basal cell layer and represent adenocarcinoma. ×100. b: Same case as in Fig. a. ×250. Grading and Scoring of Stainings Statistics The extent of the stainings was registered on a semiquantitative scale of 0–3, using the following criteria: 0, no detectable staining; 1, <25% stained cells; 2, $25%<50% stained cells; and 3, staining in >50% of cells. The intensity of the stainings was registered according to the following criteria: 0, no detectable staining; 1, weakly stained cells; 2, moderately stained cells; and 3, strongly stained cells. A system for evaluation of the immunostaining pattern with the multiple of extent × intensity to give scores of 0–9 was used ; the possible scores obtained (0, 1, 2, 3, 4, 6, 9) were further denoted negative (score 0), low (scores 1 and 2), intermediate (scores 3 and 4), and high (scores 6 and 9). Each sample was quantitated on two separate occasions by one investigator (G.L.) in a blinded manner, in an effort to test the reproducibility of the method and minimize bias. Fisher’s exact test and Pearson correlation coefficients were used. P-values of #0.05 were considered statistically significant. RESULTS In most cases, or 97% (33 of 34 evaluable specimens), positive immunostaining with CK8 was observed in the tumor cells. PSA and PAP were expressed in tumor cells in 94% (34/36) and 81% (30/37) of evaluable cases, respectively. The distribution of staining scores is illustrated in Table I; low, intermediate, and high PSA scores were observed in 29%, 21%, and 50% of the positive stainings, respectively. The corresponding PAP scores were found in 50%, 23% and 27% of positive cases, respectively. The intraobserver reproducibility of the method was 81% 94 Ljung et al. Fig. 2. PSA immunostaining of poorly differentiated adenocarcinoma. Almost general positivity, but with heterogeneous intensity. ×250. Fig. 3. PAP immunostaining in the same case as Fig. 2. Incomplete weak positivity and heterogeneous intensity. ×250. Fig. 4. High magnification (×400) of PSA immunostaining of poorly differentiated adenocarcinoma. (29/36) for the PSA stainings and 86% (32/37) for the PAP stainings. The PSA and PAP expression was heterogeneous between individuals, a phenomenon that was most marked in G3 tumors. Heterogeneity was also observed in the individual patient; between tumor cells in the individual biopsy (Figs. 2, 3), as well as between different parts of the gland where tumor was found at biopsy. A difference of >2 scores between tumor lesions from different biopsy locations in the individual case, was observed in 51% (19/37) of cases as regards to the PSA stainings and 43% (16/37) of cases as regards to the PAP stainings. A correlation between the tissue expression of PSA and PAP was observed, although tumor cells generally expressed PSA more strongly than PAP (r= 0.6; P=0.0002) (Figs. 2, 3, 4; Table 1). There was no apparent correlation between the expression of the biochemical markers and the tumor grade after treatment (Fisher’s exact two-tailed test, P $0.5). Endocrine treatment in addition to RRT had PSA, PAP, CK8 in Residual Tumor Cells After Radiation Therapy TABLE I. PSA and PAP Staining Results and Scores Score N (%) Low Intermediate High PSA Neg. 2/36 (6%) Pos. 34/36 (94%) 10/34 (29%) 7/34 (21%) 17/34 (50%) PAP Neg. 7/37 (19%) Pos. 30/37 (81%) 15/30 (50%) 7/30 (23%) 8/30 (27%) Low, immunohistochemical scores 1 and 2; intermediate, scores 3 and 4; high, scores 6 and 9. been administered in 12 of the 37 patients and did not affect PSA or PAP immunoexpression (Fisher’s exact two-tailed test, P $0.5). In 12 patients, comparisons of PSA and PAP expression in pretreatment TURP specimens and post-treatment core biopsies were possible. PSA and PAP were expressed in all pretreatment specimens, and a reduction of >2 scores of PSA as well as PAP was demonstrated in 83% (10/12) of the cases after RRT. PAP expression was enhanced in 1 of 12 patients while, in the remaining cases, both biochemical markers remained unchanged after treatment. DISCUSSION CK8 was expressed in the vast majority of cases (97%), indicating that tumor cells present after RRT retain their epithelial characteristics. PSA and PAP immunostainings were positive in 94% and 81% of cases, respectively, which confirms that the cells were not ‘‘sterilized’’ after radiation but capable of protein synthesis and biochemical activity. This is in agreement with the few studies previously performed [11,29,30]. While none of the data presented here was new, further information was gained, in that the number of patients, the sampling technique, and follow-up exceeded the data from previous reports. In the 12 cases of the present series, where preand post-RRT immunoreactivity for PSA and PAP were available, a reduced expression of both biochemical markers was observed in 83% of cases after treatment. This may reflect radiation injury to the cells, but also a dedifferentiation of the cells over time (observed in 51% of cases ), with a loss or impairment of specialized cell functions. Furthermore, these results must be interpreted with caution because the pretreatment TURP specimens and posttreatment core biopsies reflect different parts of the prostate gland and may not be readily compared due 95 to the notorious heterogeneity of prostatic adenocarcinoma. A correlation between tissue PSA and PAP expression in untreated prostatic adenocarcinoma has been reported [31–33]; the present study demonstrates the same results in the post-treatment situation. In the present study, PSA expression was generally stronger than the expression of PAP. However, conflicting data concerning whether PSA or PAP is the more sensitive method when labeling prostatic adenocarcinoma cells has been reported [34–36] and, as a consequence, the use of both markers has been recommended . PSA and PAP tissue expression was heterogenous in the irradiated tumor cells, both in the individual tumor and between patients. This finding is in accordance with the observations of Brawer et al.  in the post-treatment situation, as well as in untreated prostatic tumors, of which the majority have been reported to exhibit a heterogeneous PSA and PAP immunoreactivity [7,12,15,16,29,30,32,34,35,38–41]. We could see no apparent correlation between PSA and PAP tissue expression and tumor grade after treatment, as opposed to the results reported by several investigators who demonstrate an inverse correlation between PSA/PAP immunoexpression and tumor grade in the pretreatment situation [31–33,35, 38,41–44]. Data regarding PSA and PAP expression viz. tumor grade post-RRT must, however, be interpreted with caution, due to the difficulties involved in tumor grading after radiation . In the present study, the expression of PSA and PAP in tumor cells after RRT did not correlate to disease progression in terms of distant metastases at biopsy control (Fisher’s extact two-tailed test, P >0.2, data not shown). However, due to the unknown regional lymph node status at initial diagnosis, conclusive data viz. progression of disease cannot be obtained from the present series. An adverse prognosis for tumors with poor PSA and/or PAP immunoreactivity has been reported [33,38,44–48], but whether these markers provide independent prognostic information requires further investigation. The expression of PSA/PAP in tumor cells was not abolished or decreased when assessed three years after androgen withdrawal or suppression. Contradictory results have been reported [16,43], but our results are in line with other reports [29,49] (Magnus Hellstrom, Christer Busch, personal communication) and may suggest altered regulatory mechanisms for PSA/PAP production in tumor cells after androgen depletion. Furthermore, the addition of endocrine manipulation to radiation therapy may not improve long-term local cure in prostatic adenocarcinoma. Owing to the low number of cases, the statistical 96 Ljung et al. power of this study is limited. However, the number of cases presented exceed the numbers of patients reported in previous studies on this subject [11,29, 30]. Despite the fact that the interpretation of immunohistochemical stainings is subjective, the staining results obtained in the present study were highly reproducible. Whether immunohistochemistry adequately reflects the PSA and PAP tissue expression is unknown, but a close correlation of PSA content assessed by immunohistochemistry and mRNA in situ hybridization [31,41] and Northern blot techniques  has been demonstrated. 10. CONCLUSIONS 11. Residual tumor cells retain their epithelial characteristics immunohistochemically after high doses of radiation. The biochemical markers PSA and PAP were expressed in the tumor cells in the majority of cases, which demonstrates a biological and functional capacity in residual tumor cells several years after radiation treatment. These results imply that radiation therapy for prostatic adenocarcinoma must be substantially altered and improved if local cure is to be achieved. ACKNOWLEDGMENTS This study was supported by grants from the Lions Cancer Research Fund Uppsala, the Swedish Cancer Research Fund, and the Research Fund of Sörmlands Country Council, Sweden. Our thanks to Frank Bittkowski for pleasant and skillful cooperation at the photomicroscope. REFERENCES 1. Kabalin JN, Hodge KK, McNeal JE, Stamey TA: Identification of residual cancer in the prostate following radiation therapy; roles of transrectal ultrasound guided biopsy and prostate specific antigen. J Urol 142: 326–331, 1989. 2. Herr HW, Whitmore WF: Significance of prostate biopsies after radiation therapy for carcinoma of the prostate. Prostate 3:339–350, 1982. 3. Ljung G, Norberg M, de la Torre M, Egevad L, Holmberg L, Busch C, Nilsson S: Transrectal ultrasonicallyguided core biopsies in the assessment of local cure of prostatic cancer after radical external beam radiotherapy. Acta Oncol 34:945–952, 1995. 4. Ljung G, Egevad L, Norberg M, de la Torre M, Holmberg L, Busch C, Nilsson S: Assessment of proliferation indicators in residual prostatic adenocarcinoma cells after radical external beam radiotherapy. Prostate 29: 303–310, 1996. 5. Mostofi FK, Sesterhehn I, Sobin LH: ‘‘Histological Typing of Prostatic Tumours. International Histological Classifications of Tumours.’’ No. 22. Geneva: World Health Organization, 1980. 6. Wang MC, Valenzuela LA, Murphy GP, Chu TM: Pu- 7. 8. 9. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. rification of a human prostate specific antigen. Invest Urol 17:159–163, 1979. Brawer MK: Prostate specific antigen. A review. Acta Oncol 30:161–168, 1991. Oesterling JE, Chan DW, Epstein JI, Kimball AW Jr., Bruzek DJ, Rock RC, Brendler CB, Walsh PC: Prostate specific antigen in the preoperative and postoperative evaluation of localized prostate cancer treated with radical prostatectomy. J Urol 139:766–772, 1988. Stamey TA, Vang N, Hay AR, McNeal JE, Freiha FS, Redwine E: Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med 317: 909–916, 1987. Hudson MA, Bahnson RB, Catalona WJ: Clinical use of prostate specific antigen in patients with prostate cancer. J Urol 142:1011–1017, 1989. Brawer MK, Nagle RB, Pitts W, Freiha FS, Gamble SL: Keratin immunoreactivity as an aid to the diagnosis of persistent adenocarcinoma in irradiated human prostates. Cancer 63:454–460, 1989. Nadji M, Talei SZ, Castro A, Chu TM, Murphy GP, Wang MC, Morales AR: Prostatic-specific antigen: An immunohistologic marker for prostatic neoplasms. Cancer 48:1221–1232, 1981. Lilja H: A kallekrein-like serine protease in prostatic fluid cleaves the predominant seminal vesicle protein. J Clin Invest 76:1899–1903, 1985. Abalin RJ: Immunochemical identification of prostatic tissue-specific acid phosphatase. Clin Chem 19:786– 788, 1973. Jöbsis AC, De Vries GP, Anholt RRH: Demonstration of the prostatic origin of metastases: An immunohistochemical method for formalin-fixed embedded tissue. Cancer 41:1788–1793, 1978. Jöbsis AC, De Vries GP, Meijer AEFH, Ploem JS: The immunohistochemical detection of prostatic acid phosphatase: Its possibilities and limitations in tumour histochemistry. Histochem J 13:961–973, 1981. Nadji M, Tabei SZ, Castro A, Chu TM, Morales AR: Prostatic origin of tumors. An immunohistochemical study. Am J Clin Pathol 73:735–739, 1980. Vihko P, Kontturi M, Korhonen K: Purification of human prostatic acid phosphatase by affinity chromatography and isoelectric focusing. Part I. Clin Chem 24: 466–470, 1978. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R: The catalog of human cytokeratins: Patterns of expression in normal epithelia, tumours and cultured cells. Cell 31:11–24, 1982. Quinlan RA, Schiller DL, Hatzfeld M, Achstätter T, Moll R, Jorcano JI, Magin TM, Franke WW: Patterns of expression and organization of cytokeratin intermediate filaments. In Wang E, Fishman D, Liem RRH, Sun TT (eds): ‘‘Intermediate Filaments.’’ Ann NY Acad Sci 455:282–306, 1985. Cooper D, Schermer A, Sun TT: Classification of human epithelia and their neoplasms using monoclonal antibodies to keratins: Strategies, applications and limitations. Lab Invest 52:243–256, 1985. Ramackers FCS, Puts JJG, Moesker O, Kant A, Huysmans A, Haag D, Jap PHK, Herman CJ, Vooijs GP: Antibodies to intermediate filament proteins in the immunohistochemical identification of human tumours: an overview. Histochem J 15:691–713, 1983. Silen Å: ‘‘Monoclonal Anti-cytokeratin Antibodies and PSA, PAP, CK8 in Residual Tumor Cells After Radiation Therapy 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Cytokeratin Fragments as in vitro and in vivo Tumour Markers.’’ Comprehensive Summaries of Uppsala Dissertions from the Faculty of Medicine 506. Acta Universitatis Upsaliensis, Uppsala, 1994. TNM Classification of Malignant Tumours: UICC, 3rd Ed. Berlin: Springer-Verlag, 1987. Mostofi FK: Grading of prostate carcinoma. Cancer Chemother Rep 59:111–120, 1975. Hsu S-M, Raine L, Fanger H: A comparative study of the peroxidase-antiperoxidase method and an avidinbiotin complex method for studying polypeptide hormones with radioimmunoassay antibodies. Am J Clin Pathol 75:734–738, 1981. Bourne JA: ‘‘Handbook of Immunoperoxidase Staining Methods.’’ DAKO Corp, Santa Barbara, CA, 1983. Busch C, Malmström PU, Norlen BJ, Sällström J, Brodin T, Lundblad A: A grading system for the immunostaining of A, B and H blood group isoantigens in bladder carcinoma. Anticancer Res 8:81–88, 1988. Vernon SE, Williams WD: Pre-treatment and post-treatment evaluation of prostatic adenocarcinoma for prostatic specific phosphatase and prostatic specific antigen by immunohistochemistry. J Urol 130:95–98, 1983. Mahan DE, Bruce AW, Manley PN, Franchi L: Immunohistochemical evaluation of prostatic carcinoma before and after radiotherapy. J Urol 124:488–491, 1980. Hakalahti L, Vihko P, Henttu P, Autio-Harmainen H, Soini Y, Vihko R: Evaluation of PAP and PSA gene expression in prostatic hyperplasia and prostatic carcinoma using Northern-blot analyses, in situ hybridization and immunohistochemical stainings with monoclonal and bispecific antibodies. Int J Cancer 55:590– 597, 1993. Garde SV, Sheth AR, Venkatestan VM, Panchal CJ, Porter AT, Grignon DJ: Prostate inhibin peptide (PIP) in prostate cancer: A comparative immunohistochemical study with prostate-specific antigen (PSA) and prostatic acid phosphatase (PAP). Cancer Lett 78:11–17, 1994. Stege R, Lundh B, Tribukait B, Pousette Å, Carlström K, Hasenson M: Deoxyribonucleic acid ploidy and the direct assay of prostatic acid phosphatase and prostate specific antigen in fine needle aspiration biopsies as diagnostic methods in prostatic carcinoma. J Urol 144: 299–302, 1990. Vitali A, Ardoino S, Durante P, Ferro MA, Li Causi F, Parodi C, Sanguinetti G, Gaffuri M, Paerachino M, Salvadori RP: Correlation between immunohistochemical patterns and serum levels of PSA and PSAP in prostatic pathology: Evaluation of 198 prostatic fine needle biopsies. Anticancer Res 14:1503–1508, 1994. Svanholm H: Evaluation of commercial immunoperoxidase kits for prostatic specific antigen and prostatic specific acid phosphatase. Acta Pathol Microbiol Immunol Scand A 94:7–12, 1986. Oesterling JE: Prostate specific antigen: A critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J Urol 145:907–923, 1991. Epstein J, I. PSA and PAP as immunohistochemical markers in prostate cancer. Urol Clin North Am 55:757– 770, 1993. 97 38. Bazinet M, Hamdy SM, Bégin LR, Stephenson RA, Fair WR: Prognostic significance of antigenic heterogeneity, Gleason grade and ploidy of lymph node metastases in patients with prostate cancer. Prostate 20:311–326, 1992. 39. Pretlow TG, Yang B, Kaetzel CS, Delmoro CM, Kamis SM, Bodner DR, Kursh E, Resnick MI, Bradley EL Jr.: Tissue concentrations of prostate-specific antigen in prostatic carcinoma and benign prostatic hyperplasia. Int J Cancer 49:645–649, 1991. 40. Abrahamsson PA, Falkmer LS, Wadström LB: Immunohistochemical distribution of the three predominant secretory proteins in the parenchyma of hyperplastic and neoplastic prostatic glands. Prostate 12:39–46, 1988. 41. Qiu SD, Young CYF, Bilhartz DL, Prescott JL, Farrow GM, He WW, Tindall DJ: In situ hybridization of prostate specific antigen m RNA in human prostate. J Urol 144:1550–1556, 1990. 42. Aihara M, Lebovitz RM, Wheeler TM, Kinner BM, Ohori M, Scardino P: Prostate specific antigen and Gleason grade: An immunohistochemical study of prostate cancer. J Urol 151:1558–1564, 1994. 43. Berner A, Nesland JM, Waehre H, Fosså SJ: Hormone resistant prostatic adenocarcinoma. An evaluation of prognostic factors in pre- and post-treatment specimens. Br J Cancer 68:380–384, 1993. 44. Sakai H, Shiraishi K, Minami Y, Yushita Y, Kanetake H, Saito Y: Immunohistochemical prostatic acid phosphatase level as a prognostic parameter in prostatic carcinoma. Prostate 19:265–272, 1991. 45. Epstein JI, Egglestone JC: Immunohistochemical localization of prostate-specific acid phosphatase and prostate-specific antigen in stage A2 adenocarcinoma of the prostate: Prognostic implications. Hum Pathol 15:853– 859, 1984. 46. Berner A, Waere H, Nesland JM, Paus E, Danielsen HE, Fosså SD: DNA ploidy, serum prostate specific antigen, histological grade and immunohistochemistry as predictive parameters of lymph node metastases in T1T3/M0 prostatic adenocarcinoma. Br J Urol 75:26–32, 1995. 47. Stege R, Tribukait B, Lundh B, Carlström K, Pousette Å, Hasenson M: Quantitative estimation of tissue prostate specific antigen, deoxyribonucleic acid ploidy and cytological grade in fine needle aspiration biopsies for prognosis of hormonally treated prostatic carcinoma. J Urol 148:833–837, 1992. 48. Hammond ME, Sause WT, Martz KL, Pilepich NV, Asbell SO, Rubin P, Myers RP, Farrow GM: Correlation of prostate-specific acid phosphatase and prostate-specific antigen immunocytochemistry with survival in prostate carcinoma. Cancer 63:461–466, 1989. 49. de Winter JAR, Janssen PJ, Sleddens HM, VerleunMooijman MC, Trapman J, Brinkman AO, Santerse AB, Schröder FH, van der Kwast TH: Androgen receptor status in localized and locally progressive hormone refractory human prostate cancer. Am J Pathol 144:735– 746, 1994.