Prostate cancer is a disease of older men, and is second only to lung cancer in cancer mortality for men. For 1998, it was estimated that 184,500 new cases of prostate cancer would be diagnosed, and 39,200 men would die of prostate cancer that year.
In 1994, the total Medicare expenditure for treatment of prostate cancer was $1,411,687,900. Of the total, $477,851,000 was for androgen suppression therapy using luteinizing hormone-releasing hormone (LHRH) agonists. The prevalence of prostate cancer, and the expenditures for its treatment, are likely to increase with the aging of the population and the trend to earlier detection of the disease.
Three key issues are addressed in the report:
Two supplementary analyses were also conducted for each key question: (1) meta-analysis of overall survival at 2 or 5 years (as permitted by the data), and (2) cost-effectiveness analysis.
The following agents were included in the assessment: LHRH agonists (leuprolide, goserelin, buserelin); nonsteroidal antiandrogens (flutamide, nilutamide, bicalutamide); and the steroidal antiandrogen, cyproterone.
Each agent was compared to orchiectomy or diethylstilbestrol (DES, 3 mg/day), which was confirmed to be equivalent to orchiectomy as a comparator. No trials compared the effectiveness of two LHRH agonists or two antiandrogens as monotherapies.
The evidence was examined in the aggregate (i.e., combined androgen blockade using any antiandrogen) and by class of antiandrogen (nonsteroidal antiandrogens or cyproterone). In addition, the class of nonsteroidal antiandrogens was examined by agent.
The population of interest to this report is men with advanced prostate cancer, including: (1) regional or disseminated metastases (D1 or D2; N+/M0 or M1); (2) minimally advanced disease (C; T3-4/N0 or Nx/M0); and, only for immediate versus deferred therapy, (3) rising PSA, or other signs of progression after definitive therapy for early stage disease.
Outcomes of interest are: overall, cancer-specific, and progression-free survival; time to treatment failure; adverse effects; and quality of life. Where available, data on patient preferences are included. We also looked for treatment outcomes that were analyzed by patient subgroups based on prognostic factors such as tumor grade, extent of disease, and performance status.
The MEDLINE, Cancerlit, and EMbase databases were searched from 1966 to March 1998, and Current Contents on Diskette through August 24, 1998, for all articles that included at least one of the following terms in their titles, their abstracts, or their keyword lists: leuprolide (Lupron®); goserelin (Zoladex®); buserelin (Suprefact®); flutamide (Eulexin®); nilutamide (Anandron®, Nilandron®); bicalutamide (Casodex®); cyproterone acetate (Androcur®); diethylstilbestrol (DES); and orchiectomy (castration, orchidectomy).
The search results were then limited to studies on human subjects indexed under the Medical Subject Heading (MeSH) "prostatic neoplasms." Randomized controlled trials were identified using the UK Cochrane Center search strategy. Total retrieval was 1,477 references. These references were checked against the Cochrane Controlled Trials Register, the CENTRAL register, and trials cited in two recent meta-analyses. No additional trials were identified.
The study selection criteria for this review limited reports of efficacy outcomes to randomized controlled trials. Randomized controlled trials that only compared different doses of the same agent were excluded. For adverse events, phase II studies that reported on withdrawals from therapy were also included. All studies reporting on quality of life were included.
The meta-analysis combined data on overall survival using a random effects model. A hazard ratio of 1.0 indicates that patients treated with the therapy of interest and patients treated with the control had an equal chance of death from any cause. An initial analysis was performed to determine whether the results of orchiectomy and diethylstilbestrol are comparable and thus whether it is valid to pool studies in which the control arms used either of these monotherapies. Separate analyses were also used to compare the available monotherapies, compare monotherapy with combined androgen blockade, and compare the outcomes of immediate androgen suppression with those of deferred treatment. Sensitivity analyses were performed. One restricted the analysis to subjects with stage D2/M1 disease; the second was restricted to higher quality trials. A trial was classified as higher quality when it was double-blinded (except when orchiectomy was one of the treatments) and reported outcomes based on an intention-to-treat analysis. For combined androgen blockade, sensitivity analyses compared trials reporting 5-year survival to those reporting 2-year survival to test for a publication bias effect.
This report also evaluated the cost-effectiveness of androgen suppression strategies for patients with advanced prostate cancer. Using a decision analysis model, the cost-effectiveness analysis accounts for the benefits and harms of the interventions, including effects on quality of life, and captures a broad range of costs. Fundamental to the model is a representation of the natural history of advanced prostate cancer, in which a hypothetical cohort of patients is followed over time. Health states associated with prostate cancer and transitions between health states were derived from the literature. The model used point estimates and confidence intervals for overall survival rates derived from the meta-analyses. The model is conducted from a societal perspective, in which all costs and benefits are expressed in present value terms with a rate of time discount of 3 percent. This approach is consistent with the guidelines of the Panel on Cost-Effectiveness in Health and Medicine. The results are reported as a cost-effectiveness ratio, or the dollar cost per unit improvement in health in comparison with a well-defined alternative. The cost-effectiveness ratio is incremental.
This report has undergone extensive peer review. The first draft was reviewed by the Blue Cross and Blue Shield Association Technology Evaluation Center Medical Advisory Panel, which includes nationally recognized experts in technology assessment, cost-effectiveness analysis, oncology, and urology. The revised draft of this report was reviewed by 10 external experts on prostate cancer, an expert in cost-effectiveness analysis, an expert in systematic review and meta-analysis, and two patient representatives. The American Urological Association, the American Society of Clinical Oncology, and the American Society for Therapeutic Radiology and Oncology each appointed one of the 10 prostate cancer experts. In addition, we requested reviews from all pharmaceutical companies whose product was included in this report. Three companies commented on the draft.
Ten trials (n=1,908) compared an LHRH agonist either to orchiectomy or to DES. Nine of the 10 reported data on overall survival and no trial found a statistically significant difference between treatments. The meta-analysis found that 2-year overall survival with an LHRH agonist is essentially equivalent to orchiectomy (hazard ratio 1.1262; 95 percent confidence interval 0.915 to 1.386).
No trial found a statistically significant difference between any LHRH agonist and orchiectomy or DES. The meta-analysis found similar hazard ratios when compared to orchiectomy: leuprolide, 1.0994 (95 percent confidence interval 0.207 to 5.835); goserelin 1.1172 (95 percent confidence interval 0.898 to 1.390); and buserelin 1.1315 (95 percent confidence interval 0.533 to 2.404).
Eight trials (n=2,717) compared a nonsteroidal antiandrogen to orchiectomy, DES, or an LHRH agonist. Two found a statistically significant difference favoring the control arm; an additional 14-15 percent of patients survived at 2 years. The meta-analysis found that the hazard ratio relative to orchiectomy was 1.2158 for nonsteroidal antiandrogens as a class (95 percent confidence interval 0.988 to 1.496), compared to 0.9835 for DES (95 percent confidence interval 0.764 to 1.267), and 1.1262 for LHRH agonists (95 percent confidence interval 0.915 to 1.386).
The frequency of withdrawal from therapy due to adverse events is the most reliable index reported for comparing the tolerability of the two drug classes. Withdrawals occurred less often among patients treated with an LHRH agonist (0-4 percent) than among patients treated with nonsteroidal antiandrogens (4-10 percent).
Impotence was more common among patients treated with orchiectomy or LHRH agonists compared to patients treated with nonsteroidal antiandrogens, but the available data are too inconsistent to quantify the differences. Hot flashes were more common and gynecomastia was less common among patients treated with LHRH agonists than among those treated with nonsteroidal antiandrogens.
Among the LHRH agonists, local pain, reactions, hypersensitivity, or development of a mass at the site of depot injections were very infrequent and occurred as often with leuprolide as they did with goserelin in the only study that directly compared the two.
Eighteen trials (n=5,485) reported no significant difference in overall survival, including the largest single trial (Intergroup trial [INT 0105]), conducted by the Southwestern Oncology Group (SWOG), which included 1,382 patients.
Three trials (n=1,386) reported a statistically significant difference in overall survival favoring the combined androgen blockade arm. The reported advantage in median survival ranged from 3.7 to 7 months; the advantage in 5-year survival ranged from 3 to 9 percent.
The meta-analysis found no difference between monotherapy and combined androgen blockade in overall survival at 2 years (hazard ratio = 0.970; 95 percent CI = 0.866 to 1.087). There was an advantage in overall survival for combined androgen blockade at 5 years (hazard ratio = 0.871; 95 percent CI= 0.805 to 0.942).
Only 10 trials reporting 2-year survival also reported 5-year survival, which represents 66 percent of the patients in the meta-analysis. The results of sensitivity analyses suggest that if complete 5-year data were available, the magnitude of benefit from combined androgen blockade would not be of greater clinical significance. The estimated combined hazard ratio for 5-year survival from all trials was 0.9146 for combined androgen blockade compared to monotherapy (95 percent CI= 0.8461 to 0.9887).
Only six trials reported outcomes stratified by prognostic group. Two trials reported that combined androgen blockade was of greater benefit than monotherapy for patients with good prognostic factors, but did not report whether these results were statistically significant. Three other trials, which reported on both good and poor prognostic subgroups, found no statistically significant differences in outcome between treatment arms for either subgroup. The SWOG Intergroup trial (INT 0105), the only trial prospectively designed and adequately powered to compare outcomes for good-risk patients, also found no significant difference in survival between combined androgen blockade and monotherapy.
Of the three trials that reported a statistically significant difference in survival favoring combined androgen blockade, two used flutamide and one used nilutamide. The meta-analysis found that combined androgen blockade using flutamide or nilutamide appears to be equivalent. The hazard ratio is 0.878 (95 percent CI = 0.564 to 1.368) in trials using nilutamide and 0.945 (95 percent CI = 0.779 to 1.147) in trials using flutamide. In the only trial that directly compared two different regimens for combined androgen blockade, there was no statistically significant difference in survival between men given flutamide or bicalutamide (hazard ratio = 0.87; 95 percent CI = 0.72 to 1.05).
Patients randomized to combined androgen blockade (10 percent) withdrew from treatment due to adverse effects more frequently than patients randomized to monotherapy (4 percent). The other evidence comparing the adverse effects of these treatments is limited. The available evidence is inconsistent with respect to which adverse effects are reported and how these adverse effects are measured.
In the recent SWOG trial (INT 0105) sub-study, the only trial that compared quality of life end points, patients randomized to combined androgen blockade reported more problems with emotional functioning and diarrhea than those randomized to monotherapy.
The body of evidence to address this question is limited to three trials (n=2,143), two of which were trials conducted in the 1960s (the Veterans Administration Cooperative Urological Research Group [VACURG] trials). The VACURG trials represent a markedly older and sicker population than patients treated in current clinical practice, so these results may not be generalizable. None of the three trials had a uniform protocol for initiating deferred therapy, so deferred therapy in these trials reflects the varied practices of the treating physicians. Indeed, some patients in the deferred treatment arms of these trials received no hormonal therapy prior to death.
Two (n=1,190) of the three trials reported a statistically significant difference in 5-year overall survival favoring the immediate therapy arm; 12 percent in a VACURG trial and 4 percent in a trial by the Medical Research Council (MRC). In the MRC trial, the benefit of immediate hormonal therapy was limited to the subgroup of M0 patients. The meta-analysis found no significant difference between immediate primary hormonal therapy and deferred therapy for survival at 5 years (hazard ratio = 0.914; 95 percent CI = 0.815 to 1.026), although the 95 percent confidence interval approaches statistical significance.
Two trials (n=506) reported a statistically significant difference (17 and 14 percent) in 5-year overall survival in favor of radiotherapy plus continued androgen suppression. Two trials (n=1,059) found no statistically significant difference, but one of these reported a statistically significant difference in 5-year overall survival in favor of radiotherapy plus continued androgen suppression (11 percent) in the subgroup of patients with Gleason scores of 8-10. However, none of these trials included a group of patients treated with androgen suppression alone. Consequently, there is no evidence to determine whether radiation therapy plus androgen suppression increases survival when compared with androgen suppression alone.
The meta-analysis found a significant difference in 5-year overall survival in favor of radiation therapy plus continued androgen suppression compared to radiation therapy alone (hazard ratio = 0.631, 95 percent CI = 0.479 to 0.831).
There is a large body of randomized controlled trials showing that orchiectomy and the available LHRH agonists are equally effective, and no LHRH agent is superior when adverse effects are considered. Combined androgen blockade has not been demonstrated to be of greater benefit than monotherapy for the aggregate population of patients with advanced prostate cancer or for the subpopulation of patients with good prognostic factors. Other patient subgroups that might benefit more from combined androgen blockade than monotherapy have not been well defined.
Randomized controlled trials are needed to assess the effectiveness of various strategies for the timing of androgen suppression, and should be prospectively designed to address whether there are subgroups of patients more likely to benefit from early initiation of androgen suppression.
Although there is uncertainty over whether there is a survival advantage, earlier and more intensive strategies of androgen suppression are being adopted. Moreover, there are scant data on how quality of life is affected. Evidence on the effects of alternative androgen suppression strategies on the quality of life is urgently needed.
AHCPR Pub No. 99-E011
Current as of January 1999
Internet Citation:
Relative Effectiveness and Cost-Effectiveness of Methods of Androgen Suppression in the Treatment of Advanced Prostatic Cancer. Summary, Evidence Report/Technology Assessment: Number 4, January 1999. Agency for Health Care Policy and Research, Rockville, MD. http://www.ahcpr.gov/clinic/prossumm.htm
Return to PSA Rising Magazine Upfront
Return to PSA Rising Magazine Grassroots