We thank the patients and their families for their participation in this study. We also thank the physicians and support staff in each institute and Dr. Sharon Gladwin from Complete Medical Communications, who provided medical writing support.
Institutions participating in the study: Hokkaido University Hospital, Sapporo Medical University Hospital, Hirosaki University School of Medicine and Hospital, Tohoku University Hospital, Yamagata University Hospital, Tsukuba University Hospital, Teikyo University Chiba Medical Center, Asahi Hospital, Nihon University Itabashi Hospital, Toranomon Hospital, Keio University Hospital, Showa University Hospital, the University of Tokyo Hospital, the Jikei University School of Medicine, Tokyo Medical University Hospital, Tokyo Women's Medical University Hospital, Yokohama City University Hospital, Kitasato University Hospital, Fujieda Municipal General Hospital, Niigata University Medical and Dental Hospital, Niigata Cancer Center Hospital, Kanazawa University Hospital, Nagoya Daini Red Cross Hospital, Nagoya City University Hospital, Gifu University Hospital, University Hospital Kyoto Prefectural University of Medicine, Kyoto University Hospital, Nara Medical University Hospital, Osaka University Hospital, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka City University Hospital, Kansai Medical University Takii Hospital, Kobe University Hospital, Nishi-Kobe Medical Center, Okayama University Hospital, Kawasaki Medical School Hospital, Kurashiki Central Hospital, Hiroshima University Hospital, Shimane University Hospital, Tokushima University Hospital, Shikoku Cancer Center, Kochi Medical School Hospital, Kyushu University Hospital, Harasanshin Hospital, Nagasaki University Hospital of Medicine and Dentistry, Sasebo General Hospital, and Nagasaki Medical Center.
See also on pages 3376-8.
A previously reported, double-blind, randomized, multicenter phase 3 trial in 205 patients with stage C/D prostate cancer compared combined androgen blockade (CAB) with luteinizing hormone-releasing hormone agonist (LHRH-A) plus bicalutamide 80 mg versus LHRH-A plus bicalutamide-matching placebo (LHRH-A monotherapy). The analysis at a median follow-up of 2.4 years indicated that CAB significantly (P < .001) prolonged the time to progression and the time to treatment failure. In the current report, survival data from a long-term follow-up (median, 5.2 years) were analyzed.
All deaths irrespective of cause and all prostate cancer-specific deaths were recorded. The data were analyzed using Cox regression analysis and the log-rank test.
At a median follow-up of 5.2 years, a significant overall survival advantage was observed in favor of CAB over LHRH-A monotherapy (Cox regression analysis: hazard ratio, 0.78; 95% confidence interval, 0.60-0.99; P = .0498; log-rank test: P = .0425). The difference in cause-specific survival between the 2 groups was not significant. The achievement of a prostate-specific antigen (PSA) nadir concentration ≤1 ng/mL was a prognostic factor for improved survival. More patients attained PSA nadir concentrations ≤1 ng/mL with CAB compared with patients who received LHRH-A monotherapy (81.4% vs 33.7%; P < .001).
Combined androgen blockade (CAB), consisting of a nonsteroidal antiandrogen plus either a luteinizing hormone-releasing hormone agonist (LHRH-A) or bilateral orchiectomy, is a standard treatment for advanced prostate cancer. Furthermore, early initiation of CAB is a potential option for patients who are at high risk of prostate cancer-specific death after failing radical treatment for clinically localized disease.1, 2
The use of CAB therapy over castration alone has been widely debated because of conflicting efficacy data from individual clinical trials as well as tolerability and cost issues. In 2000, the Prostate Cancer Trialists' Collaborative Group (PCTCG) published a large meta-analysis of all randomized trials initiated before 1991 that compared CAB with castration alone in patients with advanced prostate cancer (27 studies; n = 8275).3 The results demonstrated that CAB with a nonsteroidal antiandrogen (flutamide or nilutamide) reduced the risk of death by 8% compared with castration alone (P = .005). However, the survival benefit was so small that CAB could not be widely recommended in clinical practice.
Bicalutamide (Casodex; Astra-Zeneca Pharmaceuticals, Wilmington, Del), a nonsteroidal antiandrogen with a better tolerability profile than flutamide and nilutamide,4, 5 was not included in the PCTCG meta-analysis, because no randomized trial data for CAB with bicalutamide versus castration alone were available at the time. However, Klotz and colleagues6 estimated the efficacy of CAB with bicalutamide 50 mg using validated statistical methodology7 to combine PCTCG meta-analysis data with data from a phase 3 trial of 2 CAB regimens (LHRH-A plus bicalutamide 50 mg vs LHRH-A plus flutamide). That analysis suggested that CAB with bicalutamide 50 mg could reduce the risk of death by 20% compared with castration alone (hazard ratio [HR] 0.80; 95% confidence interval [CI], 0.66-0.98).
A multicenter, double-blind, controlled trial has compared CAB directly with bicalutamide 80 mg (the 80 mg dose is licensed for CAB and monotherapy in Japan) versus castration alone.8, 9 In that phase 3 trial, 205 Japanese patients with stage C/D prostate cancer were randomized to receive either CAB with an LHRH-A plus bicalutamide 80 mg (n = 102) or LHRH-A monotherapy (LHRH-A plus bicalutamide-matching placebo; n = 103). The proportion of patients who achieved a prostate-specific antigen (PSA) level ≤4 ng/mL at 12 weeks was significantly greater for CAB compared with LHRH-A monotherapy (79.4% vs 38.6%, respectively; P < .001).8 CAB also improved the 12-week overall tumor response rate versus LHRH-A monotherapy (77.5% vs 65.3%, respectively; P = .063).8 At a median follow-up of 2.4 years, CAB significantly prolonged the median time to treatment failure (117.7 weeks vs 60.3 weeks; P < .001) and time to progression (not reached vs 96.9 weeks; P < .001), compared with LHRH-A monotherapy.9 It is noteworthy that the incidence of adverse events and withdrawals was similar between the 2 treatment groups.8, 9
Results from that trial, along with those from the analysis by Klotz and colleagues,6 are cited in guidelines published by the American Society of Clinical Oncology to support their recommendation that CAB with bicalutamide should be considered for patients with advanced prostate cancer.10 A subsequent exploratory analysis of the phase 3 study revealed that, in terms of the median time to progression, the benefit of CAB versus LHRH-A monotherapy was greater in patients with stage C disease (median not reached vs 134.1 weeks; P < .001)9 than in patients with stage D disease (98.4 weeks vs 64.1 weeks; P = .024) at diagnosis.11 Furthermore, in patients with stage C disease, CAB significantly prolonged the median time to progression compared with LHRH-A monotherapy, irrespective of histologic grade, patient age, or PSA level at diagnosis.11 Those data support the use of CAB with bicalutamide in patients with locally advanced, metastasis-free prostate cancer.
When the phase 3 study was completed at a median follow-up of 2.4 years, survival data were immature and, consequently, no significant differences were observed between the 2 treatment groups in terms of overall or cause-specific survival.9 We instigated the Study Group for the Combined Androgen Blockade Therapy of Prostate Cancer, comprised of the investigators who had participated in the original study, to conduct a long-term follow-up of patients who were enrolled in the original study and remained alive after the trial was completed. The current article presents the results of the survival analysis from the long-term follow-up study.
MATERIALS AND METHODS
Detailed methods for the original phase 3 study were published previously8, 9 and are briefly summarized here. Patients with histologically confirmed, previously untreated, advanced (stage C/D12) prostate cancer were recruited at 49 centers in Japan between February 2000 and December 2001. All patients received an LHRH-A according to the investigator's choice, either goserelin acetate (Zoladex; Astra-Zeneca Pharmaceuticals) 3.6 mg or leuprorelin acetate (Leuprin; Takeda Chemical Industries, Osaka, Japan) 3.75 mg, given as a subcutaneous injection every 4 weeks. In addition, patients were randomized in a 1:1 ratio to receive either oral bicalutamide 80 mg or matching placebo once daily. Randomized treatment was given in a double-blind manner until September 2002, when the code was broken for ethical reasons. Subsequently, patients in the LHRH-A monotherapy group discontinued placebo, and patients in the CAB group continued their treatment in an open-label manner. Patients continued to receive randomized treatment until the end of November 2003 or until there was evidence of disease progression or any other event leading to withdrawal. Patients who had disease progression during LHRH-A monotherapy were treated at the investigator's discretion, with the option of adding bicalutamide 80 mg (deferred CAB). For patients who experienced disease progression in the CAB group, bicalutamide was discontinued, and the patient was monitored for antiandrogen withdrawal syndrome at the investigator's discretion. The primary endpoints were PSA normalization rate and overall tumor response at 12 weeks and the percentage of withdrawals because of adverse drug reactions. Secondary endpoints included time to treatment failure, time to progression, survival, quality of life (QoL), time to PSA normalization, and the incidence of adverse events and adverse drug reactions.
After the original phase 3 study, the Study Group for the Combined Androgen Blockade Therapy of Prostate Cancer conducted a follow-up study of patients who were enrolled in the original study and were still alive when the original study was completed (from December 2003). The institutional review board at each medical center approved the follow-up study, and all patients provided written informed consent.
Assessments and Statistical Analyses
All deaths, irrespective of cause, and all prostate cancer-specific deaths were recorded for 3 years from the completion of the original study (the follow-up period ended in March 2007). Confirmed deaths from the original study plus deaths that were recorded during the follow-up period were analyzed at the University of Tsukuba. Overall survival and cause-specific survival were assessed using a Cox proportional hazards regression model with covariates for randomized treatment, clinical stage, age, performance status, PSA level at diagnosis, histologic grade, and the type of LHRH-A received. An additional Cox regression analysis of overall survival was performed with covariates for the PSA nadir level achieved during the original study irrespective of whether randomized treatment had been discontinued (≤1 ng/mL or >1 ng/mL), randomized treatment, clinical stage, age, performance status, histologic grade, and the type of LHRH-A received. Results from the Cox regression analyses were confirmed using the log-rank test.
For patients who achieved a PSA nadir of ≤1 ng/mL during the original study, the time taken to reach the nadir level was assessed. The number of patients achieving PSA nadir levels >4 ng/mL, from ≤4 to >1 ng/mL, from ≤1 to >0.2 ng/mL, and ≤0.2 ng/mL during the administration of randomized treatment only also was investigated.
Of 205 patients who entered the original study, 203 patients (CAB, n = 102; LHRH-A monotherapy, n = 101) received randomized treatment (Fig. 1).9 Two patients in the LHRH-A monotherapy group withdrew from the study before commencing treatment (because of deterioration of performance status and failure to attend a hospital visit, respectively). The demographics and baseline characteristics of the patients who received randomized treatment were similar between the 2 treatment arms and are summarized in Table 1.9 For the majority of men in both treatment arms, the choice of LHRH-A was goserelin acetate (75.5% and 78.2% for the CAB and LHRH-A arms, respectively).
Table 1. Demographic and Baseline Characteristics of Patients in the Original Study Population*
In total, 172 patients (CAB, n = 89; LHRH-A monotherapy, n = 83) remained alive in December 2003 after the original study had completed. At the end of March 2007, in total, 139 patients (CAB, n = 76; LHRH-A monotherapy, n = 63) remained alive.
At a median follow-up of 5.2 years, there were fewer overall deaths with CAB than with LHRH-A monotherapy (26 deaths vs 38 deaths, respectively). A significant overall survival advantage was observed in favor of CAB over LHRH-A monotherapy (Cox regression analysis: HR, 0.78; 95% CI, 0.60-0.99; P = .0498; log-rank test: P = .0425) (Fig. 2). The 5-year overall survival rate estimated by the Kaplan-Meier method was 75.3% for CAB versus 63.4% for LHRH-A monotherapy. The results from the subgroup analysis according to disease stage (stage C/D1 and stage D2) are shown in Figure 3.
CAB also was associated with fewer cause-specific deaths compared with LHRH-A monotherapy (14 deaths vs 22 deaths, respectively). The difference in cause-specific survival between the 2 groups was not significant (Cox regression analysis: HR, 0.79; 95% CI, 0.55-1.11; P = .1703; log-rank test: P = .0918) (Fig. 4).
Overall Survival and Prostate-Specific Antigen Nadir Level
During the original study, PSA levels decreased to ≤1 ng/mL in 137 of 203 patients (67%). Overall survival was prolonged significantly in patients who attained a PSA nadir ≤1 ng/mL compared with those who did not (death rate: 19.7% [27 of 137 patients] vs 56.1% [37 of 66 patients], respectively; HR, 0.34; 95% CI, 0.20-0.59; P = .0001; log-rank test: P < .0001) (Fig. 5). In total, 75% of patients who achieved a PSA nadir ≤1 ng/mL had reached that level within the first 192 days of the study (Fig. 6). During randomized treatment, PSA nadir concentrations ≤1 ng/mL were achieved by 83 of 102 patients (81.4%) who received CAB and by 34 of 101 patients (33.7%) who received LHRH-A monotherapy (Fisher exact test: P < .001) (Table 2).
Table 2. Patients Who Achieved Defined Prostate-Specific Antigen Nadir Levels During Randomized Treatment in the Original Phase 3 Study
In this report, long-term follow-up data from a phase 3 study of CAB with bicalutamide 80 mg versus LHRH-A monotherapy alone have demonstrated a significant overall survival advantage in favor of CAB. The overall survival advantage for CAB is consistent with previous observations from this study of prolonged time to treatment failure and time to progression.9 In particular, the magnitude of the reduction in risk of death reported for CAB with bicalutamide 80 mg (22%) concurs with that estimated by Klotz and colleagues6 for CAB with bicalutamide 50 mg (20%). In most countries, bicalutamide is licensed at a dose of 50 mg daily for use in CAB. However, based on pharmacokinetic and pharmacodynamic data,13 the only approved dose of bicalutamide in Japanese men is 80 mg per day for monotherapy. A previous pilot study of LHRH-A in combination with bicalutamide 80 mg identified no safety concerns14; therefore, the 80 mg dose of bicalutamide is used both for monotherapy and for CAB in Japan. A comparison between our study results and Western CAB data with bicalutamide 50 mg should be considered as the next step.
In total, 30 patients in the CAB group experienced disease progression during the original phase 3 study, including at least 18 patients who were observed for antiandrogen withdrawal syndrome, and 7 patients (39%) responded (median response duration, 58 weeks).9 Of 57 patients in the LHRH-A monotherapy group who had disease progression during the phase 3 study, at least 40 patients subsequently received second-line CAB with bicalutamide 80 mg, and 31 patients (78%) responded to that treatment (median response duration, 40 weeks).9 Currently, CAB is used widely in Japan and accounts for approximately 70% of primary hormone therapy for prostate cancer.15 For patients who receive LHRH-A monotherapy as initial treatment and subsequently experience disease progression, second-line therapy is usually the addition of an antiandrogen to their regimen (deferred CAB therapy). Because the majority of patients who progressed in the LHRH-A monotherapy group received second-line CAB therapy, our study can be considered a comparison of immediate versus deferred CAB. Consequently, results from the current follow-up study suggest that immediate CAB may be superior to deferred CAB in terms of prolonging overall survival.
Although it was not predefined in the protocol, a subgroup analysis of overall survival by clinical stage was performed for reference. Consequently, the difference in overall survival between CAB and LHRH-A monotherapy was greater in the patients who had stage C/D1 disease. In the original phase 3 study, the same tendency was observed in the time to progression for CAB in the patients who had stage C disease, suggesting that the long-term prognosis for patients who have stage C disease and are treated with CAB can be expected to be markedly better than that of the patients who are treated with LHRH-A monotherapy.9 Sylvester et al. reported that, among patients with stage D2 prostate cancer who either underwent orchiectomy or received CAB (goserelin + flutamide), the survival benefit of CAB was greater for patients who had mild bone metastasis than for those who had more advanced disease.16 On the basis of these results, the survival benefit of CAB versus LHRH-A monotherapy is expected to be much greater for patients who have early stage disease.
Our follow-up study revealed no significant difference in cause-specific survival between CAB and LHRH-A monotherapy (P = .0918). This is unsurprising, because the analysis lacked statistical power to detect a significant difference in cause-specific mortality in light of the low number of prostate cancer-related deaths (14 patients on CAB and 22 on LHRH-A monotherapy). To observe a treatment difference in cause-specific survival, longer follow-up or a larger patient population may be necessary.
Previous studies have suggested that the normalization of PSA by hormone therapy may be associated with prolonged time to progression and survival.17, 18 Because of an exploratory multivariate analysis with PSA cutoff levels of 4 ng/mL, 2 ng/mL, 1 ng/mL, 0.5 ng/mL, and 0.2 ng/mL, the use of 1 ng/mL produced a stable and better fitting model with a small P value and variance of estimated values. Therefore, we used a cutoff level of 1 ng/mL for our analysis of overall survival. Data from our study indicated that patients who attained a PSA nadir ≤1 ng/mL survived significantly longer than patients who had PSA levels that remained >1 ng/mL. It also was apparent that patients who received CAB achieved lower PSA nadir levels than patients who received LHRH-A monotherapy. It is noteworthy that PSA levels fell below 0.2 ng/mL (the detection limit) in 89% of patients who had a PSA nadir ≤1 ng/mL in the CAB group, compared with only 41% of patients who had a PSA nadir ≤1 ng/mL in the LHRH-A monotherapy group. Therefore, the PSA reduction associated with CAB appears to be important clinically in terms of prolonging overall survival. Among the patients who achieved a PSA nadir ≤1 ng/mL in the original study, 75% had attained this nadir within approximately 6 months (192 days). This suggests that, if no therapeutic effect is observed within the first 6 months of treatment, then a change of therapy should be considered.
A primary obstacle to the wider use of CAB is the potential for increased side effects and costs compared with castration alone. Indeed, compared with castration alone, CAB with flutamide is associated with an increased incidence of gastrointestinal disorders and hepatotoxicity, whereas CAB using nilutamide is associated with an increased incidence of visual disorders.19 However, in the phase 3 study of CAB with bicalutamide 80 mg versus LHRH-A monotherapy, there was no difference between the 2 treatment arms regarding the percentage of withdrawals because of adverse drug reactions (primary safety endpoint) or adverse drug reaction profiles.8, 9 QoL was assessed as a secondary endpoint in this study using the Japanese version of the Functional Assessment of Cancer Therapy-Prostate questionnaire.20 These data demonstrated that, compared with LHRH-A monotherapy, CAB with bicalutamide did not reduce overall QoL and provided an early improvement in QoL related to micturition disorder and pain.21
Nishimura and colleagues22 conducted a cost-effectiveness analysis of CAB with bicalutamide 80 mg based on efficacy data from the phase 3 study and medical costs in Japan. Those authors concluded that CAB was a cost-efficient therapy with an incremental cost effectiveness ratio of approximately ¥1,560,000 (approximately $14,000 in US dollars). This is consistent with results from similar analyses conducted in the United States. For example, Ramsey and colleagues23 demonstrated that the incremental cost per quality-adjusted life-year (QALY) gained for CAB with bicalutamide 50 mg versus CAB with flutamide was $22,000 at 5 years and $16,000 at 10 years. Likewise, Penson and colleagues24 estimated that the cost per QALY of CAB with bicalutamide 50 mg was $33,677 and $20,053 at 5 years and 10 years, respectively, compared with castration alone. These studies support CAB with bicalutamide as a cost-effective treatment strategy for patients with advanced prostate cancer.
In conclusion, the long-term follow-up of the first double-blind controlled study to directly compare CAB with bicalutamide 80 mg versus LHRH-A monotherapy has demonstrated a statistically significant overall survival benefit in favor of CAB. The advantage in overall survival, together with the previously reported significant improvements in time to treatment failure and time to progression, which were achieved without reducing tolerability, indicate that CAB with bicalutamide is a recommendable first-line therapy option for patients with locally advanced or metastatic prostate cancer.
Conflict of Interest Disclosures
Financial sponsorship for this study and publication was provided by the Advanced Clinical Research Organization (a nonprofit organization).