Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Prostate Cancer and Prostatic Diseases (2004) 7, 243–248 & 2004 Nature Publishing Group All rights reserved 1365-7852/04 $30.00 www.nature.com/pcan Clusterin as a possible predictor for biochemical recurrence of prostate cancer following radical prostatectomy with intermediate Gleason scores: a preliminary report MR Pins1, JE Fiadjoe1,2, F Korley1, M Wong1, AW Rademaker1, B Jovanovic1, TK Yoo1,3, JM Kozlowski1, A Raji1, XJ Yang1 & C Lee1* 1 Departments of Urology, Preventive Medicine, Pathology, and Robert H. Lurie Comprehensive Cancer Center, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA Disease recurrence following radical prostatectomy is a major concern in prostate cancer patients. Gleason scores are useful in predicting recurrence. Low Gleason scores are usually associated with long disease-free intervals, while high Gleason scores are suggestive of early recurrence. However, prediction of recurrence has been difficult with intermediate Gleason scores. Clusterin is a ubiquitous secretory sulfated glycoprotein. It is also an antiapoptotic mediator in prostate cancer. The objective of the present study is to determine if clusterin can serve as a predictive biomarker for recurrence of prostate cancer with intermediate Gleason scores in patients following radical prostatectomy. Prostatic specimens with Gleason score of 6 (3 þ 3) or 7 (3 þ 4) were obtained from the archival bank. Three groups of specimens were investigated. The first group was from nine patients who developed recurrent disease according to a persistent rise of serum PSA within 3 years following radical prostatectomy. Those in the second group and the third group were from patients who showed no evidence of disease recurrence for at least 5 y (11 patients) and 10 y (eight patients), respectively following the surgery. Histological sections were subjected to immunohistochemical staining using a monoclonal antibody specific for clusterin. The staining intensity was scored as 0, 1, 2, and 3, with 0 being no staining, 1 showing less than 25% positive staining, 2 being 25–50% positive, and 3 showing greater than 75% positive staining. One-way ANOVA with Bonferroni correction was used for statistical analysis. Evaluation of the scores of clusterin staining was carried out according to four specific areas in each specimen. They were (a) benign epithelial cells, (b) malignant epithelial cells (cancer epithelia), (c) stromal cells surrounding benign *Correspondence: C Lee, Department of Urology, Tarry Building, Room 16-733, Northwestern University Feinberg School of Medicine, 303 East Chicago Avenue, Chicago, IL 60611, USA. E-mail: [email protected] An abstract of this report was presented at the NCI sponsored prostate cancer SPORE workshop, Westfield Conference Center, Chantilly, Virginia, July 14–16, 2002. 2 Current address: Department of Anesthesiology, Downstate Medical Center, 450 Clarkson Avenue, P.O. Box 6, Brooklyn, NY 11210, USA 3 Current address: Department of Urology, Eul Ji Hospital, 280-1, Hagye 1 Dong, Nowon Gu, Seoul 139-711, South Korea An abstract of this report was presented at the NCI sponsored prostate cancer SPORE workshop, Westfield Conference Center, Chantilly, Virginia, July 14–16, 2002. Received 20 November 2003; revised 20 January 2004; accepted 10 February 2004 Clusterin and recurrent prostate cancer MR Pins et al 244 cells, and (d) stromal cells surrounding malignant cells (cancer stroma). Staining score in prostatic epithelial cells, benign as well as malignant, showed no significant relationship among the three patient groups. However, when staining scores in stromal cells were compared, there was a significant difference between patients with recurrent disease and those showed no evidence of disease recurrence for at least 10 y. Results of this preliminary study support the important role of clusterin in the stromal component for prostate cancer progression. Clusterin immunostaining may be useful to aid the prediction of chance of disease recurrence in patients with Gleason score 6 or 7 prostate cancer following radical prostatectomy. Further studies with a large number of cases are warranted to verify this preliminary finding. Prostate Cancer and Prostatic Diseases (2004) 7, 243–248. doi:10.1038/sj.pcan.4500722 Keywords: recurrence clusterin immunostaining; stromal cells; Gleason scores; PSA Introduction Radical prostatectomy is the treatment of choice with an intention to provide a cure for patients with organconfined prostate cancer. When the postoperative serum prostate-specific antigen (PSA) is undetectable, it is suggestive of a cure. The resurgence of a persistently detectable serum PSA is indicative of biochemical recurrence of prostate cancer. In at least 30–50% of patients, who enjoy an initial disease free state; recur at various postoperative intervals.1 Although the exact factors controlling recurrence in prostate cancer remains unclear, clinical experience dictates that Gleason scores have been useful in predicting disease recurrence.2–5 Cases with high Gleason scores have a tendency of recurring shortly after radical prostatectomy; while those with low Gleason scores often enjoy a prolonged diseasefree interval. The dilemma lies in those with intermediate Gleason scores, when prediction of disease recurrence has been difficult. A tumor biomarker that can aid the prediction of recurrence would be valuable in the management of this special group of patients. Clusterin, also known as SGP-2, CLI, Apo-J, and TRPM-2, is a ubiquitous secretory sulfated glycoprotein.6–10 It is a multifunctional protein, which has been implicated in diverse processes as cell aggregation,6 inhibition of complement-mediated cytotoxicity,8 and lipid transport.11 Clusterin has been shown to play an antiapoptotic role in prostate cancer cells.12,13 The level of clusterin expression has been correlated with Gleason scores in prostate cancer.14 Clusterin overexpression has resulted in the development of resistance to androgen ablation.15 Recently, the use of antisense oligodeoxynucleotide targeting the clusterin gene enhanced apoptosis induced by either radiation or chemotherapeutic agents further support the importance of clusterin expression in tumor progression.16–18 The objective of this preliminary study is to determine if clusterin can be used as a predictor of disease recurrence in prostate cancer with intermediate Gleason scores following radical prostatectomy. identify patients who had radical prostatectomy and subsequent clinical follow-up data from 1989 to 2001. A series of exclusion criteria was applied. In order to test our hypothesis, we assembled a cohort of patients who had undergone prostatectomy and had regularly (annually or semiannually) documented postoperative serum PSA levels. The cohort was limited to patients who had undetectable serum PSA levels (defined as o0.5 ng/ml) at the time of their first postoperative PSA determination and who had no adjuvant therapy prior to surgery. Next, we searched clinical specimens corresponding to the cohort from the archival bank of the Department of Pathology of Northwestern University Medical School. We selected three groups of patients under three broad categories. First, we limited the specimens to be included into this study only with Gleason scores 6 (3 þ 3) and 7 (3 þ 4). This inclusion criterion defines the scope of the present study, as prediction of disease recurrence with extreme Gleason scores has been well established.3–5 Another inclusion criterion is the reduction of serum PSA levels to an undetectable level, defined as less than 0.5 ng/ml; 3–6 months following surgery in all patients. Patients with a detectable serum PSA level immediately following radical prostatectomy were excluded from this study. Finally, patients showing positive surgical margins in their surgical specimens were also excluded from this study. We restricted a group of patients with recurrence within 3 y of surgery (Group A). Biochemical recurrence was defined as patients who had elevated serum PSA levels above the baseline for at least two consecutive office visits. We then selected two additional groups of patients. One group was those who remained diseasefree with a 5-y follow-up (Group B) and another group was those who were disease-free with a 10-y follow-up (Group C). The assignment of patient groups and the number of study subjects in each group are indicated in the following chart: Patient data Group A consisted of patients who recurred within 3 y after prostatectomy (nine patients) Group B consisted of patients without recurrence after 5 y of followup (11 patients) Group C consisted of patients without recurrence after 10 y of follow-up (eight patients) After receiving Institutional Review Board approval, a search of the hospital tumor registry was performed to There were a total of 28 patients. In many cases, where more than one specimen per patient was obtained, each Materials and methods Prostate Cancer and Prostatic Diseases Clusterin and recurrent prostate cancer MR Pins et al specimen was considered as one determination and was analyzed toward the overall mean and variation for that patient. Immunohistochemistry All specimens were from radical prostatectomy specimens and were of formalin-fixed paraffin-embedded tissues. Sections of 6 mm thick were cut and subjected to immunohistochemical staining for clusterin. Sections were first deparaffinized in xylene and then rehydrated in graded alcohol solutions. Endogenous peroxidase activity was quenched using a commercial peroxidase blocking solution (DAKO, Carpinteria, CA, USA). Antigen retrieval was performed using DAKO Target retrieval solution in a steamer at 901C. Sections were then incubated with a 1:200 dilution with a mouse monoclonal antibody against the clusterin alpha chain (Upstate Biotech, Lake Placid, NY, USA) for 30 min at room temperature followed by incubation with a secondary antibody–polymer–HRP (horseradish peroxidase) conjugate (DAKO’s EnVision þ system) for 30 min at room temperature. Clusterin expression was visualized by incubation with 3,30 -diaminobenzidine (DAB) for 2 min followed by counterstaining with Gill’s hematoxylin. All slides were processed at the same time, thus eliminating the run-to-run variation. Negative control slides were processed in a similar manner, with substitution of the primary antibody with nonimmune mouse IgG at the same concentration for the primary antibody. No color reaction was observed in any of the negative control slides. Evaluation of staining patterns Since the intensity of clusterin staining was different from one cell type to another in each specimen, it is important that comparisons were made according to the specific cell type. In an attempt to elucidate characteristics of clusterin staining in specific areas of each specimen, the following regions were evaluated from each slide: (1) Staining status of benign epithelium (2) Staining status of adenocarcinoma (3) Staining status of stroma associated with benign epithelium (4) Staining status of stroma associated with adenocarcinoma The staining pattern of all specimens was evaluated by two authors (MRP and JEF) independently. The evaluators had no knowledge of follow-up information at the time of the evaluation. Occasional discrepancies were resolved by re-evaluation and consensus. Results were scored as 0, 1 þ , 2 þ , and 3 þ , where 0 representing no staining, 1 þ representing less than 25% of cells stained positive, 2 þ representing 25–75% of cells stained positive, and 3 þ representing greater than 75% of cells stained positive. All comparisons of staining patterns were made at 40 magnification. The resulting scores were then tabulated and the patients sorted into the three groups as described earlier. 245 Statistical analysis The status of positive clusterin staining was compared across the three groups: Group A (recurrence within 3 y), Group B (5 y disease-free), and Group C (10 y diseasefree), using a one-way nested analysis of variance.19,20.The correlation among multiple determinations on the same patient was taken into account when calculating the standard error of the mean for each group. This resulted in an effective sample size between the number of persons (if the correlation was near 1) and the number of determinations (if the correlation was near Table 1 Summary of clusterin immunostaining status in prostate cancer specimens for all patients Patient ID Replicate Benign Benign number epithelia stroma Cancer Cancer Total epithelia stroma Group A 1 1 1 2 3 3 4 5 5 6 6 7 8 8 9 9 1 2 3 1 1 2 1 1 2 1 2 1 1 2 1 2 1 1 2 2 1 3 2 2 1 1 2 2 1 1 1 1 3 2 3 2 2 3 3 3 2 3 3 2 2 3 3 2 2 1 1 1 2 3 3 3 3 2 2 3 3 2 1 3 3 2 2 1 1 1 1 2 2 1 2 3 3 1 2 2 9 6 8 6 6 10 9 10 8 7 9 10 9 7 7 8 Group B 18 18 19 20 21 22 22 23 23 24 24 25 25 26 26 26 26 27 28 1 2 1 1 1 1 2 1 2 1 2 1 2 1 2 3 4 1 1 1 2 0 1 1 3 1 2 1 2 2 1 2 1 3 3 1 1 1 1 3 1 1 2 2 2 2 3 3 3 3 3 1 3 3 1 2 1 3 3 1 2 1 2 2 1 3 0 3 2 3 3 3 2 3 0 2 2 3 2 1 0 1 1 2 3 2 3 2 3 1 2 1 1 0 1 7 11 4 5 4 8 6 7 10 7 11 8 11 6 11 9 6 3 5 Group C 10 11 12 13 13 13 14 14 15 15 16 17 1 1 1 1 2 3 1 2 1 2 1 1 1 1 2 2 2 1 1 1 1 1 1 1 1 2 1 3 3 3 1 1 1 1 0 0 2 0 1 1 2 1 1 0 1 2 3 0 0 1 1 1 2 0 0 1 0 1 1 0 4 4 5 7 9 5 3 3 3 5 5 1 Prostate Cancer and Prostatic Diseases Clusterin and recurrent prostate cancer MR Pins et al 246 Figure 1 Clusterin immunostaining in benign epithelial cells, prostate cancer cells, and in stromal cells. (a–c) Examples of clusterin staining in benign prostatic epithelial cells in prostatectomy specimens: (a) An example of staining score of 1, in which less than 25% of epithelial cells showed positive staining for clusterin. (b) An example of staining score of 2, in which 25–75% of epithelial cells showed positive staining for clusterin. (c) An example of staining score of 3, in which more than 75% of epithelial cells showed positive staining for clusterin. (d–f) Examples of clusterin staining in malignant prostatic epithelial (prostate cancer) cells: (d) An example of staining score of 1, in which less than 25% of cancer cells showed positive staining for clusterin. (e) An example of staining score of 2, in which 25–75% of cancer cells showed positive staining for clusterin. (f) An example of staining score of 3, in which more than 75% of cancer cells showed positive staining for clusterin. (g–i) Examples of clusterin staining in stromal cells of prostate cancer specimens: (g) An example of staining score of 1, in which less than 25% of stromal cells showed positive staining for clusterin. (h) An example of staining score of 2, in which 25–75% of stromal cells showed positive staining for clusterin. (i) An example of staining score of 3, in which more than 75% of stromal cells showed positive staining for clusterin. zero). An overall analysis of variance P-value for group was determined and Po0.05 was used as the criterion for significance for this overall P-value. To control the overall chance of making a Type I error at 5% for the three pairwise comparisons (Group A vs Group B, Group A vs Group C, and Group B vs Group C), these comparisons among groups were made using t-tests with the Bonferroni correction,20 so that a P-value o0.0167 ¼ 0.05/3 (two-tail test) was needed for statistical significance. Statistical analyses were performed using PROC MIXED in SAS (SAS Institute Inc., Cary, NC, USA). Results Typical patterns of immunohistochemical staining for clusterin are shown in Figure 1. It is known that clusterin is expressed in both benign and malignant prostatic epithelial cells.14,21 Clusterin can also be expressed in prostatic stromal cells.22 Therefore, the staining intensity was evaluated separately in following four distinct types of cells: (a) benign epithelium, (b) cancerous epithelium, (c) stroma associated with benign epithelium, and (d) stroma associated with cancerous epithelium. Table 1 shows results of clusterin staining in the above four types of cells for each slide. Prostate Cancer and Prostatic Diseases The staining intensity of clusterin expression was compared across the three groups using one-way analysis of variance,19 taking into account that there were multiple determinations per individual. An overall ANOVA P-value was determined. Table 2 shows the result of statistical analysis comparing each type of cells and the cumulative score. There was no significant difference in staining intensity when only the benign prostatic epithelium was subjected for comparison (P ¼ 0.41). A statistically significant difference was observed, when the staining intensity in the stromal cells adjacent to the benign prostatic epithelium (P ¼ 0.006) the staining intensity of cancerous epithelium (P ¼ 0.03), or the staining intensity of stromal cells adjacent to the cancerous prostatic epithelium was evaluated (P ¼ 0.008). Significant differences were observed between patients who had recurred within 3 y of prostatectomy and those who had no recurrence in 10 y. For cancer stromal cells, there was also a difference for those who had no recurrence in 5 and 10 y. When the staining intensity of the cumulative scores was subjected to analysis, a similar degree of significant difference was noted (Po0.001). In addition to the above statistics, we performed comparisons using different calculation parameters. Table 2 also shows that when the scores of two types of stroma cells (associated with benign and malignant Clusterin and recurrent prostate cancer MR Pins et al 247 Table 2 Statistical analysis of clusterin staining in human prostate cancer specimens Group No. of patients A B C P-value 9 11 8 No. of specimens 16 19 12 Benign epithelia Benign stroma Cancer epithelia Cancer stroma Total stroma Total score 1.4870.16 1.5970.14 1.2770.19 0.41a 2.5470.24 2.0570.22 1.2670.27 0.006b 2.2170.26 1.9870.24 1.1670.29 0.03b 1.7970.23 1.6070.21 0.6570.26 0.008c 4.3270.39 3.62+0.36 1.89+0.44 0.001c 8.0770.56 7.2070.52 4.4270.64 0.001c % Total score45.0 10070 58711 9713 0.001d Each value represents mean and standard error of the mean (mean7s.e.m.). In each designated area, the intensity of the staining was scored from 0 to 3, with 0 being no staining and 3 showing the strongest staining. Group A ¼ patients who recurred within 3 y after prostatectomy (nine patients). Group B ¼ patients without recurrence after 5 y of follow-up (11 patients). Group C ¼ patients without recurrence after 10 y of follow-up (eight patients). a No significant pairwise comparisons. b Significant pairwise comparisons: recurrence (Group A) vs 10-y (Group C). c Significant pairwise comparisons: recurrence (Group A) vs 10-y (Group C), 5-y (Group B) vs 10-y (Group C). d Significant pairwise comparisons: recurrence (Group A) vs 5-y (Group B), recurrence (Group A) vs 10-y (Group C), 5-y (Group B) vs 10-y (Group C). epithelia) were combined, the differences among different groups were also significant (Po0.001). The best significance was obtained when the cumulative total scores were calculated as the percentage of the total scores that were greater than 5, it was highly significant (Po0.001) among all three groups (last column in Table 2). Discussion Findings of this preliminary study have indicated that clusterin staining in stromal cells may be a significant factor in predicting disease recurrence in prostate cancer patients following radical prostatectomy irrespective of whether or not the stromal cells are associated with the cancerous tissues. As prediction of recurrence in prostate cancer with extreme Gleason scores has been well characterized, the present study has focused on cases with intermediate Gleason scores (6 or 7). In earlier studies, clusterin immunostaining has correlated with Gleason scores in prostate cancer specimens.14 Since clusterin is an antiapoptotic mediator,12 prostate cancer cells expressing more clusterin would be endowed with a greater ability for survival. This is consistent with cancer cells with high Gleason scores with a greater potential for survival. However, results of the present study seem to suggest that the intensity of clusterin staining in the cancer tissue is unable to provide a prediction for disease recurrence among cases with Gleason score 6 and 7. The exact mechanism for disease recurrence in prostate cancer patients following a successful radical prostatectomy remains a medical mystery. Radical prostatectomy is intended to cure, if the malignancy is confined to the prostate. Theoretically, if postoperative serum PSA remains undetectable, it is suggestive that the disease has been cured. However, the high frequency of disease recurrence from our clinical experience is inconsistent with the possibility that there might be the presence of occasional extraprostatic carcinoma cells at the time of radical prostatectomy, especially in the face of a negative finding from surgical margins and from pelvic lymph nodes. The release of carcinoma cells during the course of radical prostatectomy can be considered as one of the possibilities for disease recurrence, unless the host environment is hostile to the released cancer cells. The present findings have provided some insights regarding the role of stromal cells in determining the fate of prostate cancer cells. It is inevitable that there are viable prostatic cancer cells released during prostate removal.23 Although the fate of these intraoperative PSApositive cells remains unclear, some of these cells can be found later in the circulation.24,25 It is possible that these released PSA-positive cells attach locally or in a distant site and eventually develop into viable cancer site. Since all circulating PSA-positive cells are subjected to the same host environment, while the majority of released cancer cells are destined to apoptosis, those endowed with the ability to survive or those met with a favorable survival environment will have a growth advantage and eventually proliferate and develop into clinically significant malignancy. We hypothesize that clusterin expression in the stromal cells may reflect the overall ability of the host to produced clusterin into the circulation may be able to protect these circulating cancer cells from death and thereby increase the chance of tumor recurrence after prostatectomy. A recent report of a downregulation of clusterin expression in transformed epithelial cells and upregulation in stromal cells in prostate cancer of intermediate and high Gleason scores supported our observation.22 In the present study, cases showing weak stromal staining for clusterin were associated with a prolonged disease free interval; while those with strong staining for clusterin were associated with early disease recurrence. A strong clusterin staining in stromal cells is related to disease recurrence regardless whether they are associated with cancerous tissues or benign tissues. Since these stromal cells have been removed from the patients at the time of prostatectomy, it is not likely that they have a direct impact on the survival of recurrent cancer cells. It is likely that these stromal cells reflect the intrinsic overall ability of the host to produce clusterin, which provides a protection to released cancer cells. The strong efficacy of i.p. administration of clusterin antisense oligonucleotides in inhibiting tumor progression also supports a systemic role of clusterin in malignancy.16–18 This possibility is plausible and remains to be tested. In summary, results of this preliminary study support the notion that high levels of clusterin produced by Prostate Cancer and Prostatic Diseases Clusterin and recurrent prostate cancer MR Pins et al 248 prostatic stromal cells contribute to the biologically aggressiveness of prostate cancer cells. These results have indicated that clusterin immunohistochemical staining in prostatic stromal cells could be used as a predictor for disease recurrence in patients following radical prostatectomy, for cases with intermediate Gleason score. The finding that the intensity of clusterin staining in prostatic stromal cells can determine the future fate of prostate cancer cells suggests the importance of noncancer cells in the host in disease recurrence. It is important to point out that the cases targeted in the present study were those with intermediate Gleason scores and were otherwise difficult in providing prediction of disease recurrence. Owing to the fact that the present study followed a stringent exclusion criterion, we were only able to obtain a total of 28 cases that satisfied the requirement at Northwestern University. Further studies with a large number of cases are warranted to verify this preliminary finding. 9 10 11 12 13 14 15 Acknowledgements This study was supported in part by NIH Grants CA80963 and CA90386. We thank for the constructive comments of Dr Beatrice Knudson of the Fred Hutchinson Cancer Research Center. 16 17 References 1 Croawford ED, Kiker JD. Radical retropubic prostatectomy. J Urol 1983; 129: 1145–1156. 2 Gleason DF. Classification of prostatic carcinomas. Cancer Chemother Rep 1966; 50: 125–128. 3 Grayhack JT, Assimos DG. Prognostic significance of tumor grade and stage in the patient with carcinoma of the prostate. Prostate 1983; 4: 13–31. 4 Epstein JL, Partin AW, Sauvageot J, Walsh PC. Prediction of progression following radical prostatectomy. A multivariate analysis of 721 men with long-term follow-up. Am J Surg Pathol 1996; 20: 286–292. 5 Oefelein MG, Colangelo L, Grayhack J, McVary KT. Survival after radical retropubic prostatectomy in men with clinically localized high grade carcinoma of the prostate. Cancer 1995; 76: 2523. 6 Fritz IB, Burdzy BK, Setchell B, Blaschuk O. Ram rete testis fluid contains a protein (clusterin) which influences cell–cell interactions in vitro. Biol Reprod 1983; 28: 1173–1188. 7 de Silva HV et al. Apolipoprotein J: structure and tissue distribution. Biochemistry 1990; 29: 5380–5389. 8 Jenne DE, Tschopp J. Molecular structure and functional characterization of a human complement cytolytic inhibitor found in blood and seminal plasma: identity to sulfated Prostate Cancer and Prostatic Diseases 18 19 20 21 22 23 24 25 glycoprotein-2, a constituent of rat testis fluid. Proc Natl Acad Sci USA 1989; 86: 7123–7127. Jenne DE, Tschopp J. Clusterin: the intriguing guises of a widely expressed glycoprotein. Trends Biochem Sci 1992; 17: 154–159. Leger JG, Montpetit ML, Tenniswood MP. Characterization and cloning of androgen-repressed mRNAs from rat ventral prostate. Biochem Biophys Res Commun 1987; 147: 196–203. Sylveste SR, Morales C, Oko R, Griswold MD. Localization of sulfated glycoprotein-2 (clusterin) on spermatozoa and in the reproductive tract of the male rat. Biol Reprod 1991; 31: 1087–1101. Sensibar JA et al. Prevention of cell death induced by tumor necrosis factor alpha in LNCaP cells by overexpression of sulfated glycoprotein-2 (clusterin). Cancer Res 1995; 55: 2431– 2437. Sintich S et al. Cytotoxic sensitivity to tumor necrosis factor-a in prostate cancer cells is regulated by extracellular levels of SGP-2 (clusterin). Prostate 1999; 39: 87–93. Steinberg J et al. Intracellular levels of SGP-2 (Clusterin) correlate with tumor grade in prostate cancer. Clin Cancer Res 1997; 3: 1707–1711. Miyake H, Nelson C, Rennie PS, Gleave ME. Testosteronerepressed prostate message-2 is an antiapoptotic gene involved in progression to androgen-independence in prostate cancer. Cancer Res 2000; 60: 170–176. Miyake H, Hara I, Kamidono S, Gleave ME. Synergistic chemosensitivity and inhibition of tumor growth and metastasis by the antisense oligodeoxynucleotide targeting clusterin gene in a human bladder cancer model. Clin Cancer Res 2001; 7: 4245– 4252. Zellweger T et al. Antitumor activity of antisense clusterin oligonucleotides is improved in vitro and in vivo by incorporation of 20 -O-(2-methoxy)ethyl chemistry. J Pharmacol. Exp Ther 2001; 298: 934–940. Zellweger T et al. Enhanced radiation sensitivity in prostate cancer by inhibition of the cell survival protein clusterin. Clin Cancer Res 2002; 8: 3276–3284. Rosner B. Fundamentals of Biostatistics, 5th edn. Duxbury Press: Pacific Grove, CA 2000. Winer BJ, Brown DR, Michels KM. Statistical Principles in Experimental Design, 3rd edn. McGraw-Hill: New York, NY 1991; pp 158, 358. Sensibar JA et al. Prostatic ductal system in rats: regional variation in localization of an androgen-repressed gene product, sulfated glycoprotein-2. Endocrinology 1991; 128: 2091–2102. Bettuzzi S et al. Clusterin gene expression is down-regulated in transformed epithelial cells but up-regulated in fibroblasts from prostate cancer. J Urol 2003; 169: 159 (Abstract 617). Ignatoff JM et al. Prostate specific antigen reverse transcriptasepolymerase chain reaction assay in preoperative staging of prostate cancer. J Urol 1997; 158: 1870–1874. Oefelein MG et al. Molecular detection of prostate epithelial cells from the surgical field and peripheral circulation during radical prostatectomy. J Urol 1995; 155: 238–242. Oefelein MG et al. Clinical follow-up after radical retropubic prostatectomy. J Urol 1999; 162: 293–306.