Fax: (604) 263-4306
Final results of the Canadian prospective phase II trial of intermittent androgen suppression for men in biochemical recurrence after radiotherapy for locally advanced prostate cancer
Article first published online: 16 JUN 2006
Copyright © 2006 American Cancer Society
Volume 107, Issue 2, pages 389–395, 15 July 2006
How to Cite
Bruchovsky, N., Klotz, L., Crook, J., Malone, S., Ludgate, C., Morris, W. J., Gleave, M. E. and Goldenberg, S. L. (2006), Final results of the Canadian prospective phase II trial of intermittent androgen suppression for men in biochemical recurrence after radiotherapy for locally advanced prostate cancer. Cancer, 107: 389–395. doi: 10.1002/cncr.21989
- Issue published online: 5 JUL 2006
- Article first published online: 16 JUN 2006
- Manuscript Accepted: 8 MAR 2006
- Manuscript Revised: 12 FEB 2006
- Manuscript Received: 2 DEC 2005
- Abbott Laboratories Ltd. (1995-1997)
- Berlex Canada Inc. (1995-2001)
- Liber Ero Foundation of Vancouver (2001-2006)
- biochemical recurrence;
- intermittent androgen suppression
This prospective Phase II study was undertaken to evaluate intermittent androgen suppression as a form of therapy in men with localized prostate cancer who failed after they received external beam irradiation.
Patients who demonstrated a rising serum prostate-specific antigen (PSA) level after they received radiotherapy and who were without evidence of distant metastasis were accepted into the study. Treatment in each cycle consisted of cyproterone acetate given as lead-in therapy for 4 weeks, followed by a combination of leuprolide acetate and cyproterone acetate, which ended after a total of 36 weeks.
Of 109 patients registered, 103 patients were eligible for interruption of treatment, yielding a PSA response rate of 95%. The study continued for 6 years with a mean follow-up of 3.7 years (median follow-up, 4.2 years). The time off treatment averaged 53% of the total cycle time but, in absolute terms, decreased with each succeeding cycle, ranging from 63.7 weeks in Cycle 1 to 25.6 weeks in Cycle 5. Prostate volume was reduced by 40% in Cycle 1 and by 34% in Cycle 2, and there were no decreases in Cycle 3 or Cycle 4. At the end of the trial, 38.5% of patients still were receiving treatment, 23.9% of patients had failed, and 15.6% of patients had died. Only 2% of deaths were cancer-specific.
Biochemical recurrence after irradiation for localized prostate cancer was amenable to cyclic androgen withdrawal therapy and showed a high response rate. Despite progressively shorter treatment cycles, the off-treatment interval remained appreciable, ranging from 65% in Cycle 1 to 46% in Cycle 5. Cancer 2006. © 2006 American Cancer Society.
Intermittent androgen suppression attempts to bring about long-term control of prostate cancer while minimizing the psychological and biologic side effects associated with permanent castration. The idea of producing more than 1 regression of androgen-dependent malignancy arose from the observation that the involution of prostate brought on by castration is an active process that involves the rapid elimination of a large number of epithelial cells.1–4 It was postulated first that the replacement of androgens, even in small amounts, would have a conditioning effect on surviving cells, allowing them to conserve or regain desirable traits of differentiation,5, 6 including the potential to undergo apoptosis.7 Such thinking began to shape the concept of cyclic versus continuous administration of hormones, especially the possibility that hormonal resistance may be delayed or avoided by interspersing treatment and rest periods.8 The underlying assumption was that the maintenance of apoptotic potential, otherwise defined as androgen dependence, by successive rounds of androgen withdrawal and replacement would result in multiple tumor regressions and would forestall tumor progression.9
When this regimen was tested in the Shionogi androgen-dependent mouse mammary carcinoma, an animal model that mimics certain features of prostate cancer,10–12 it was demonstrated that multiple regressions of a tumor cell population were possible if androgens were withdrawn and replaced successively.9 In addition, the massive increase in the proportion of androgen-independent stem cells that characterizes recurrent androgen-independent disease12 was delayed.13
Promising results from early feasibility studies9, 14, 15 stimulated the initiation of a number of Phase II clinical trials,16–29 including the current prospective Phase II study, which was conducted in 4 Canadian centers between 1995 and 2001. The clinical trial focused on men in biochemical recurrence after they had received radiotherapy for locally advanced prostate cancer and has been used as a guideline for several subsequent large, Phase III, randomized studies.30–35
MATERIALS AND METHODS
Patients with histologically confirmed adenocarcinoma who had rising serum PSA levels after they received external beam irradiation of the prostate were considered eligible for the study. The inclusion criteria were as follows: clinical Stage T1b/T1c, T2, or T3 disease at initial diagnosis prior to radiotherapy; performance status of 0 or 1; serum PSA before androgen suppression >6.0 μg/L; pretreatment serum testosterone within normal range (6.3–27 nmol/L); and no previous hormonal manipulations with the exception of neoadjuvant androgen-suppression therapy ≤3 months' duration. Patients must have completed irradiation at least 6 months prior. Patients were excluded from the study if they displayed the following characteristics: abnormal bone scan suggestive of metastatic osseous disease; previous hormonal manipulation, including any medication with antiandrogenic activity (e.g., antiandrogens, estrogens, ketoconazole, spironolactone, cimetidine, finasteride); any systemic chemotherapy or curative radiotherapy within 6 months; hepatic dysfunction (bilirubin >35 mmol/L and/or aspartate aminotransferase or lactate dehydrogenase >3 times the upper limit of normal range); renal dysfunction (blood urea nitrogen >15 mmol/L and/or creatinine >300 mmol/L); history or presence of other malignancy within the last 5 years (except treated squamous/basal cell carcinoma of the skin or superficial, low-grade bladder carcinoma); and known hypersensitivity to cyproterone acetate or leuprolide acetate.
Requisite evaluation consisted of a history, physical examination, and digital rectal examination; a transrectal sonogram of the prostate with sextant biopsies; a computed tomographic scan of the abdomen and pelvis to exclude regional lymph node involvement; a bone scan; and X-rays of the chest, pelvis, and other skeletal sites, as required. Laboratory tests included a complete blood count, bilirubin, alkaline phosphatase, lactate dehydrogenase, aspartate aminotransferase, urinalysis, blood urea nitrogen, creatinine, serum PSA, and total serum testosterone. The volume of the prostate was measured by using the formula for an ellipsoid volume calculation, π/6 (transverse diameter × anteroposterior diameter × cephalocaudal diameter).36
This was a prospective, noncomparative, open-label, multicenter study. Patients were followed on-study until the development of androgen-independent disease (time to progression), as defined below. Survival data were recorded until death. Side effects and quality of life also were evaluated (to be published). The treatment protocol is outlined in Figure 1. Cyproterone acetate was administered orally at a dose of 100 mg twice daily for 4 weeks to block flare. The luteinizing hormone-releasing hormone agonist, leuprolide acetate, was then given intramuscularly at a dose of 7.5 mg and was repeated every 4 weeks up to a total of 8 doses. After the initial injection of luteinizing hormone-releasing hormone agonist, cyproterone acetate was continued at the starting dose of 100 mg twice daily for another 2 weeks, when the dose was decreased to 50 mg twice daily to reduce the incidence of hot flashes. The overall duration of treatment was 36 weeks. If the serum PSA level was <4.0 μg/L at both 24 weeks and 32 weeks with a constant or decreasing slope, then therapy was stopped at 36 weeks. If the serum PSA level was not <4.0 μg/L at both 24 weeks and 32 weeks, then the patient was not eligible for interruption of therapy and was taken off-study. Serum PSA and testosterone levels were monitored every 4 weeks, and therapy was resumed when the serum PSA level was ≥10 μg/L. Additional sonographic estimates of prostatic volume were obtained at the start and end of each treatment period. Loss of androgen-dependence (progression) was defined as 3 sequential increases of serum PSA >4.0 μg/L despite castrate levels of serum testosterone.
This study was approved for implementation by the Health Protection Branch of the Ministry of Health of Canada. Submissions were made to the Clinical Research Ethics Review Board of each participating center to obtain the necessary authorization to begin enrolment of patients. All patients provided written informed consent in accordance with institutional guidelines.
In total, 109 patients were registered for entry into the trial. Four patients withdrew from the study before completing the initial 36 weeks of androgen suppression for reasons unrelated to treatment, and 2 patients did not meet the criteria for interruption of therapy at 36 weeks (Fig. 1). The resultant response rate was in excess of 95%. The characteristics of 109 patients are given in Table 1. The mean age at baseline was 73.3 years; the mean pretreatment PSA level was 21.2 μg/L; the mean pretreatment testosterone level was 13.2 nmol/L; the mean time from irradiation to the start of treatment was 4.6 years; the percentage of patients with positive biopsy results was 83.7%; the percentage of patients who had negative biopsy results was 16.3%; and patients primarily were Caucasian (96.1%). The mean follow-up was 3.7 years (median. 4.2 years).
|Characteristic||Mean ± SE||Median||Range|
|Age at baseline, y||73.3 ± 0.7||73.2||45.0–89.1|
|Pretreatment serum PSA, μg/L||21.2 ± 1.6||15.1||5.1–229|
|Pretreatment serum testosteron, nmol/L||13.2 ± 0.4||13.0||0.4–23.0|
|Time from irradiation to start of treatment, y||4.6 ± 0.3||4.1||0.8–14.7|
|Biopsy positive at registration, %||83.7||—||—|
|Biopsy negative at registration, %||16.3||—||—|
|Race, %||Caucasian, 96.3||—||—|
|Follow-up, y||3.7 ± 0.2||4.2||0.05–6.3|
The total number of cycles initiated was 277; of these, 232 cycles were completed, and an additional 45 cycles were initiated but not completed. The mean number of completed cycles per patient was 2.3 ± 0.1 cycles (median, 3.0 cycles; range, 1–5 cycles). The total number of treatment periods available for analysis was 495 divided between 277 on-treatment periods and 218 off-treatment periods.
The distribution of clinical stages (TNM) at the time of initial diagnosis prior to radiotherapy in the group of 103 evaluable patients was as follows: T1b, n = 6 patients (5.8%); T2a, n = 16 patients (15.5%); T2b, n = 37 patients (35.9%); T3a, n = 41 patients (39.8%); unknown, n = 3 patients (2.9%). Thus, the majority of patients (i.e., 76%) were diagnosed with Stage T2b and T3 disease.
The Gleason scores obtained at the time of initial diagnosis before radiotherapy were as follows: Gleason scores of 2, 3 and 4, n = 19 patients (18.5%); Gleason scores of 5 and 6, n = 35 patients (34.0%); Gleason score of 7, n = 29 patients (28.2%); Gleason scores of 8 and 9, n = 19 patients (18.5%); and unknown Gleason score, n = 1 patient (1.0%). The proportion of patients who had Gleason scores ≥7 was 47.6%, and the proportion of patients who had Gleason scores ≤6 was 52.4%.
Mean Time On and Off Treatment by Cycle
Table 2 shows that the mean time on treatment in Cycles 1 and 2 was approximately 36 weeks, in agreement with the protocol guidelines (Fig. 1); this on-treatment time showed an apparent decrease in Cycles 3, 4, and 5, because some patients in those groups progressed or reached the end of the study before their last cycle was completed. The time off treatment, which averaged 63.7 weeks (median, 53.7 weeks; range, 4–262 weeks) in Cycle 1, was reduced significantly by a relatively constant 22% to 27% in Cycles 2, 3, and 4 to 46.2 weeks (median, 37.7 weeks; range, 13–130 weeks; t test; P = .0028), 35.7 weeks (median, 31.7 weeks; range, 8–78 weeks; t test; P < .0001), and 27.2 weeks (median, 27.7 weeks; range, 11–59 weeks; t test; P < .0001), respectively. The proportion of the total length of each cycle spent in the off-treatment phase averaged 53% and varied from 64.9% in Cycle 1 to 46.4% in Cycle 5. Shortening of the off-treatment period was observed both in patients who progressed on treatment and in those who did not (Table 3), consistent with the conclusion that the decrease in length of the off-treatment period of successive cycles is not linked specifically to androgen independence.
|Cycle||Time on treatment (weeks)||Time off treatment (weeks)|
|Mean ± SE||Percent||Median||No. of patients||Mean ± SE||Percent||Median||No. of patients|
|1||34.5 ± 0.6||35.1||36.0||103||63.7 ± 4.4||64.9||53.7||93|
|2||34.9 ± 0.8||43.0||36.0||86||46.2 ± 3.3||57.0||37.7||71|
|3||31.6 ± 1.5||46.9||35.4||55||35.7 ± 2.8||53.1||31.7||37|
|4||33.4 ± 2.5||55.1||35.6||26||27.2 ± 3.2||44.9||27.7||15|
|5||29.6 ± 3.8||53.6||35.1||7||25.6 ± 10.5||46.4||25.6||2|
|Cycle||Time off treatment (weeks)|
|Mean± SE||No. of patients||Mean ± SE||No. of patients|
|1||68.3 ± 5.2||68||51.3 ± 7.4||25|
|2||50.8 ± 3.9||55||30.5 ± 3.8||16|
|3||38.0 ± 3.1||31||23.8 ± 2.0||6|
|4||27.7 ± 3.3||14||19.7 ± 0.0||15.6 ± 10.5|
|5||25.6 ± 10.5||25.6 ± 10.5||—||—|
Mean Time Off Treatment versus Biopsy Status
In general, the biopsy status had little effect on outcomes, except perhaps in the later cycles, when the mean time off treatment in the negative biopsy group was 19% longer in Cycle 2 and 75% longer in Cycle 3 than in the corresponding positive biopsy groups, but the study lacked sufficient power to lend confidence to these observations. When the times off treatment were averaged for all cycles, the time off treatment for the negative biopsy group was 52.9 ± 4.3 weeks (mean ± standard error [SE]; n = 33 cycles; range 14.3–130 weeks), slightly greater than that for the positive biopsy group (50.1 ± 2.8 weeks [mean ± SE]; n = 176 cycles; range, 4.0–261.7 weeks), but the difference was not statistically significant (t test; P = .0764).
Mean Time Off Treatment versus Gleason Score
The Gleason score had no statistically significant effect on cycle times.
Changes in Prostate Volume by Cycle
Pronounced changes in the volume of the prostate were observed in Cycles 1 and 2, as shown in Figure 2. In Cycle 1, from a mean volume of 24.7 cm3 at baseline, the mean volume decreased to 14.7 cm3 over the 36-week interval of androgen withdrawal, representing a reduction of 40%. The mean volume of the prostate recovered to 20.9 cm3 and then underwent a 34% decrease in volume to 13.9 cm3. The volume then recovered to a volume of 18.6 cm3 at the start of Cycle 3 and then failed to undergo the reduction that was observed previously. No effects of androgen withdrawal therapy were observed in Cycle 4. Decreases in serum PSA continued to be observed in Cycles 3 and 4 (data not shown), inferring that intermittent androgen suppression selects for patients with androgen-sensitive disease37 but not for patients with androgen-dependent disease.
Reasons for Ending Participation
At the termination of this trial in 2001, 42 patients still were undergoing active treatment on study (Table 4). However, it should be noted that there were variations in the numbers of patients still being followed at the participating institutions (22 of 31 patients, 15 of 33 patients, 1 of 3 patients, and 4 of 26 patients at each of the 4 sites, respectively). The number of patients who failed treatment by showing signs of early progression to androgen independence was 26, and there were 17 deaths from all causes. Two patients died from prostate cancer, 13 patients died from nonprostate cancer causes, and 2 patients died from unknown causes. The number of patients who moved into successive cycles is shown in Table 5 with information regarding the basis for failing to complete a cycle. The majority of treatment failures occurred in Cycles 2 and 3, whereas attrition for other reasons was observed mostly in Cycles 1 and 2. The mean time to progression was 165.6 ± 15.4 weeks (median, 175.1 weeks; range, 35.9–294.1 weeks). Morbidity and mortality statistics will be discussed in more detail in a subsequent report along with quality-of-life results.
|Reason for ending participation||No. of patients||Percent|
|End of study||42||38.5|
|Lost to follow-up||5||4.6|
|Cycle||Started cycle||Completed cycle||Progression||Attrition for other reasons||In off-treatment|
Although a number of Phase II studies of intermittent androgen suppression have been undertaken previously,16–31 the current clinical trial is unique, in that it was designed as a prospective study focusing on a well defined group of patients. The trial spanned an interval of 6 years with a mean follow-up of 3.7 years and a median follow-up of 4.2 years. The mean number of completed cycles per patient was 2.3, with a wide range from 1 to 5 cycles per patient. Information about the clinical stage of disease was based on the initial evaluation at the time of diagnosis and before radiotherapy. The majority of patients (76%) had been diagnosed with Stage T2b or T3 disease, as may be expected. The patient population was divided almost equally between men with Gleason scores ≥7 and men with Gleason scores ≤6.
The duration of the off-treatment period decreased with each succeeding cycle (Table 1). This was 63.7 weeks in Cycle 1, 46.2 weeks in Cycle 2, 35.7 weeks in Cycle 3, 27.2 weeks in Cycle 4, and 25.6 weeks in Cycle 5. The percentage reduction in the time off treatment between Cycles 1 and 2, Cycles 2 and 3, and Cycles 3 and 4 was constant, averaging 25%. The off-treatment times in Cycles 4 and 5 were not different. Despite the shortening of the overall cycle time with increasing numbers of cycles, the prostate cancer remained biochemically androgen sensitive, as indicated by continuing PSA responses. This is in keeping with the conclusion that intermittent androgen suppression selects for the patients who have the most androgen-sensitive tumors, giving rise to increasingly shorter treatment cycles. It is also important to consider the possibility that shortening the off-treatment period may be correlated to the development of early androgen independence. However, somewhat contrary to this notion is the finding that the mean number of cycles completed by patients who experienced treatment failure (2.77 cycles), compared with the mean number completed by patients who continued to be treated successfully (2.66 cycles), did not differ significantly (t test; P = .5917). Although shortening of the overall cycle time was characteristic of all patients, including those who progressed, the off-treatment periods in the latter group tended to be 25% to 40% shorter than the periods in the no-progression group (Table 3), accounting for the slight difference in the mean number of cycles completed. Irrespective of these considerations, the regimen was successful in making it possible for the patients to enjoy freedom from active treatment for as much as 65% of Cycle 1, 57% of Cycle 2, 53% of Cycle 3, 45% of Cycle 4, and 46% of Cycle 5. It is noteworthy that the side effects related to low testosterone levels induced by treatment may have persisted into the early off-treatment period and temporarily may have compromised the sense of well being of a patient until the function of the testis returned to normal.
In exploring the relation between time off treatment and biopsy status, the data were suggestive of a weak trend favoring the negative biopsy group in Cycles 2 and 3. Whether the apparent disparities in the times off treatment between the groups may have been related to differences in biologic characteristics (for example, the possibility of greater responsiveness of unirradiated, subclinical, micrometastatic disease compared with disease confined to the prostate) is unknown. When time off treatment was related to Gleason score, no clear relation was demonstrated except for a suggestive trend that a Gleason score ≤6 may be associated with a slightly longer time off treatment in the initial 2 cycles.
The observations that the prostate volume was reduced by 40% with treatment in Cycle 1 and 34% in Cycle 2 were somewhat unexpected but consistent with the possibility that the dose of radiation received as primary therapy was not totally ablative; no further decreases in volume were observed in Cycles 3 and 4. A possible explanation for these divergent responses is that, as the number of cycles increases, the prostate is rendered androgen-resistant with regard to apoptosis but remains biochemically androgen-dependent, as evidenced by the continuing responses of serum PSA. The dissociation of apoptosis and serum PSA has been observed experimentally in the LNCaP tumor model of prostate cancer.38 Alternatively, it is possible that the responses in later cycles are dominated by extracapsular and metastatic disease rather than by disease confined to the prostate, which conceivably may be more susceptible to elimination in the first 2 cycles.
At the end of 6 years of follow-up, 42 of 109 patients (38.5%) who initially registered for the clinical trial still were being treated actively (Table 4). There were 26 patients (23.9%) who had progressed and 17 patients (15.6%) who had died. The mean and median times to disease progression were 165.6 weeks and 175.1 weeks, respectively.
The data presented in this report highlight the clinical parameters associated with the treatment of men in biochemical recurrence after they receive radiotherapy for locally advanced prostate cancer. Our observations reinforce the impression that intermittent androgen suppression is applicable to this group of men and is associated with relatively few adverse events. Most of the deaths were attributed to causes other than prostate cancer, which may be expected in a cohort of men whose mean age at entry was 74 years. Biochemical parameters and observations on quality of life will be covered in later communications.
We thank Trial Coordinator Maureen Palmer; Trial Data Manager Leslie Godwin; Site Monitors Irene McNeil, Lorraine Miller, and Fern Lowe; Data Entry Assistants Kira Lynne and Corrina Ialungo; Information Technology Analysts Doug Hoffart and Jonas Abersbach; Alla Sekunova for typing and editing the manuscript. We also thank Dr. George Armitage, Dr. Andy Coldman, and Dr. Norman Phillips for helpful discussions and other contributions.
- 5Neoplastic development, vol 1. New York: Academic Press, 1969; 73..
- 8Basis for the use of drug and hormone combinations in the treatment of endocrine-related cancer. In: BruchovskyN, GoldieJH, editors. Drug and hormone resistance in neoplasia, vol II, clinical concepts. Boca Raton: CRC Press, 1983: 129–162., .
- 22Intermittent androgen suppression for the treatment of prostate cancer. In: BelldegrunA, KirbyRS, OliverRTD, editors. New perspectives in prostate cancer. Oxford: Isis Medical Media, 1998: 273–282., , .
- 31Second international workshop on intermittent androgen ablation therapy, meeting report. Eur Urol Today. 2000; 11: 4–5., .