Combined androgen blockade with nonsteroidal antiandrogens for advanced prostate cancer: a systematic review2

ADULT UROLOGY

COMBINED ANDROGEN BLOCKADE WITH NONSTEROIDAL ANTIANDROGENS FOR ADVANCED PROSTATE CANCER: A SYSTEMATIC REVIEW BRIAN SCHMITT, TIMOTHY J. WILT, PAUL F. SCHELLHAMMER, VINCENZO DEMASI, OLIVER SARTOR, E. DAVID CRAWFORD, AND CHARLES L. BENNETT

ABSTRACT Objectives. Combined androgen blockade with medical or surgical castration plus a nonsteroidal antiandrogen for metastatic prostate cancer has been the subject of 20 randomized trials. The findings range from no expected increase in survival in 17 studies to an estimated 3.7 to 7 months’ survival improvement noted in 3 studies. Most recently, a 1999 evidence report from the Agency for Healthcare Research and Quality and a 2000 overview from the Prostate Cancer Trialists Collaborative Group indicated that combined androgen blockade was associated with an approximately 3% to 5% increase in 5-year survival. We report herein a systematic review on combined androgen blockade performed by the Cochrane Collaborative Review Group on Prostate Diseases. Methods. Controlled trials that included a randomization of immediate nonsteroidal antiandrogens with castration versus castration alone for metastatic prostate cancer and provided information on survival were reviewed. Information on overall survival, toxicity, progression-free survival, cancer-specific survival, and type of nonsteroidal antiandrogen and castration therapies was abstracted by two independent reviewers. Results. Twenty trials (n ⫽ 6320 patients) were included. The pooled odds ratio (OR) for overall survival with combined androgen blockade was 1.03 (95% confidence interval [CI] 0.85 to 1.25; n ⫽ 4970 from 13 trials), 1.16 (95% CI 1.00 to 1.33; n ⫽ 5286 from 14 trials), and 1.29 (95% CI 1.11 to 1.50; n ⫽ 3550 from 7 trials) at 1, 2, and 5 years, respectively. Progression-free survival was improved at 1 year (OR ⫽ 1.38; 95% CI 1.15 to 1.67; n ⫽ 2278 from 7 trials). Cancer-specific survival was improved at 5 years (OR ⫽ 1.58; 95% CI 1.05 to 2.37; n ⫽ 781 from 2 trials). When analysis was limited to studies identified as being of high quality, the pooled OR for overall survival progressively increased but was not significant at any follow-up interval. Conclusions. We find that there is a 5% improvement in the percentage of men surviving at 5 years (30% vs. 25%) with combined androgen blockade with nonsteroidal antiandrogens as well as improvements in progression-free survival at 1 year. Appropriate patients with metastatic prostate cancer should be informed of the potential benefits, toxicities, and out-of-pocket expenditures. UROLOGY 57: 727–732, 2001. © 2001, Elsevier Science Inc. Supported by American Cancer Society and the Department of Veterans Affairs, Health Services Research and Development Program C.L. Bennett is on the speaker’s bureau for Schering-Plough and ALZA, has received unrestricted grant support from Schering-Plough, AstraZeneca, and Cell Pathways, and was a technical advisor and contributor to the report on hormonal therapies for advanced prostate cancer issued by the Agency for Healthcare Research and Quality and prepared by the Blue Cross and Blue Shield Association’s Technology Evaluation Center, one of the Agency’s Evidence-Based Practice Centers. P.F. Schellhammer is on the speaker’s bureau for Schering-Plough and TAP Pharmaceuticals. E.D. Crawford is on the speaker’s bureau for ScheringPlough and TAP Pharmaceuticals. O. Sartor is on the speaker’s bureau for Schering-Plough. T. Wilt was a technical advisor and contributor to the Evidence-Based Practice Center report on hormonal therapies for advanced prostate cancer. B. Schmitt is a co-author, with the Evidence-Based Practice Center of the Blue Cross and Blue Shield Association’s Technology Evaluation Center and others, of a draft manuscript on hormonal therapies for advanced prostate cancer. A preliminary version of this report has been published in electronic format as a Cochrane Systematic Review, which is available from the Cochrane Collaboration at www.cochranelibrary.com.© 2001, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED

From the Veterans Administration Chicago Healthcare System/ Lakeside Division; the Robert H. Lurie Comprehensive Cancer Center, the Division of Hematology/Oncology, the Department of Medicine, and the Institute of Health Services Research and Policy Studies of Northwestern University; the Division of General Internal Medicine, the Department of Medicine, Northwestern University, Chicago, Illinois; the Department of Urology, Eastern Virginia Medical School, Norfolk, Virginia; Atlantic Cancer Center, Salzbury, Maryland; the Stanley Scott Cancer Center and the Division of Hematology/Oncology, Louisiana State University, New Orleans, Louisiana; the Division of Urology, the University of Colorado, Denver, Colorado; and the Minneapolis Veterans Administration Center for Chronic Disease Outcomes Research, and the Veterans Administration Coordinating Center for the Cochrane Collaborative Review Group in Prostate Diseases and Urologic Malignancies, Minneapolis, Minnesota Reprint requests: Charles L. Bennett, M.D., Ph.D., VA Chicago Healthcare System, Lakeside, Northwestern University Medical School, Medical Science Building, Suite 205, 400 East Ontario Street, Chicago, IL 60611 Submitted: August 2, 2000, accepted (with revisions): November 20, 2000 0090-4295/01/$20.00 PII S0090-4295(00)01086-4 727

P

rostate cancer is the second leading cause of cancer death in men.1 The primary hormonal treatment for advanced disease has been castration.2 It has been hypothesized that counteracting adrenal androgens can improve survival, which has been evaluated in 20 studies of combined androgen blockade therapy with nonsteroidal antiandrogens.3–21 The Southwest Oncology Group (SWOG) trial (NCI 0105) of combined androgen blockade with orchiectomy plus flutamide versus orchiectomy failed to detect a statistically significant increase in survival with combined androgen blockade,11,22 whereas an earlier SWOG study (NCI 0036) of patients randomized to combined androgen blockade with flutamide plus daily leuprolide injections versus daily leuprolide injections found a 24% increase in survival.7 Smaller trials and several overviews2,23–25 have found conflicting results. A 1995 meta-analysis did not find a statistically significant increase in survival with combined androgen blockade,24 whereas recent overviews found a 3% to 5% increase in 5-year survival with combined androgen blockade when nonsteroidal antiandrogens were used.2,23,25 An evidence report on this topic was recently issued by the Agency for Healthcare Research and Quality (AHRQ), which was prepared by the Blue Cross and Blue Shield Association’s Technology Evaluation Center, one of the AHRQ’s Evidence-Based Practice Centers. That report questioned the clinical significance of this magnitude of difference, in light of the limited published data at 5 years and the adverse effect profile associated with combined androgen blockade.2 A subsequent overview from the Prostate Cancer Trialists Collaborative Group (PCTCG) stated that the “collaborative group makes no comment on whether, if real, a difference of a few percent in 5-year survival should be considered clinically significant.”25 Comment on these findings ranged from supporting a shared decisionmaking approach26 to suggesting that there is no role for combined androgen blockade.27 In this study, our objectives were (a) to disseminate the available information on combined androgen blockade through the Cochrane Collaboration, an international effort that conducts, disseminates, and maintains systematic reviews23 and (b) to present our interpretation of combined androgen blockade in the context of the patient– physician relationship. MATERIAL AND METHODS This meta-analysis was designed to evaluate the relative efficacy of combined androgen blockade on overall survival, 728

using any nonsteroidal antiandrogen compared to castration alone for advanced prostate cancer. Secondary objectives were to evaluate the relative efficacy of combined androgen blockade on progression-free survival and/or cancer-specific survival using any nonsteroidal antiandrogen compared to castration alone for advanced prostate cancer and to determine the incidence and type of adverse effects and quality of life from combined androgen blockade using any nonsteroidal antiandrogen when compared to castration alone. For efficacy, randomized controlled trials of nonsteroidal antiandrogen therapy in addition to castration compared to castration alone were reviewed. For adverse events, randomized and nonrandomized studies were included. Trials were included if they enrolled men with advanced prostate cancer who were not previously treated with hormonal therapy. The following groups were included: (a) disseminated and/or symptomatic metastases and (b) asymptomatic metastatic disease or advanced, nonmetastatic, disease. Sensitivity analyses were restricted to studies that provided separate data for stage D2 (M1) patients. The data set produced by the Blue Cross and Blue Shield Association’s Technology Evaluation Center,2 the Cochrane Controlled Trials Register, the CENTRAL register, and the Veterans Administration Cochrane Prostate Disease and Urologic Malignancy register were searched for trials focused on nonsteroidal antiandrogens for prostate cancer. Titles and abstracts were screened by two reviewers and evaluated against inclusion and exclusion criteria, independently abstracting data on trial identifiers; study methods (including enrollment and withdrawal numbers); patient characteristics; and overallsurvival, progression-free survival, and cancer-specific survival. When not reported, survival was estimated from published curves. Disagreements were resolved by consensus. Methodologic quality parameters included success in concealing randomization, successful randomization or equivalent baseline susceptibility to the outcomes of interest, completeness of follow-up, equivalent performance of outcome assessment, and blinded assessment of outcomes. Quality differences were assessed by comparing how each study performed against each criterion. The effect of interstudy quality differences on outcomes was determined by sensitivity analysis. Quantitative analyses of outcomes were based on intention to treat. Approximate chi-square tests for heterogeneity were used to assess outcome data for comparability with the assumption of a uniform risk ratio (P ⬎0.10). Pooled odds ratios (ORs) for survival were derived from (a) directly abstracted information from individual studies providing the “number of surviving patients” at each time period or (b) the “percent survival” listed in the report or calculated from the hazard rate. Second, the “number of patients at risk” and the “number reaching each endpoint” at 1, 2, or 5 years were obtained, based on the provided “number at risk at baseline” and the “percent survival” at the corresponding time interval as calculated from the individual study hazard rates. The number at risk at follow-up was assumed to be equal to the number at risk at baseline, providing an underestimate of survival in trials with relatively few patients. It is unlikely to affect differences in survival between treatment arms. Sensitivity analyses were conducted. ORs for these studies did not vary by more than 10%. Additionally, the effect size and statistical significance of our results were compared to those in the evidence report prepared by the Blue Cross and Blue Shield Association’s Technology Evaluation Center, an Evidence-Based Practice Center of the AHRQ.2 Data were analyzed with the random-effects model. If sensitivity analyses UROLOGY 57 (4), 2001

indicated that smaller or less well-designed studies changed the results, the studies were analyzed separately. We identified 27 randomized controlled trials that compared survival with monotherapy to the outcomes of combined androgen blockade. Seven trials were excluded because a steroidal antiandrogen, cyproterone, was used. One trial that compared four regimens for combined androgen blockade was excluded because there was no comparison to monotherapy.28 Twenty trials with 6320 patients were included in this review. Most patients (n ⫽ 6095) had distant metastases. Eight studies (n ⫽ 3271) included only patients with distant metastases.3,7–9,11,12,17 Two studies had no information about the percentage with M1 disease (n ⫽ 225).18,19 In the remaining trials most patients (mean weighted percentage 78%) had M1 disease (n ⫽ 2824). Twelve trials (n ⫽ 4672) used flutamide, eight (n ⫽ 1648) used nilutamide, and none used bicalutamide. Of the 6320 enrolled patients, 5523 were randomized to treatment or control, and 5432 (98.4%) were analyzable. Nineteen trials were two-arm trials. One of these trials was a two-by-two factorial design.19 For this review, the two monotherapy arms were pooled together as were the two combined androgen blockade arms. One three-arm study compared monotherapy to two combined androgen blockade arms, each with a different dose of nilutamide.6 Among the 20 eligible trials, the monotherapy was orchiectomy in 9 studies (n ⫽ 2671), a luteinizing hormone-releasing hormone (LhRH) agonist in 10 studies (n ⫽ 2959), and orchiectomy or a LhRH agonist in 1 study (n ⫽ 690). All trials with flutamide used 250 mg three times daily. All but one of the trials with nilutamide used 300 mg daily. In one study and in one of the three arms in a second trial a dose of 150 mg daily was used.6,10 Data were available to calculate overall survival at 1, 2, and 5 years in 13 studies (n ⫽ 4970 patients), 14 studies (n ⫽ 5286 patients), and 7 studies (n ⫽ 35 patients) respectively; progression-free survival at 1, 2, and 5 years in 7 studies (n ⫽ 2278 patients), 5 studies (n ⫽ 2141 patients), and 2 studies (n ⫽ 1017 patients), respectively; and cancer-specific survival at 1, 2, and 5 years in 5 studies (n ⫽ 1270 patients), 4 studies (n ⫽ 1232 patients), and 2 studies (n ⫽ 781 patients), respectively. Eight studies specified the method of randomization.3–5,7,9,10,14,16 Allocation was adequately concealed in six studies.3–5,7,9,10 Higher quality studies were defined as those in which the patients and investigators were blinded to group assignment and used intent-to-treat analysis of outcomes.3,6,7,10,13,17,19,21

RESULTS Three studies reported a statistically significant increase in survival that favored combined androgen blockade with a 5-year survival advantage, ranging from 3% to 9%.7,9,10 The remaining 17 trials reported no significant difference. The proportion of patients surviving with combined androgen blockade compared to control was 84.6% versus 84.0% at 1 year, 58.6% versus 55.6% at 2 years, and 30.1% versus 24.9% at 5 years. The pooled estimate of the OR for overall survival progressively increased with time (OR ⫽ 1.03 [95% CI 0.85 to 1.25] at 1 year; OR ⫽ 1.16 [95% CI 1.00 to 1.33] at 2 years; and OR ⫽ 1.29 [95% CI 1.11 to 1.50] at 5 years) (Fig 1A-C). When only studies with more than 90% M1 disease were included, the point estimate of the OR for overall survival was significant only at 5 years. The OR was 1.08 (95% CI 0.86 to UROLOGY 57 (4), 2001

1.38) at 1 year, 1.12 (95% CI 0.94 to 1.32) at 2 years, and 1.25 (95% CI 1.05 to 1.48) at 5 years. When analysis was limited to studies identified as being of high quality, the pooled OR for overall survival progressively increased but was not significant at any follow-up interval. The OR was 1.02 (95% CI 0.67 to 1.55) at 1 year, 1.21 (95% CI 0.93 to 1.32) at 2 years, and 1.34 (95% CI 0.96 to 1.87) at 5 years. Studies used varying definitions for progressionfree survival, generally based on prostate-specific antigen failure, bone scan changes, or clinical evaluation. The pooled OR for progression-free survival was determined for specific follow-up time intervals and progressively decreased at 2 years and 5 years. There were no differences detected in progression-free survival at later follow-up. The pooled OR at 1 year was 1.38 (95% CI 1.15 to 1.67), 1.19 (95% CI 0.97 to 1.46) at 2 years, and 1.14 (95% CI 0.77 to 1.68) at 5 years. Similarly, cancer-specific survival progressively increased over time and became statistically significant at 5 years. However, the results are limited by the fact that only two studies reported 5-year results for progression-free and cancer-specific survival, and both of these studies were positive.9,10 The pooled OR of cancer-specific survival was 1.20 (95% CI 0.92 to 1.57) at 1 year, 1.22 (95% CI 0.86 to 1.73) at 2 years, and 1.58 (95% CI 1.05 to 2.37) at 5 years. Adverse events were categorized by system and were pooled to estimate the frequency of event occurrence. The major differences between those assigned to monotherapy compared to those assigned to combined androgen blockade included diarrhea (1.8% versus 9.7%), gastrointestinal pain (1.6% versus 7.4%), and nonspecific ophthalmologic events (5.4% versus 29%). The occurrence of adverse events resulted in a withdrawal rate of 10% for combined androgen blockade. Only one study assessed the effect of combined androgen blockade on quality of life.22 After a 3-month follow-up patients receiving flutamide and orchiectomy reported more diarrhea (8.6% versus 2.7%) than those treated with orchiectomy alone. At both the 3-month and 6-month follow-up, patients receiving combined androgen blockade reported significantly worse emotional functioning. COMMENT We found that the pooled estimates of the ORs for survival increased in favor of combined androgen blockade over time and were small, but statistically significant, when evaluated at 5-years of follow-up but were not significantly different at 1-year and 2-year follow-up periods. Progressionfree survival appeared to be improved with com729

FIGURE 1. Studies for advanced prostate cancer that compared combined androgen blockade (CAB) versus monotherapy for overall survival of (A) 1 year, (B) 2 years, and (C) 5 years.

bined androgen blockade at the end of the first year, based on positive findings reported for three of seven studies4,7,9 –11,16,19 but not at later followup. Quality-of-life data were reported in one study, 730

thus limiting our ability to incorporate this outcome measure in the overview analyses. Methodologic considerations are important to consider. All three recent overviews included inUROLOGY 57 (4), 2001

formation on 20 clinical trials with nonsteroidal antiandrogens. However, the effort of the PCTCG was based on patient-level data, and the two other overviews were based on literature reviews. Our Cochrane Review used the OR as the effect measure; the evidence report prepared by the Blue Cross and Blue Shield Association’s Technology Evaluation Center, an AHRQ Evidence-Based Practice Center, used the hazard ratio; and the PCTCG used log-rank mortality statistics. The use of ORs in deriving treatment-effect estimates leads to conservative assessments. However, the different analytic methods yielded surprisingly similar estimates. Also, analytic difficulties occur when a variety of therapies and controls are evaluated. The limitations of this study must be addressed. First, the quality of the analyses in this literaturebased overview depends on the quality of the original data. Only seven of the trials included 5-year follow-up for overall survival data, with three of these studies being ones for which statistically significant increased survival with combined androgen blockade were noted. However, it is reassuring that the estimates of increased survival derived in this literaturebased effort is about the same as that included in the PCTCG overview (2.9% increase in 5-year survival, 2P ⫽ 0.005). The follow-up by the Trialists was typically for 5 years.25 Similarly, although progressionfree survival was improved at 1-year of follow-up, this estimate was based on data included in only seven clinical trials. Three of these seven studies were the clinical trials that found statistically significant increases in survival with combined androgen blockade.7,9,10 Second, randomized clinical trials have not addressed second-line use of nonsteroidal antiandrogen, after initial progression. This strategy is undoubtedly less costly. Third, class differences in nonsteroidal antiandrogens exist, such that diarrhea, the most common toxicity of flutamide therapy, may be less of a concern with agents such as bicalutamide.28 Phase III studies with bicalutamide as combined androgen blockade versus monotherapy have not been reported. In conclusion, the most recent overviews consistently indicate that there is about a 2% to 5% improvement in 5-year survival rates with combined androgen blockade.2,25 Uncertainty centers on whether this increase is clinically meaningful. We feel that clinicians and patients should discuss the likelihood of a small to modest benefit for some individuals, the likelihood of $3600 annually in out-of-pocket costs for most patients, and the potential for side effects. There is empirical evidence that these discussions are occurring in some health care settings. For example, the most common reason that veterans with prostate cancer shift to the Veterans Administration system is because of antiandrogen costs, reflecting gaps in the availability of UROLOGY 57 (4), 2001

insurance coverage for prescription drugs in this population.29 We conclude that clinicians and patients should discuss costs, potential toxicities, and potential benefits when considering the addition of a nonsteroidal antiandrogen to castration. ACKNOWLEDGMENT. To the Blue Cross and Blue Shield Association’s Technology Evaluation Center, an Evidence-Based Practice Center of the AHRQ, for contributing their combined androgen blockade database to support the Cochrane review on which this report is based. The authors also acknowledge Roderick MacDonald and Indulis Rutkis for their technical assistance with the Cochrane Review. REFERENCES 1. Greenlee RT, Murray T, Bolden S, et al: Cancer statistics, 2000. CA Cancer J Clin 50: 7–33, 2000. 2. Aronson N, Seidenfeld J, Samson DJ, et al: Relative effectiveness and cost-effectiveness of methods of androgen suppression in the treatment of advanced prostatic cancer. Evidence Report/Technology Assessment Number 4. (Prepared by the Blue Cross and Blue Shield Association EvidenceBased Practice Center under Contract No. 290-97-0015.) Agency for Health Care Policy and Research Publication No. 99-E0012. Rockville, Md: Agency for Health Care Policy and Research, May 1999. 3. Beland G, Eihali M, Fradet Y, et al: A controlled trial of castration with and without nilutamide in metastatic prostate carcinoma. Cancer 66S(5 suppl): 1074 –1079, 1990. 4. Boccardo F, Pace M, Rubagotti A, et al, for the Italian Prostatic Cancer Project (PONCAP) Study Group: Goserelin acetate with or without flutamide in the treatment of patients with locally advanced or metastatic prostate cancer. Eur J Cancer 29A: 1088 –1093, 1993. 5. Bono AV, DiSilverio F, Robustelli della Cuna G, et al, and the members of the Italian Leuprolide Group. Complete androgen blockade versus chemical castration in advanced prostatic cancer: analysis of an Italian multicentre study. Urol Int 60(suppl 1): 18 –24, 1998. 6. Brisset JM, Boccon-Gibod L, Botto H, et al: Anandron compared to surgical castration in previously untreated stage D prostate cancer: a multicenter comparative study of two doses of the drug and of a placebo. Prog Clin Biol Res 243A: 411– 422, 1987. 7. Crawford ED, Eisenberger M, McLeod DG, et al: A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med 321: 419 – 424, 1989. 8. Crawford ED, Kasmis BS, Gandara D, et al: A randomized, controlled clinical trial of leuprolide and anandron versus leuprolide and placebo for advanced prostate cancer. Proc Am Soc Clin Oncol 9: 523A, 1990. 9. Denis LJ, Carnelro de Moura JL, Bono A, et al, for the EORTC GU Group and EORTC Data Center: Goserelin acetate and flutamide versus bilateral orchiectomy: a phase III EORTC trial (30853). Urology 42: 119 –129, 1993. 10. Dijkman GA, Janknegt RA, De Reijke TM, et al, for the International Anandron Study Group: Long-term efficacy and safety of nilutamide plus castration in advanced prostate cancer, and the significance of early prostate specific antigen normalization. J Urol 158: 160 –163, 1997. 11. Eisenberger M, Crawford ED, Blumenstein B, et al: A controlled trial of surgical castration with and without flutamide (NCI 0105). N Engl J Med 339: 1036 –1042, 1998. 12. Ferrari P, Castagnetti G, Ferrari G, et al: Combination treatment versus LhRH alone in advanced prostatic cancer. Urol Int 56(suppl 1): 13–17, 1996. 13. Fourcade RO, Colombel P, Mangin M, et al: Zoladex 731

plus flutamide versus zoladex plus placebo in advanced prostatic carcinoma: extended follow-up of the French multi-center study, in Murphy G, and Khoury S (Eds): Proceedings of the Third International Symposium on Recent Advances in Urological Cancers. 1993, p 102–106. 14. Iverson P, Rasmussen F, Klarskov P, et al: Long-term results of Danish Prostatic Cancer Group trial 86. Goserelin acetate plus flutamide versus orchiectomy in advanced prostate cancer. Cancer 72S: 3851–3854, 1993. 15. Knonagel H, Bolle JF, Hering F, et al: Therapy of metastatic prostatic cancer by orchiectomy plus Anandron versus orchiectomy plus placebo. Initial results of a randomized multicenter study. Helv Chir Acta 56: 343–345, 1989. 16. Namer M, Toubol J, Caty A, et al: A randomized doubleblind study evaluating Anandron associated with orchiectomy in stage D prostate cancer. J Steroid Biochem Mol Biol 37: 909 –915, 1990. 17. Navratil H. Double-blind study of Anandron versus placebo in stage D2 prostate cancer. Preliminary results of a randomized, double-blind study. Am J Clin Oncol 11(suppl 2): S191–S196, 1988. 18. Periti P, Rizzo M, Mazzei T, et al: Depot leuporelin acetate alone or with nilutamide in the treatment of metastatic prostate carcinoma: interim report of a multi-center, doubleblind study. Can J Infect 6(suppl C): 292C, 1995. 19. Schulze H, Kaldenoff H, and Senge T: Evaluation of total versus partial androgen blockade in the treatment of advanced prostate cancer. Urol Int 43: 193–197, 1988. 20. Tyrrell CJ, Altwein JE, Klippel F, et al, for the International Prostate Cancer Study Group: A multicenter randomized trial comparing the luteinizing hormone-releasing hormone analog goserelin acetate alone with flutamide in the treatment of advanced prostate cancer. J Urol 146: 1321–1326, 1991.

732

21. Zalcberg JR, Raghaven D, Marshall V, et al: Bilateral orchidectomy and flutamide versus orchidectomy alone in newly diagnosed patients with metastatic carcinoma of the prostate—an Australian multicentre trial. Br J Urol 77: 865– 869, 1996. 22. Moinpour CM, Savage MJ, Troxel A, et al: Quality of life in advanced prostate cancer: results of a randomized therapeutic trial. J Natl Cancer Inst 90: 1537–1544, 1998. 23. Schmitt B, Bennett CL, Seidenfeld J, et al: Maximal androgen blockade for advanced prostate cancer (Cochrane Review), in: The Cochrane Library, Issue 3, 2000. Oxford: Update Software. 24. Prostate Cancer Trialists Collaborative Group. Maximum androgen blockade in advanced prostate cancer: an overview of 22 randomized trials with 3283 deaths in 5710 patients. Lancet 346: 265–269, 1995. 25. Prostate Cancer Trialists Collaborative Group. Maximum androgen blockade in advanced prostate cancer: an overview of the randomized trials. Lancet 355: 1491–1498, 2000. 26. Reese DM. Choice of hormonal therapy for prostate cancer. Lancet 355: 1474 –1475, 2000. 27. Laufer M, Denmeade SR, Sinibaldi VJ, et al: Complete androgen blockade for prostate cancer: what went wrong? J Urol 164: 3–9, 2000. 28. Schellhammer P, Sharifi R, Block N, et al: A controlled trial of bicalutamide versus flutamide, each in combination with luteinizing hormone-releasing hormone analogue therapy in patients with advanced prostate cancer. Urology 45: 745–752, 1995. 29. Kim SP, Moran EM, Bowes ED, et al: The antiandrogen transfer. Prostate J 2: 88 –93, 2000.

UROLOGY 57 (4), 2001