Fax: (608) 265-8133
Results from a pilot Phase I trial of gefitinib combined with docetaxel and estramustine in patients with hormone-refractory prostate cancer
Article first published online: 27 MAR 2006
Copyright © 2006 American Cancer Society
Volume 106, Issue 9, pages 1917–1924, 1 May 2006
How to Cite
Wilding, G., Soulie, P., Trump, D., Das-Gupta, A. and Small, E. (2006), Results from a pilot Phase I trial of gefitinib combined with docetaxel and estramustine in patients with hormone-refractory prostate cancer. Cancer, 106: 1917–1924. doi: 10.1002/cncr.21831
- Issue published online: 18 APR 2006
- Article first published online: 27 MAR 2006
- Manuscript Accepted: 2 DEC 2005
- Manuscript Revised: 1 NOV 2005
- Manuscript Received: 9 JUN 2005
- epidermal growth factor receptor;
- tyrosine kinase inhibitor;
- hormone-refractory prostate cancer;
Gefitinib, which is an orally active epidermal growth factor receptor tyrosine kinase inhibitor, has demonstrated activity against hormone-refractory prostate cancer (HRPC) in preclinical studies. In this pilot Phase I trial, the authors evaluated the tolerability, efficacy, and pharmacokinetics of gefitinib combined with estramustine and docetaxel in patients with HRPC.
Patients received gefitinib (at a dose of 250 mg/day or 500 mg/day) on each day of a 21-day treatment cycle. Docetaxel (at a dose of 60 mg/m2) was administered on Day 1, and estramustine (at a dose of 280 mg) was administered 3 times daily on Days 1 through 5.
Fifteen patients were recruited at each gefitinib dose level. The most common adverse events observed were consistent with the known profiles of gefitinib, docetaxel, and estramustine. No dose-limiting toxicity was observed. Adverse events considered to be gefitinib related included diarrhea (n = 23 patients), rash (n = 8 patients), nausea (n = 7 patients), dry skin (n = 6 patients), and emesis (n = 6 patients). Overall, 9 of 22 evaluable patients (40.9%) experienced a pain response. and 9 of 30 patients (30%) had a prostate-specific antigen response. A partial objective tumor response was demonstrated in 1 of 13 evaluable patients (7.7%) in each dose group; the median time to progression for both doses combined was 185 days (range, 28–233 days). Data comparisons within individual patients suggested that docetaxel and estramustine had no effect on gefitinib steady-state levels. Gefitinib had no effect on docetaxel exposure at the 250-mg dose but decreased exposure at the 500-mg dose. However, gefitinib may increase exposure to estramustine, particularly at the 500 mg/day dose.
The results of the current study demonstrated that gefitinib combined with estramustine and docetaxel had acceptable and predictable tolerability. However, it is unclear whether gefitinib provides an additional clinical benefit over docetaxel and estramustine alone. Cancer 2006. © 2006 American Cancer Society.
Prostate cancer is the most common cancer in men in the U.S., with approximately 230,000 new diagnoses and 30,000 deaths reported in 2004.1 Although increasing numbers of patients are presenting with early-stage disease, accounting for approximately 86% of diagnoses in 2004, 6% of patients still present with advanced, metastatic disease, and these patients have a 5-year survival rate of 34%.1
Established treatment options for patients with localized disease include surgery and radiotherapy; however, from 22% to 48% of patients develop recurrent disease after they undergo radical prostatectomy.2 For patients with advanced disease, treatment includes medical or surgical androgen ablation. The majority of patients who initially respond to hormone manipulation eventually progress to hormone-refractory prostate cancer (HRPC),2 and treatment options for this stage of disease include second-line hormone manipulation, radiotherapy, and chemotherapy. Current chemotherapy regimens for patients with HRPC include estramustine, vinca alkaloids, mitoxantrone, and taxanes as single agents and in combination.3 Results of a recent randomized studies in patients with metastatic HRPC demonstrated improved overall survival of patients who received estramustine plus docetaxel or docetaxel alone compared with patients who received mitoxantrone plus prednisone therapy. Moreover, the time to progression (TTP), reductions in prostate-specific antigen (PSA) levels, and objective tumor response rates were greater in patients who received the docetaxel-containing regimens.4, 5
Progression to HRPC is a multistep process and often involves mutation of tumor suppressor genes or overexpression of oncogenes or growth factors and growth factor receptors.3, 6 One such growth factor receptor is the epidermal growth factor receptor (EGFR), which often is expressed or highly expressed in prostate cancer7 and correlates with recurrence after radical prostatectomy and disease progression to HRPC.8
Gefitinib (IRESSA™; AstraZeneca, Macclesfield, U.K.), which is an orally active EGFR tyrosine kinase inhibitor, has shown activity against prostate cancer cell lines and xenografts.9–11 In Phase I trials, 252 patients with a range of solid tumors received oral gefitinib, including 28 patients with HRPC.12–15 One patient with HRPC had a reduction in measurable disease in a lymph node metastasis, palliation of disease-related pain, and a reduction in PSA.12 The current pilot trial was designed to investigate the tolerability, efficacy (objective tumor response, palliative pain, and PSA response), and pharmacokinetics of gefitinib in combination with docetaxel and estramustine in patients with HRPC.
MATERIALS AND METHODS
Patients were recruited into this study over 9 months from 4 centers in the U.S. and France. For inclusion in the study, patients were required to have a diagnosis of prostate cancer (histologic and/or cytologic confirmation was desirable but not essential if the patient had a markedly elevated PSA level and metastatic bone lesions) that was not curable by surgery or radiotherapy and that had progressed (determined by radiologically imaged metastatic disease, PSA, or cancer-related pain) after the patient underwent surgical or medical castration (serum testosterone <50 ng/dL). Patients who received peripheral antiandrogens (bicalutamide, nilutamide, ketoconazole, or flutamide) must have discontinued these agents and must have been observed for a withdrawal response before study entry (≥8 weeks' observation for bicalutamide, nilutamide, and ketoconazole; ≥4 weeks' observation for flutamide).
Patients had to be age 18 years or older with a World Health Organization (WHO) performance status of 0 or 2 and a life expectancy of ≥12 weeks. They had to provide written, informed consent to participate in the study, which was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.
Patients were excluded from the study for any of the following reasons: neutrophils <1.5 × 109/L or platelets <100 × 109/L; serum bilirubin level greater than the upper limit of reference range (ULRR); alanine aminotransferase or aspartate aminotransferase levels >1.5 times the ULRR if accompanied by a serum alkaline phosphatase level <2.5 times the ULRR; serum creatinine >1.5 times the ULRR; any evidence of severe or uncontrolled systemic diseases, including hepatitis B and human immunodeficiency virus infection; radiotherapy completed within 4 weeks of starting treatment or strontium-89 therapy within 8 weeks of starting treatment; any standard chemotherapy, herbal preparations, or otherwise prohibited medications were to be completed 30 days prior to administration of study treatment; incomplete recovery from any previous cancer treatment or other surgery; active spinal cord compression; evidence of any other significant clinical disorder or laboratory findings that made it undesirable for the patient to participate in the study; a history of another solid tumor within the past 5 years that would confound the diagnosis of metastatic prostate cancer; contraindications to the use of estramustine, such as a history of thrombotic disease; previous treatment with anticoagulants for thrombotic disease; a PR interval of >217 msec.
This Phase I, multicenter, noncomparative, dose-escalation study in patients with HRPC was designed to evaluate the safety of 2 doses of gefitinib (250 mg/day and 500 mg/day) in combination with docetaxel and estramustine. Secondary objectives included evaluations of the palliative response rate in patients with cancer-related pain at study entry, PSA response rate, TTP, objective tumor response rates in patients with evaluable/measurable disease, and an assessment of whether steady-state exposure of gefitinib is altered by the addition of docetaxel and estramustine or whether exposure to docetaxel or estramustine is altered when given with gefitinib.
Treatment was given in 21-day cycles (Fig. 1). Estramustine (at a dose of 280 mg) was administered orally 3 times daily (total of 840 mg/day) on Days 1 through 5 of each treatment cycle. Prophylactic use of antiemetics was encouraged during treatment with estramustine. Docetaxel was administered at a dose of 60 mg/m2 as a 1-hour infusion on Day 2 of each treatment cycle. To prevent hypersensitivity reactions, patients received an antihistamine (diphenhydramine at a dose of 50 mg) and an H2 blocker (cimetidine at a dose of 300 mg) before the infusion of docetaxel. Dexamethasone at a dose of 8 mg every 12 hours also was given for a minimum of 2 days before, on, or after the day of the infusion. Gefitinib was administered once daily as an oral dose of 250 mg/day or 500 mg/day beginning on Day 3 of the first cycle, when 2 doses were given to ensure that steady-state levels of gefitinib were reached rapidly.
Gefitinib dose escalation
Initially, 6 patients were assigned to receive a dose of 250 mg/day of gefitinib in combination with estramustine and docetaxel, and treatment continued for 6 cycles or for as long as the patient continued to demonstrate benefit. The first full safety evaluation was performed when 6 patients had completed 21 days of treatment with docetaxel/estramustine and gefitinib. If no drug-related dose-limiting toxicity (DLT) had been reported by the time three patients had completed the first cycle of 250 mg/day, then the 500 mg/day dose was opened for recruitment. DLT was defined as Grade 3 or 4 thrombocytopenia and/or neutropenia for >7 days or neutropenia of any duration with fever; Grade 3 or 4 skin toxicity; Grade 3 or 4 gastrointestinal toxicity (diarrhea, nausea, and/or emesis persisting at the same or higher grade for >4 days despite antidiarrheal or antiemetic therapy); Grade 3 or 4 cardiac toxicity, including prolongation of the PR interval >217 msec; Grade 3 or 4 central nervous system, peripheral nervous system, lung, renal, or hepatic toxicity; and any significant symptomatic ocular toxicity consistent with preclinical corneal changes.
If 2 or 3 of the initial 6 patients had a drug-related DLT, then the cohort was expanded to obtain 12 evaluable patients; if ≥4 of these 12 patients had a drug-related DLT, then this dose was not considered safe in combination, and no dose escalation occurred. If ≥4 of the initial 6 patients had a drug-related DLT, then no dose escalation occurred. Once the highest safe combination of gefitinib in combination with docetaxel and estramustine had been determined, ≤20 patients were treated at that dose.
Gefitinib treatment could be interrupted for ≤14 days if Grade 3 or 4 nonhematologic toxicity or unacceptable toxicity occurred. If the adverse event (AE) decreased in severity to Grade 1, then gefitinib treatment could be resumed at the original dose. Dose reductions for gefitinib were permitted if the AE decreased in severity to Grade 2; 500 mg/day could be reduced to 250 mg/day, and 250 mg/day could be reduced to 100 mg/day. If toxicity occurred at the same grade on rechallenge, then a second dose interruption was allowed.
In the absence of Grade ≥3 toxicities after the first course of docetaxel, the dose could be escalated to 70 mg/m2. One dose reduction to 45 mg/m2 was allowed, and docetaxel treatment could be interrupted for ≤14 days to allow for recovery from toxicity. No dose changes were allowed for estramustine, although any delays in the administration of docetaxel would delay administration of estramustine.
Toxicity was assessed according to the National Cancer Institute Common Toxicity Criteria (version 2.0). Patients were monitored throughout the study and for 30 days after treatment cessation.
The area under the concentration-time curve during the 24-hour dosing interval at steady state (AUC24ss) was determined for gefitinib on Day 21 (gefitinib alone) and Day 23 (gefitinib in combination with chemotherapy). Venous blood samples (4.5 mL) were taken prior to gefitinib and at 3 hours, 7 hours, and 24 hours postgefitinib. The AUC from time 0 to time t (AUC0-t) was determined for docetaxel and for estramustine, which is the main metabolite of estramustine, on Day 2 (without gefitinib) and Day 23 (with gefitinib). For estramustine analysis, samples were taken predose and 2 hours, 4 hours, and 6 hours postestramustine dosing. For docetaxel analysis, samples were taken 1 hour, 2 hours, 3 hours, 4 hours, 7 hours, and 24 hours after the start of infusion of docetaxel. No formal statistical analyses were performed on pharmacokinetic data.
Pain was assessed by using the McGill-Melzack Pain Questionnaire.16 Questionnaires were completed at baseline, on Day 1 of each cycle, and at study completion/discontinuation. Patients were asked to rate their average pain level during the previous 24 hours on a scale from 0 (no pain) to 5 (worst pain). A response was defined as a 2-point reduction on 2 consecutive evaluations ≥3 weeks apart with no increase in analgesic score.
PSA response was determined from serum analysis. PSA was measured at baseline, on Day 1 of each cycle, and at study completion/discontinuation. A PSA response was defined as a reduction ≥50% in PSA compared with study entry that was confirmed ≥4 weeks later17 in patients who had no clinical or radiologic evidence of disease progression between assessments.
Objective tumor response in bidimensionally measurable disease (complete response, partial response, stable disease, or disease progression) was evaluated using Southwest Oncology Group-modified WHO objective tumor response criteria at baseline; at the beginning of Cycles 3, 6, 9, and 12; and at study completion/discontinuation. TTP was defined as the time from study entry to the earlier of the following: an increase in the pain intensity scale of ≥1 point compared with the nadir of the McGill-Melzack Pain Questionnaire; an increase in analgesic score of >25% compared with baseline on 2 consecutive visits; unequivocal evidence of new lesions or radiologic progression; or a requirement for radiotherapy. PSA was not used to determine TTP.
Fifteen patients were recruited at each dose level, and their characteristics at study entry are presented in Table 1. Patients in each group were well matched at study entry for previous cancer treatment and performance status, with the majority having a performance status of 1.
|Characteristic||No. of Patients (%)|
|Gefitinib, 250 mg/Day||Gefitinib, 500 mg/Day|
|No. of patients||15||15|
|Mean age in y (range)||63.6 (42–78)||62.7 (51–77)|
|WHO performance status|
|0||5 (33.3)||5 (33.3)|
|1||9 (60.0)||9 (60.0)|
|2||1 (6.7)||1 (6.7)|
|M0||5 (33.3)||9 (60.0)|
|M1||10 (66.7)||6 (40.0)|
|Prior cancer treatment|
|Chemotherapy||10 (66.7)||5 (33.3)|
|Brachytherapy||0 (0.0)||2 (13.3)|
|Radical prostatectomy||3 (20.0)||5 (33.3)|
|Radiotherapy||8 (53.3)||11 (73.3)|
|Medical castration||12 (80.0)||12 (80.0)|
|Surgical castration||3 (20.0)||3 (20.0)|
All patients were included in the safety population. The most common AEs that occurred in >10% of patients are presented in Table 2 and were consistent with the known profiles of gefitinib, estramustine, and docetaxel and with the patients' underlying condition. No patients experienced an acute (within 21 days) DLT that we believed was related to gefitinib. Common AEs that we believed were related to gefitinib included diarrhea (n = 23 patients), rash (n = 8 patients), nausea (n = 7 patients), dry skin (n = 6 patients), and emesis (n = 6 patients). The most common AEs that we believed were related to docetaxel and/or estramustine included asthenia (n = 19 patients), diarrhea (n = 17 patients), nausea (n = 14 patients), emesis (n = 12 patients), taste perversion (n = 12 patients), anorexia (n = 11 patients), peripheral edema (n = 11 patients), and alopecia (n = 6 patients). Hematologic AEs that we believed were related to docetaxel and/or estramustine included anemia (n = 10 patients), neutropenia (n = 9 patients), and leukopenia (n = 7 patients). There were no hematologic AEs that we believed were related to gefitinib.
|Event||No. of Patients*|
|Gefitinib, 250 mg/Day (n = 15)||Gefitinib, 500 mg/Day (n = 15)|
|Grade 1/2||Grade 3/4||Grade 1/2||Grade 3/4|
|Urinary tract infection||2||0||2||0|
No major differences were apparent between the dose groups with respect to Grade 3 or 4 AEs that we considered gefitinib related. Gefitinib-related Grade 3 AEs at the 250 mg/day dose were diarrhea (n = 3 patients), asthenia (n = 1 patient), dehydration (n = 1 patient), nausea (n = 1), thrombocytopenia (n = 1 patient), and emesis (n = 1 patient); at the 500 mg/day dose, the gefitinib-related Grade 3 AEs were diarrhea (n = 1 patient), asthenia (n = 1 patient), and hypocalcemia (n = 1 patient). No Grade 4 AEs that we considered gefitinib related occurred at the 250 mg/day dose. A Grade 4 myocardial infarction occurred at the 500 mg/day dose and was considered to be related to gefitinib and estramustine/docetaxel.
Neutropenia was the most frequently reported Grade 3 or 4 AE that we considered to be estramustine/docetaxel related. The estramustine/docetaxel-related Grade 3 AEs that were recorded in patients who received 250 mg/day gefitinib were neutropenia (n = 4 patients), leukopenia (n = 4 patients), asthenia (n = 2 patients), diarrhea (n = 2 patients), and thrombophlebitis, dehydration, nausea, thrombocytopenia, and emesis (n = 1 patient each). In the 500 mg/day group, the estramustine/docetaxel-related Grade 3 AEs were neutropenia (n = 3 patients) and leukopenia, asthenia, diarrhea, coagulation disorder, febrile neutropenia, and generalized edema (n = 1 patient each). Estramustine/docetaxel-related Grade 4 AEs included neutropenia (n = 3 patients in the 250 mg/day group and n = 2 patients in the 500 mg/day group), leukopenia (n = 1 patient in the 250 mg/day group), and myocardial infarction (n = 1 in the 500 mg/day group, as discussed earlier).
Ten patients had a docetaxel dose escalation from 60 mg/m2 to 70 mg/m2 for the second cycle of chemotherapy on Day 23. Dose interruption or delay as a result of toxicity occurred in 1 patient (6.7%) in the gefitinib 250 mg/day group and in 5 patients (33.3%) in the gefitinib 500 mg/day group. All patients resumed therapy within 15 days. Treatment was discontinued in 8 of 30 patients (3 patients in the 250 mg/day group and 5 patients in the 500 mg/day group). Two of those patients discontinued treatment as a result of AEs that were considered to be related solely to gefitinib (1 patient in the 250 mg/day group for Grade 3 diarrhea; 1 patient in the 500 mg/day group for Grade 3 acne), and 1 patient in the 500 mg/day group discontinued treatment as a result of neuropathy (Grade 2) that was considered to be estramustine/docetaxel related. The remaining 5 patients discontinued treatment as a result of AEs that were not considered to be treatment related or that were related to both gefitinib and docetaxel/estramustine.
At the 250 mg/day dose, 7 patients completed 6 cycles of therapy, and 1 patient completed 6 cycles after a dose reduction to 100 mg/day (Treatment Cycle 2). At the 500 mg/day dose, 4 patients completed 6 cycles of therapy, and 1 patient completed 6 cycles after a dose reduction to 250 mg/day (Treatment Cycle 6).
For both dose levels of gefitinib, the geometric mean (gmean) AUC24ss was similar when gefitinib was administered alone and in combination with docetaxel and estramustine (Table 3). For docetaxel, the gmean AUC0-t was similar when it was administered alone and in combination with 250 mg/day gefitinib; however, docetaxel in combination with 500 mg/day gefitinib resulted in a lowering of the gmean AUC0-t by 20% compared with docetaxel alone (Table 4). The gmean AUC0-t of estramustine, which is the main metabolite of estramustine, in combination with 250 mg/day gefitinib was similar to that estramustine alone; whereas, with 500 mg/day gefitinib, it was 47% higher than when it was administered alone (Table 4).
|AUC24ss, ng.h/mL||Gefitinib, 250 mg/Day (n = 8)||Gefitinib, 500 mg/Day (n = 7)|
|Gefitinib Alone (Day 21)||Gefitinib and Combination (Day 23)||Gefitinib Alone (Day 21)||Gefitinib and Combination (Day 23)|
|AUC24ss, ng.h/mL||Gefitinib, 250 mg/day (n = 6)||Gefitinib, 500 mg/day (n = 7)||Gefitinib, 250 mg/day (= 8)||Gefitinib, 500 mg/day (n = 7)|
|Docetaxel Alone (Day 2)||Docetaxel and Gefitinib (Day 23)||Docetaxel Alone (Day 2)||Docetaxel and Gefitinib (Day 23)||Estromustine Alone (Day 2)||Estromustine and Gefitinib (Day 23)||Estromustine Alone (Day 2)||Estromustine and Gefitinib (Day 23)|
Pain and PSA responses are presented in Table 5. Twenty-two patients were evaluable for pain response (13 patients at the 250 mg/day level and 9 patients at the 500 mg/day level). An improvement in pain was experienced by 5 of 13 evaluable patients (38.5%) and by 4 of 9 evaluable patients (44.4%) at the 250 mg/day and 500 mg/day dose levels, respectively.
|Assessment||No. of Patients (%)|
|Gefitinib, 250 mg/Day||Gefitinib, 500 mg/Day|
|Evaluable for PSA response||15 (100.0)||15 (100.0)|
|Responder||4 (26.7)||5 (33.3)|
|Stable||4 (26.7)||5 (33.3)|
|Nonresponder||4 (26.7)||0 (0.0)|
|Noncalculable*||3 (20.0)||5 (33.3)|
|Evaluable for pain response||13 (100.0)||9 (100.0)|
|Responder||5 (38.5)||4 (44.4)|
|Nonresponder||8 (61.5)||5 (55.6)|
All patients in the intent-to-treat population (n = 15 patients at both dose levels) were evaluable for PSA response. Response was similar between the 2 dose groups: Four of 15 patients (26.7%) in the 250 mg/day group and 5 of 15 patients (33.3%) in the 500 mg/day group had a PSA response.
Thirteen patients in each group were evaluable for objective tumor response. One patient in each treatment group had an objective tumor response (partial response), and there was a response rate of 7.7% for both dose groups. For the patients who demonstrated a partial response, the duration of response was 84 days and 91 days for the 250 mg/day and 500 mg/day gefitinib groups, respectively. The patient in the 250 mg/day group required a dose reduction from 250 mg/day to 100 mg/day at Cycle 2. Overall, 8 of 30 patients (26.7%) had disease progression. The Kaplan–Meier median TTP was estimated at 185 days (range [including censored progression times], 28–233 days).
The results from the current study demonstrate that oral gefitinib in combination with docetaxel and estramustine can be administered safely to patients with HRPC. The tolerability profile of gefitinib in combination was consistent with the known tolerability profile of each agent, such that common AEs included skin rash and diarrhea (gefitinib), bone marrow toxicity (docetaxel), and nausea and thrombotic events (estramustine); however, there was no evidence of increased or cumulative toxicity. Likewise, there was no DLT observed at either dose of gefitinib in patients who completed the first cycle of combined treatment.
In the current study, gefitinib in combination with docetaxel and estramustine demonstrated efficacy in terms of pain and PSA response to a similar extent in both dose groups. Approximately 40% of patients in this study who reported pain at baseline experienced palliation of their disease-related pain, and approximately 30% experienced a PSA response. In previous studies that evaluated combinations of docetaxel and estramustine as a treatment option in HRPC, pain and PSA reductions were observed in from 59% to 81% and 39% to 68% of patients, respectively,18–21 although some of those studies used different doses and schedules. There may have been somewhat lower than expected response with gefitinib in the current study, because patients who had received prior chemotherapy were permitted to enroll.
Steady-state levels of gefitinib at either dose level were not altered by the presence of docetaxel and estramustine. Although no marked interaction was observed between gefitinib and docetaxel, the gmean AUC0-t for docetaxel in combination with 500 mg/day gefitinib resulted in a lowering of the gmean AUC0-t by 20% compared with docetaxel alone. However, steady-state levels of gefitinib, particularly at the higher dose, increased exposure to estramustine. This may reflect an effect of gefitinib on the biliary clearance of estramustine, the primary route of elimination for both estramustine and estromustine.22 Recent data by Solit et al.23 indicate that the dose timing of gefitinib with taxanes may affect both the dose level and the efficacy of the combination.
The results of a second Phase I trial of gefitinib in combination with mitoxantrone and prednisone for the treatment of patients with HRPC also suggest that gefitinib in combination with standard treatment is feasible and has an acceptable tolerability profile. Although no objective tumor responses were observed, PSA responses of 33% to 60% and pain responses of 20% to 56% were demonstrated, and no unexpected or cumulative toxicities were recorded.24
In conclusion, the results of the current study demonstrate that the combination of gefitinib with docetaxel and estramustine has an acceptable and predictable tolerability profile. However, it is not clear from these data whether gefitinib provides additional clinical benefit over docetaxel and estramustine alone. In the absence of strong signals that suggest synergy, Phase III clinical studies of this combination cannot be recommended.
- 24ZD1839, an epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), is well tolerated in combination with mitoxantrone and prednisone in patients with hormone-refractory prostate cancer (HRPC) [abstract]. Ann Oncol. 2002; 13(Suppl 5): 91., , , et al.