3

SWOG 99-16: Randomized Phase III Trial of Docetaxel (D) / Estramustine (E) Versus Mitoxantrone(M) / Prednisone(p) in Men with Androgen-independent Prostate Cancer (AIPCA)

D. P. Petrylak, C. Tangen, M. Hussain, P. N. Lara, J. Jones, M. E. Talpin, P. Burch, G. Greene, E. Small, E. D. Crawford; New York Presbyterian Hospital, New York, NY; University of Washington, Seattle, WA; University of Michigan, Anne Arbor, MI; University of California Davis Cancer Center, Sacramento, CA; NASA Johnson Space Center, Houston, TX; Dana Farber Cancer Center, Boston, MA; Mayo Clinic, Rochester, MN; University of Arkansas, Little Rock, AR; University of California San Francisco, San Francisco, CA; University of Colorado, Denver, CO

METHODS: 770 men with progressive AIPCA were randomized to either Arm 1) Dexamethasone 60 mg premedication; D=60 mg/m² day(D) 2, E= 280 mg PO D1-5 Q 21 D or Arm 2) M 12 mg/m² + P 5 mg PO BID q 21 D. Pts were dose escalated to 70 mg/m² in Arm 1 and 14 mg/m2in Arm 2 if no Grade(G) 3 or 4 toxicities were observed in cycle 1. Assuming 310 eligible pts in each arm, the study had 0.80 power to detect a 33% difference in survival. All p-values reported are two-sided.

RESULTS: Pt Characteristics: Eligible pts Arm 1:334; Arm 2 :332. Pretreatment characteristics were equally balanced for age, race, performance status, PSA and symptoms.

RESPONSE AND SURVIVAL: The median survival of men treated on Arm 1 was 18 months and Arm 2 was 15 months (logrank p=.008) The hazard ratio was 0.77 (95% confidence interval 0.64, 0.94). Arm 1 also demonstrated a superior median time to progression (6 months) compared to Arm 2 (3 months) logrank p<0.0001. The measurable disease response rates in arm 1 and 2 were 17% and 10% (p=.30), respectively. Toxicity: G 3 or 4 toxicity was reported in 175 pts (54%) on Arm 1 and 109 pts (34%) on arm 2, due to higher rates of gastrointestinal (63pts vs. 21pts) and cardiovascular(44pts vs 20pts) events on arm 1. There was no significant difference in toxic deaths observed Arm 1 (n=7, 2%) vs. Arm 2(n=4, 1%).

CONCLUSIONS: The 23% improvement in survival in men treated with D + E supports its use as first line therapy for AIPCA. This is the first large randomized trial demonstrating a survival advantage in AIPCA

4

A Multicenter Phase III Comparison of Docetaxel (D) + Prednisone (P) and Mitoxantrone (MTZ) + P in Patients With Hormone-Refractory Prostate Cancer (HRPC)

Author(s): M. A. Eisenberger, R. De Wit, W. Berry, I. Bodrogi, A. Pluzanska, K. Chi, S. Oudard, T. Christine, N. James, I. Tannock; Sidney Kimmel Comprehensive Cancer Center, Baltimore, MD; Erasmus University, Rotterdam, Netherlands; US Oncology, Raleigh, NC; Orszagos Onkologiai Intezet, Budapest, Hungary; Onkologiczny Klinika Chemotherapii, Lodz, Poland; BC Cancer Agency, Vancouver, BC, Canada; Hopital Europeen Georges Pompidou, Paris, France; Institute Gustave Roussy, Villejuif, France; Queen Elizabeth Hospital, Edgbaston, United Kingdom; Princess Margaret Hospital, Toronto, ON, Canada

BACKGROUND: Docetaxel-based regimens have shown encouraging anti-tumor activity in phase II trials in HRPC. This study compared two schedules of D+P to MTZ+P, an accepted standard in HRPC.

METHOD: Patients (pts) with histologically-proven metastatic HRPC, testosterone < 50 ng/ml, adequate hematologic and organ function, clinical and/or biochemical (PSA) progression, and no antiandrogens within 6 wk, were stratified by pain and KPS and randomized to daily prednisone (5mg po bid) plus D (ARM A-75mg/m2 q3wk x 10 cycles; ARM B-30mg/m2/wk x 5 of 6 wk x 5 cycles) or MTZ (ARM C-12mg/m2 q3wk x 10 cycles). Primary endpoint was survival; secondary endpoints included PSA response (=50% decrease for at least 4 wk), pain response, and toxicity. A modified Bonferroni method was used to adjust for multiple comparisons.

RESULTS: 1,006 pts were randomized. Median follow up was 20.7 mo. Pts in all arms had similar baseline characteristics. Planned treatment was delivered to 98%, 96% and 99% pts in A, B, and C, respectively. Grade 3/4 toxicities (% pts A, B, C): overall (45.8%, 43.0%, 34.6%), including bone pain (7.8%, 7.3%, 9.9%), infection (5.7%, 5.5%, 4.2%), fatigue (4.5%, 5.5%, 5.1%), and diarrhea (2.1%, 4.8%, 1.2%). The most common laboratory grade 3/4 toxicity was neutropenia (32.0%, 1.5%, 21.7%).

CONCLUSIONS: The D q3wk +P schedule improved overall survival and increased pain response and PSA response vs MTZ+P. It was well tolerated, but with more grade 3/4 neutropenia than MTZ+P. This is the first phase III study to show a significant survival benefit in HRPC. (Supported by Aventis)

4503

The Impact of A Delay in Initiating Radiation Therapy on Prostate-specific Antigen Outcome for Patients with Clinically Localized Prostate Cancer

P. L. Nguyen, R. Whittington, S. Koo, D. Schultz, K. B. Cote, M. Loffredo, E. McMahon, A. A. Renshaw, J. E. Tomaszewski, A. V. D'Amico; Harvard Medical School, Boston, MA; University of Pennsylvania Radiation Oncology, Philadelphia, PA; Millersville University Dept of Mathematics, MIllersville, PA; Brigham & Womens/Dana-Farber Radiation Oncology, Boston, MA; Baptist Hospital of Miami Dept of Pathology, Miami, FL; University of Pennsylvania Dept of Pathology, Philadelphia, PA

BACKGROUND: While a delay in initiating external beam radiation therapy (RT) following diagnosis is known to decrease local control in patients with breast or head and neck cancer, its effect on PSA outcome in prostate cancer is unknown.

METHODS: We studied 460 patients who received 3D conformal RT without hormonal therapy for clinically localized prostate cancer from 1994-2001. The primary endpoint was PSA failure (ASTRO definition); estimates of PSA control were made using the Kaplan-Meier method. Delay was defined as the time between diagnosis and RT start date. Cox multivariable regression was used to determine the ability of treatment delay to predict time to PSA failure after adjusting for the distribution of baseline PSA, clinical T-category, Gleason score, and percent of cores positive for tumor.

RESULTS: Treatment delay independently predicted for time to PSA failure (p=0.013) after controlling for baseline PSA, clinical T-category, Gleason score, and percent of cores positive. Patients with high-risk disease (n=161) had 5-year estimates of PSA-failure free survival of 53% vs. 38% (P_log-rank=0.037) for those with delay less vs. greater than the median delay (2.5 months) respectively. Low-risk patients (n=299) had no detectable difference in PSA outcome (p=0.621) between those with a delay less vs. greater than the median delay (2.6 months).

CONCLUSIONS: A treatment delay of approximately 3 months adversely affected PSA outcome for patients with high-risk but not low-risk disease following RT alone. Whether this effect remains after the use of RT plus concomitant hormonal therapy in patients with high-risk disease requires further study.

4504

Does Post-operative Radiotherapy (P-RXT) After Radical Prostatectomy (Px) Improve Progression-free Survival (PFS) in pT3N0 Prostate Cancer (PC)? (EORTC 22911)

M. Bolla, H. Van Poppel, P. Van Cangh, K. Vekemans, P. Rigatti, T. De Reijke, A. Verbaeys, J.-F. Bosset, L. Collette, Radiotherapy and Genito-urinary Tract Cancer Groups Of The Eortc; CHR de Grenoble, Grenoble, France; UZ Gasthuisberg, Leuven, Belgium; Cliniques Universitaires St Luc, Brussels, Belgium; Virga Jesse Hospital, Hasselt, Belgium; Istituto Scientifico H.S. Raffaele, Milan, Italy; Academisch Medisch Centrum, Amsterdam, Netherlands; Universiteit Gent, Gent, Belgium; CHR de Besançon, Besançon, France; EORTC, Brussels, Belgium

BACKGROUND: In 1992, as radical prostatectomy was more frequently applied to clinical T1-2N0M0 prostate cancer, the EORTC has undertaken a randomized trial of immediate post-operative treatment versus wait-and-see policy, for patients with high risk factors of local relapse on pathological specimen. We present the first efficacy results of this study.

METHODS: Eligible patients were ≤75 years old, WHO performance status (PS) 0-1, had T0-3N0M0 PC preoperatively and ≥1 pathological risk factor of: capsule invasion, positive surgical margins, invasion of seminal vesicles. P-XRT was 60Gy conventional external radiation delivered over 6 weeks. The trial was powered to detect a hazard ratio (HR) of 0.77 with 80% power with two-sided α=0.05 with regard to clinical or biological progression-free survival.

RESULTS: 1005 pts were accrued by end 2001 (P-XRT: 503, Px: 502). Median age was 65.4 years, PS 0 in 93.8%, median pre- and post-operative PSA 12.3 and 0.2 ng/ml, respectively. All but 41pts on P-XRT were irradiated with median dose 60Gy (range: 50-74). The trial was reviewed by an IDMC in December 2003, after a median follow-up of 5 years and early disclosure of the trial results was recommended. Biochemical PFS (time to twice confirmed PSA increase over nadir or first clinical failure or death) at 5 years was 72.2% (CI: 67.7-76.8) on P-XRT and 51.8% (CI: 46.8-56.8), HR=0.52 (CI: 0.42-0.64), P<0.0001. Clinical PFS was improved from 74.8% to 83.3% at 5 years (HR=0.68, CI: 0.52-0.89, P=0.004). The incidence of loco-regional failure was significantly decreased (P<0.0001). Further follow-up is needed to assess the impact on distant metastases and overall survival. P-XRT is associated with an increased risk of immediate and late grade 1-2 side effects. Grade 3 side effect rates so far were less than 5% in both groups.

CONCLUSIONS: Post-operative radiotherapy results in improved biochemical and clinical PFS. This benefit is to be weighed against the treatment side effects. Further follow-up is needed to assess the impact on overall survival.

4507

Association Between Androgen Deprivation Therapy And Fracture Risk: A Population-based Cohort Study in Men with Non-metastatic Prostate Cancer

M. R. Smith, W. C. Lee, T. Krupsi, J. Brandman, Q. Wang, M. Botteman, C. Pashos, M. Litwin; Massachusetts General Hospital, Boston, MA; Abt Associates Inc, Bethesda, MD; UCLA, Los Angeles, CA; Novartis Pahrmaceutical Corporation, Florham Park, NJ

BACKGROUND: Androgen deprivation therapy with a gonadotropin-releasing hormone (GnRH) agonist decreases bone mineral density in men with prostate cancer. Small retrospective studies have suggested that GnRH agonists increase fracture risk. We conducted a population-based cohort study to better characterize the association between GnRH agonist treatment and clinical fractures.

METHODS: Using 1991-2001 medical claims data from a 5% national random sample of Medicare beneficiaries, we identified a study group (N=3887) of men with non-metastatic prostate cancer who initiated GnRH agonist treatment in 1992-94. A comparison group (N=7,774) of men with non-metastatic prostate cancer who did not receive GnRH agonist treatment during the study period was randomly matched on a 1:2 ratio based on age, race and Charlson co-morbidity index. We analyzed their inpatient, outpatient, and physician claims to identify clinical fractures between 1994-2001.

RESULTS: One third of the men who received GnRH agonist treatment experienced one or more fractures. When adjusted for survival, the fracture incidence rates during the final year of the analysis were 83% for the study group versus 56% for control group. Men in the study group were significantly more likely to develop fractures over the study period than men in control group (hazard ratio 1.4; 95% CI 1.16-1.70; P<0.001). Longer-term (>3 years) treatment with a GnRH agonist was associated with a marked increase in fracture risk compared to shorter term use (< 1year) (hazard ratio 1.5; 95% CI 1.17-1.93 ; P<0.001).

Conclusion: Non-metastatic prostate cancer patients receiving GnRH agonist treatment were 40% more likely to develop a clinical fracture over the study period relative to those not receiving this treatment. Long-term treatment is associated with higher fracture risk. These results highlight the importance of diagnosis and management of skeletal complications in this population.

4508

Effects of Atrasentan on Disease Progression and Biological Markers in Men with Metastatic Hormone-refractory Prostate Cancer: Phase 3 Study

M. Carducci, J. B. Nelson, F. Saad, C. C. Schulman, D. P. Dearnaley, D. J. Sleep, S. M. Hulting, J. D. Isaacson, A. R. Allen, P. Nisen; Sidney Kimmel Comprehensive Cancer Center, Baltimore, MD; University of Pittsburgh School of Medicine, Pittsburgh, PA; University of Montreal, Montreal, PQ, Canada; Erasme Hopital, Brussels, Belgium; Royal Marsden Hospital, London, United Kingdom; Abbott Laboratories, Abbott Park, IL

Background: In earlier trials, atrasentan, a potent, selective, oral endothelin A receptor antagonist, significantly delays time to disease progression (TTP) in a per-protocol group of hormone-refractory prostate cancer (HRPC) patients identified pre-blind break. We report the results from a phase 3 trial.

Methods: 809 subjects with metastatic HRPC enrolled into a phase 3, randomized, double-blind, placebo-controlled study of 10 mg atrasentan were analyzed (401 placebo, 408 atrasentan). Kaplan-Meier and Cox proportional hazard methodologies were used to compare TTP (a composite of radiographic and clinical measures) between treatments. Analysis of covariance compared change from baseline to final value in PSA and bone and total alkaline phosphatase (BAP and ALP). Data from a protocol-compliant subset identified prior to blind break were analyzed as well.

Results: Atrasentan exhibited a nonsignificant trend in delayed TTP (log-rank p=0.091; hazard ratio [HR] = 1.14, 95% CI = 0.98-1.34) and significantly delayed time to BAP progression (TTBAP, >50% increase from nadir) versus placebo (median: 254 vs 505 days; log-rank p < 0.001; HR = 1.78, 95% CI = 1.34-2.37). The atrasentan group showed smaller mean increases than the placebo group in PSA (p=0.025), BAP (p=0.001), and ALP (p < 0.001). In the protocol-compliant subset (329 placebo, 342 atrasentan), atrasentan significantly delayed TTP (log-rank p=0.007; HR = 1.26, 95% CI = 1.06-1.50) and TTBAP (log-rank p < 0.001) and resulted in smaller mean changes from baseline to final in the 3 biological markers. Atrasentan was well tolerated. The most common associated adverse events, peripheral edema, rhinitis, and headache, resulted in few discontinuations.

Conclusions: Atrasentan demonstrated a significant delay in TTP and attenuated increase in relevant bone and tumor markers in metastatic HRPC patients representative of the intended target population. The accumulating and consistent data for atrasentan demonstrates biologic and clinical activity of endothelin receptor blockade in metastatic HRPC.

4509

Multi-institutional Trial of The Epothilone B Analogue BMS-247550 with or without Estramustine Phosphate (EMP) in Patients with Progressive Castrate-metastatic Prostate Cancer (PCMPC): Updated Results

W. K. Kelly, M. D. Galsky, E. J. Small, W. Oh, I. Chen, D. Smith, L. Martone, T. Curley, A. Delacruz, H. I. Scher; Memorial Sloan Kettering Cancer Center, New York, NY; University of California, San Francisco, CA; Dana-Farber Cancer Institute, Boston, MA; M. D. Anderson Cancer Center, Houston, TX; University of Michigan, Ann Arbor, MI

BACKGROUND: BMS-247550 is a semi-synthetic analogue of epothilone B with preclinical activity in taxane-resistant cell lines. We have previously shown that BMS-247550 + EMP could be safely administered in PCMPC (Ann Oncol 14:1518, 2003). Subsequently, we initiated a multi-institutional phase II study to determine the efficacy and toxicity of BMS-247550 +/- EMP.

METHODS: Chemotherapy-naïve patients with PCMPC were eligible. Patients were randomized to BE: BMS-247550 (35 mg/m² IV over 3 hours every 3 weeks on day 2) + EMP (280 mg PO TID day 1-5) + Coumadin (2 mg daily) or B: BMS-247550 (35 mg/m² IV over 3 hours every 3 weeks).

RESULTS: Accrual has completed with 92 patients enrolled (BE: 45, B: 47). Patients have received a median of 5 cycles (range, 1-12) on BE and 4 cycles (range, 1-25) on B. Twenty-two patients continue on treatment. The major adverse events are detailed in the table. Neuropathy developed after a median of 3 cycles (range, 1-7). On BE, 22/32 (69%; 95% CI, 53-85%) have achieved a ≥ 50% decline in PSA and 8/18 (44%; 95% CI, 21-69%) have had a partial regression (PR) of measurable disease. On B, 18/32 (56%; 95% CI, 39-73%) have achieved a ≥ 50% decline in PSA and 6/26 (23%; 95% CI, 7-40%) have had a PR. At a median follow-up of 6.5 months (range, 0-23), 81 patients are alive and 11 have died.

CONCLUSIONS: BMS-247550, alone or in combination with EMP, has significant activity in PCMPC. Neuropathy is prominent, has proven manageable, yet requires further characterization. Phase III studies are needed to determine if BMS-247550 confers a survival benefit. Support: CTEP, Prostate Cancer Foundation, NIH N01-CM17105

4510

Epothilone B (Epo-B) Analogue BMS-247550 (NSC #710428) Administered Every 21 Days in Patients (pts) with Hormone Refractory Prostate Cancer (HRPC). A Southwest Oncology Group Study (S0111)

M. Hussain, J. Faulkner, U. Vaishampayan, P. Lara, D. Petrylak, D. Colevas, W. Sakr, E. D. Crawford; University of Michigan, Ann Arbor, MI; Southwest Oncology Group, Seattle, WA; Wayne State University, Detroit, MI; University of California, Sacramento, CA; Columbia Presbyterian Med, New York, NY; NCI, Rockville, MD; University of Colorado, Denver, CO

BACKGROUND: The epothilones are a new class of non-taxane tubulin polymerizing agents with activity both in taxane-sensitive and resistant tumor models. This Phase II trial with a 2-stage design evaluated the efficacy and toxicity of Epo-B in pts with metastatic HRPC.

METHODS: Eligible pts had chemotherapy-naive metastatic HRPC, a Zubrod performance status 0-2, and adequate organ function. All patients received Epo-B (40 mg /m²) over 3 hours every 3 weeks. Disease was assessed after every 2 cycles. The primary end point was rate of PSA response (greater or equal to 50% decrease below baseline with stable or better objective disease).

RESULTS: Two-stage accrual was completed in November, 2003. Of the 48 registered pts 41 were eligible with a median age of 74 years. To date complete data is available for the 28 pts registered to the first stage (22 analyzable, 5 ineligible, 1 did not receive protocol specified treatment). Grade (G) 3 and 4 toxicities occurred in 11 and 1(neutropenia) pts, respectively. The most frequent G3 toxicities were fatigue (3 pts) and sensory neuropathy (3 pts). There were 9 PSA responses (41%, 8 confirmed and 1 unconfirmed, 95% CI 21-64%). Ten of the 22 pts had measurable disease and 3/10 had an objective response (30%, 1 unconfirmed complete and 2 unconfirmed partial responses). The estimated median progression free survival was 8 months (95% CI 6-11 months) and the estimated 1-year survival was 75%. The median survival was not reached at this time. Final results on all patients will be presented.

CONCLUSIONS: Early results from this multicenter trial indicate that BMS-247550 has activity in patients with metastatic HRPC. Supported by CTEP.

4511

Development of Bone Metastases from Prostate Cancer: First Results of the MRC PR04 Trial (ISCRTN 61384873)

M. D. Mason, MRC PR04 Collaborators; University of Wales College of Medicine, Cardiff, United Kingdom

BACKGROUND: The skeleton is the most common site of metastases from prostate cancer (PCa). Bisphosphonates have been shown to slow development of metastases from breast cancer and myeloma and to modify morbidity from bone metastases (BM) in PCa.

METHODS: Phase III double-blind placebo-controlled randomised trial of oral bisphosphonate in men receiving standard treatment for stage T2-T4 PCa with no evidence of bone metastases and WHO performance status 0-2. The primary endpoint was time to development of symptomatic bone metastases or PCa death. Treatment consisted of either 4 tablets/day oral sodium clodronate (2,080mg Loron520) or 4 tablets/day of matching placebo control. Patients were encouraged to stay on trial medication for 5 years or until the primary trial endpoint had been reached. The trial was initiated in the pre-PSA era.

RESULTS: Patients: 508 patients were randomised (target 500) over 3.5 years (June 1994 - December 1997): 254 to active (A), 254 to control (C). Baseline characteristics were well balanced with stage T2 - 53%, T3 - 42% and T4 - 4%; WHO performance status 0 - 79%; median age 69 years, overall. The median follow-up time is 7 years for patients last alive. Medication & Toxicity: All patients have completed their allocated trial medication. The main reasons for stopping trial drug were: 5 years of trial drug (Active 38%, Control 46%), gastro-intestinal (GI) problems (Active 15%, Control 8%) and symptomatic bone progression (Active 9%, Control 9%). The proportion of patients still on trial medication at 3 years was 54% for Active and 69% for Control. 383 adverse events were reported, 202 in the Active group, 181 in Control. The most common reported AEs were GI problems (43% Active, 38% Control) and raised LDH (15% Active, 3% Control). Symptomatic bone metastases or prostate cancer deaths: With 131 events there was no evidence of a difference; a trend to better outcome for Control was observed: HR=1.29 (95%CI 0.92,1.82; p=0.13). Overall survival: there was no evidence of a difference with 170 deaths reported: HR=1.03 (95%CI 0.76,1.39; p=0.86). Combining the arms, overall survival at 5-yrs was 78%.

COMMENTS: These initial results suggest, at best, no improvement in terms of the primary outcome of symptomatic bone metastases or prostate cancer death for patients with locally advanced prostate cancer with the use oral sodium clodronate. This outcome is qualitatively different to the published results of the sister trial, MRC PR05, in hormone-sensitive metastatic disease and to studies with zoledronic acid in patients with hormone-refractory metastatic disease. The outcomes for all patients were much better than anticipated in the design of this trial, therefore the observed event rate was lower than expected. Further adjuvant studies with newer bisphosphonates are urgently required before their routine adoption as adjuvant treatment in PCa. Benefits from bisphosphonates in advanced disease cannot be assumed to apply in the adjuvant setting nor can a class effect across different tumour types be assumed.

4551

Is Prostate-specific Antigen a Surrogate for Survival in Advanced Prostate Cancer?

L. Collette, T. Burzykowski, K. Carroll, D. Newling, T. Morris, F. Schroder; EORTC Data Center, Brussels, Belgium; Limburgs Universitair Centrum, Diepenbeek, Belgium; AstraZeneca, Macclesfield, United Kingdom; Erasmus Medical Centrum, Rotterdam, Netherlands

Background: Surrogate endpoints are needed to shorten the duration of Phase III clinical trials in advanced prostate cancer. Prostate-specific antigen (PSA) is the most studied biomarker in prostate cancer. This study attempts to validate PSA endpoints as surrogates for overall survival (OS).

Methods: Individual data from 2161 patients with advanced prostate cancer treated in studies comparing bicalutamide ('Casodex'), either as monotherapy or in combination with an LHRHa, with castration were used in a meta-analytical approach to surrogate endpoint validation. PSA response, several definitions of time to PSA progression and longitudinal PSA measurements were considered.

Results: The analyses confirmed the known association between the PSA endpoints and OS at the individual patient level (biomarker association). However, when comparing patients treated with bicalutamide-based treatment or castration, the effect of hormonal intervention on the PSA endpoint did not predict the effect on OS with a high degree of precision. The association between intervention on any PSA endpoint and OS, measured by the determination coefficient R2 (ranging from 0.10-0.66 for PSA progression, to 0.69 for the whole PSA profile) was generally low.

Conclusions: It is a common misconception that a correlation at the individual level between PSA and OS is enough to demonstrate surrogacy. To demonstrate true surrogacy, a high correlation between the treatment effect on the surrogate and the treatment effect on the true endpoint needs to be established across groups of patients treated with two alternative interventions. The level of association observed in this study between the treatment effect on PSA endpoints and that observed on OS was in general low, showing that in Phase III clinical trials of hormonal treatments in advanced prostate cancer, treatment effects on OS cannot be predicted from observed treatment effects on PSA endpoints with a high degree of precision. This study indicates that PSA is unlikely to be a useful surrogate for OS in advanced prostate cancer. 'Casodex' is a trademark of the AstraZeneca group of companies

4552

Duration of Response to Androgen Deprivation Therapy and Survival After Subsequent Biochemical Relapse in Men Initially Treated with Radical Prostatectomy

F. J. Bianco, Z. A. Dotan, M. W. Kattan, P. A. Fearn, H. I. Scher, J. A. Eastham, P. T. Scardino; Memorial Sloan-Kettering Cancer Center, New York, NY

Background: Biochemical recurrence (BCR) is seen in 25-35% of patients after radical prostatectomy (RP). Without treatment the median time from BCR to metastases (Mets) is 7-8 years and from Mets to death is 5-6 years. Management of BCR varies, with some patients observed, some irradiated and some given androgen deprivation therapy (ADT). The duration of response to first line ADT, and of survival for those who progress with a rising PSA after ADT (CastratePSA), have not been well documented.

Methods: Our Prostate SPORE database traced all PSA measurements in 4037 patients treated by RP alone from 1983 to 2003. BCR was defined as 3 rises in PSA or secondary treatment. The date of initiation of ADT was recorded. The duration of response to ADT was defined as time until a subsequent PSA rise in the presence of castrate testosterone levels or the absence of PSA response to ADT. Overall survival after CastratePSA was estimated by Kaplan Meier analysis.

Results: Median time from BCR to ADT was 3.4 (95% CI, 2.5-4.3) years, and the actuarial probability of receiving ADT was 80% by 12 years after BCR. At the time of analysis 309 of 630 BCR patients had begun ADT: 255 before and 54 after clinical Mets. Median duration of response to ADT was 10.8 (95% CI, 10-11.5) yrs. In 85 patients (20 with ADT after Mets) the PSA subsequently rose (CastratePSA), and 62 of these have died, all from prostate cancer. Median survival after CastratePSA was 25.6 (range 6-54) months. Significant independent predictors of survival for patients with CastratePSA in multivariable analysis were PSA doubling time (p=0.05), PSA level before ADT (p=0.08) and duration of response to ADT (p=0.01). These variables were used to construct a prognostic nomogram for this disease state.

Conclusions: For patients with BCR after RP the duration of response to ADT is long (median 10.8 yr) and some patients respond for >15 yr. Emergence of a CastratePSA state, noted by a rise in PSA with or without detectable metastases, carries an ominous prognosis, with median survival of 2 yr and no patient alive 5 yr later.

4553

What is the Probability of a Positive Bone Scan (+BS) in Patients with a Rising PSA After Radical Prostatectomy (RP): A New Nomogram

M. W. Kattan, Z. A. Dotan, F. J. Bianco, F. Rabbani, J. A. Eastham, H. I. Scher, C. Hui-Ni, H. Schoder, H. Hricak, P. T. Scardino; MSKCC, New York, NY

Background: Physicians often order bone scans periodically to assess for metastases in patients with a rising PSA after RP (biochemical recurrence, BCR), but most scans are negative. We studied pre- and postoperative characteristics of patients to find factors predictive of a positive scan.

Methods: From our SPORE prostate cancer database we identified all patients with BCR (3 subsequent peaks in PSA =0.1 ng/ml) after RP. We analyzed the following features at the time of each BS for their association with a positive scan: preoperative PSA, neoadjuvant hormonal therapy (NHT), time to BCR, pathologic findings (surgical margin, extracapsular extension, seminal vesicle invasion, lymph node metastases, and pathologic Gleason score), PSA level before the BS (trigger PSA), PSA doubling time (PSA DT), and time from BCR to BS were incorporated into a predictive model.

Results: There were 927 scans in 330 patients with BCR. We excluded 318 scans done while the patient was on androgen deprivation therapy. Of the 609 eligible scans, only 84 (13.8%) were positive for metastases. In multivariate analysis trigger PSA (P=0.014), NHT (P=0.03), and time from BCR to bone scan (P=0.05) predicted a +BS. Pathologic variables and PSA DT did not add to the accuracy of the model. A nomogram for predicting the bone scan result was constructed with a concordence index of 0.9. A policy of ordering a BS only when the probability of a +BS is >15% would eliminate 76% of the scans currently ordered in patients with BCR after RP and will lead to a significant savings in the health care expenses.

Conclusions: Patients with BCR after RP have a low frequency of +BS (13.8%). Trigger PSA, NHT, and time from BCR to scan were significant predictors of a +BS. A highly accurate nomogram can be used to select patients according to their risk for a positive scan. Omitting scans in low risk patients could substantially reduce costs.

4554

Prostate-specific Antigen Doubling Time as a Predictor of Prostate Cancer Disease Progression and Survival

J. B. Nelson, A. R. Allen, S. M. Hulting, J. D. Isaacson, D. S. Sleep; University of Pittsburgh School of Medicine, Pittsburgh, PA; Abbott Laboratories, Abbott Park, IL

Background: Prostate-specific antigen doubling time (PSADT) has been identified as a predictor of outcome in prostate cancer. Atrasentan, a potent, selective, oral endothelin A receptor antagonist, significantly decreased time to prostate-specific antigen (PSA) progression in a randomized, double-blind, placebo-controlled phase 2 trial in 271 men with metastatic HRPC. In that study, PSADT < 4 mo was predictive of rapid disease progression and decreased survival in men with metastatic hormone-refractory prostate cancer (HRPC). Our objective was to confirm the utility of PSADT as a predictor of disease progression and survival in a larger population of subjects from combined phase 2 and 3 trials.

Methods: Data from randomized, double-blind, placebo-controlled phase 2 and 3 trials of metastatic HRPC patients given either placebo or atrasentan (10 or 2.5 mg QD) were combined. Measurements from 982 patients with a minimum of 3 PSA values on treatment were analyzed. Data from all treatment arms were pooled and stratified by PSADT < 4 mo vs ≥4 mo. Kaplan-Meier and Cox proportional hazard methodologies were used to compare time to disease progression (TTP) and survival between the PSADT < 4 mo and PSADT ≥4 mo groups.

Results: 982 patients were included in the analyses: 484 patients had a PSADT < 4 mo and 498 had a PSADT ≥4 mo. Patients with a PSADT ≥4 mo showed a statistically significant longer median TTP (178 days), more than twofold greater than those with a PSADT < 4 mo (85 days, log-rank p < 0.001). Survival in patients with a PSADT < 4 mo was significantly shorter than that in patients with a PSADT =4 mo (log-rank p < 0.001). Median survival was 445 days for patients with PSADT < 4 mo vs 746 days for patients with PSADT ≥4 mo.

Conclusions: These data confirm earlier findings that a PSADT < 4 mo is predictive of early disease progression and reduced survival of men with metastatic HRPC, independent of therapy.

4555

Prostate Specific Antigen Doubling Time (PSADT) Predicts for Distant Failure and Prostate Cancer Specific Survival (PCSS) in Men with Biochemical Relapse After Radical Prostatectomy (RP)

A. W. Partin, M. A. Eisenberger, V. J. Sinibaldi, E. Humphreys, L. A. Mangold, P. C. Walsh; Johns Hopkins Medical Institutions, Baltimore, MD

Background: Patients (pts) with biochemical (PSA) relapse after RP represent a rapidly rising population in whom appropriate standards of care remain undefined. Over the past several years we have focused on studying the natural history of these pts. Our previous report (Pound et al; JAMA 1999) suggests that the Gleason score, time of PSA relapse (>0.2 ng/ml) and PSADT are the strongest predictors of the probability of distant metastasis (DM) (Cox model). A recent update (Partin; AUA 2003, Eisenberger; ASCO 2003) suggests that PSADT overrides the other parameters in the prediction of DM. Our experience has been expanded (> 1000 additional pts and longer follow-up). We evaluate the relationship between PSADT and PCSS.

Methods: 4,415 pts underwent a RP from 1982-2003 and 825 demonstrated evidence of PSA relapse. All pts were followed with yearly PSAs with no androgen deprivation given until development of DM. Cox-proportional hazards and Kaplan-Meier survival analyses were completed with time from PSA relapse to censoring (last follow-up) or death used as endpoint.

Results: 825/4,415 demonstrated PSA relapse (mean time 8.4yrs), 170/825(20.6%) developed DM and 109/825(13%) have died of prostate cancer (PCa). PSADT was calculated on 411 pts (mean 9 PSAs /pt, range 2-33), ROC-AUC for PSADT prediction of PCSS was 0.77.

Conclusions: PSADT is a potent predictor of the probability of DM and PCSS. Pts with PSADT ≤10 mos. are at significant risk for developing DM and dying of PCa. In this group, clinical trials should evaluate progression-free and overall survival with aggressive approaches. Pts with PSADT of >10 mos. may be observed or treated to delay onset of DM. (Supported by CAPCURE).

4566

Treatment "Mismatch" in Early Prostate Cancer: An Empirical Measure of Patient-physician Communication

R. C. Chen, J. A. Clark, S. P. Mitchell, J. A. Talcott; Harvard Medical School, Boston, MA; Boston University School of Public Health, Boston, MA; Massachusetts General Hospital, Boston, MA

Background: Because complications vary by treatment, preexisting organ dysfunction may relatively contraindicate specific treatment modalities for early prostate cancer. However, dysfunction may not be fully communicated between patient and doctor. We compared patient-reported pretreatment dysfunction with the treatment modality received to empirically assess this process.

Methods: In a prospective, 438-patient cohort study, we determined how often patients with "intermediate" and "poor" urinary, sexual, and bowel function received relatively contraindicated brachytherapy (BT), nerve sparing radical prostatectomy (NSRP), or external beam radiotherapy (EBRT) respectively. Because age affects treatment choice, we assessed receipt of contraindicated (i.e.,"mismatched") BT in all men, NSRP in men = 65 years, and EBRT in men >65 years. To assess the actual impact of "mismatched" treatments, we analyzed functional outcomes.

Results: Overall, 242 patients reported "intermediate" to "poor" baseline urinary, sexual, and/or bowel function; 50 patients (21%) received "mismatched" treatments. Of 196 patients in our cohort with baseline urinary obstructive/irritative symptoms, 28 (14%) received BT (see table). Of 83 patients 65 years old or younger with intermediate or poor baseline sexual function, 17 (20%) received NSRP. Of 213 patients older than age 65, 7 had baseline bowel dysfunction, but 5 (71%) received EBRT. "Mismatched" patients tended to have greater increases in dysfunction, although small sample sizes limited analysis.

Conclusions: Patients reporting dysfunction received relatively contraindicated treatment modalities surprisingly often. While multiple factors affect treatment decisions, incomplete patient-physician communication regarding baseline patient dysfunction may compromise the quality of care.

4567

Fifteen Year Follow Up of the First Cohort of Localized Prostate Cancer Patients Treated with Brachytherapy

J. E. Sylvester, J. C. Blasko, P. D. Grimm, R. Meier, J. F. Spiegel, J. A. Malmgren; Seattle Prostate Institute, Seattle, WA; HealthSTAT Consulting, Inc., Seattle, WA

Background: The role of external beam radiation in addition to brachytherapy continues to be scrutinized for long term disease free survival.

Methods: A cohort of 223 localized prostate cancer patients treated with iodine125 or palladium103 brachytherapy and neoadjuvant external beam radiation therapy were followed for biochemical failure, local or distant disease recurrence, and vital status. Biochemical failure (BF) was assigned if there were two consecutive PSA rises with the second greater than the first after a nadir post implant. Vital status was obtained from chart review, the social security death index, the SEER registry, and death certificate records. Time to BF was calculated and compared using Kaplan-Meier plots. Risk categories were derived from multivariate modeling and are defined as low = PSA ≤10 ng/mL, Gleason score (GS) ≥7, and stage ≥T2c; intermediate PSA >10 ng/mL or GS >7 or stage >T2c; and high = 2 or more intermediate risk factors.

Results: Implants were conducted from 1987 to 1994 with average patient age at implant equal to 69 years. Median follow up time was 10 years (range 1.4 to 16.3). Twenty seven percent were low risk (n=61), 41% were intermediate (n=91) and 32% were high risk (n=71). The disease specific death rate from prostate cancer was 14% (29/212) for the sub group with known cause of death. Cause of death was unknown for 11 of 115 decedents. Fifteen year biochemical relapse free survival by risk category was the following: low = 86%, intermediate = 72% and high = 47% (log rank test score = 30.76, p=.000). Fifteen year survival was 93% for low risk, 86% for intermediate risk and 59% for high risk (log rank test score = 31.64, p = .000).

Conclusions: Excellent long term biochemical and disease specific survival outcomes are achievable with prostate brachytherapy in low and intermediate risk groups. High risk patients could possibly benefit from additional adjuvant therapy.

4568

A Phase II Study of External Beam Radiation Therapy Combined with Permanent Source Brachytherapy for Intermediate Risk Clinically Localized Adenocarcinoma of the Prostate: Preliminary Results of RTOG P-0019

W. R. Lee, M. Desilvio, C. F. Lawton, M. T. Gillin, G. C. Morton, S. Firat, M. Baikadi, M. R. Kuettel, K. M. Greven, H. M. Sandler; Wake Forest University, Winston Salem, NC; RTOG, Philadelphia, PA; Medical College of Wisconsin, Milwaukee, WI; M.D. Anderson Hospital, Houston, TX; Toronto-Sunnybrook Regional Cancer Center, Toronto, ON, Canada; University of Pennsylvania, Philadelphia, PA; Roswell Park Cancer Institute, Buffalo, NY; University of Michigan, Ann Arbor, MI

Background: External beam radiation therapy is commonly combined with permanent source brachytherapy in the treatment of prostate cancer. The primary objective of this study is to estimate the rate of acute- and late-grade 3-5 genitourinary and gastrointestinal toxicity following treatment with external beam radiation therapy and permanent source brachytherapy in a multi-institutional, cooperative group setting.

Methods: Between January 2001 and November 2001, 138 patients from 20 institutions were entered on this study. Acute toxicity information is available in 131 patients, and 127 patients are analyzable for late toxicity. All patients were treated with external beam radiation therapy (45 Gy/25 fractions) followed 2-6 weeks later by an interstitial implant using I-125 to deliver an additional 108 Gy. The median followup is 20 months. Median pretreatment PSA = 7.5.

Results: The most commonly reported acute toxicity is urinary frequency and dysuria. Acute Grade 3 toxicity was documented in 10/131 (7.6%) patients. No Grade 4-5 acute toxicity has been observed. Late Grade 3 toxicity has been observed in six men (5 urinary, 1 bowel). The 18-month month estimate of late Grade 3 GU/GI toxicity is 3.3% (95% CI 0.1-6.5). No late Grade 4-5 toxicity has been observed. In the sixty-one men that reported no impotence at baseline and received no androgen deprivation, the 18-month rate of Grade 2-3 impotence is 45.5% (95% CI 32.6-58.5%).

Conclusions: The acute and late morbidity observed in this multi-institutional, cooperative group study is consistent with previous reports from single institutions with significant prostate brachytherapy experience. The combination of external beam radiation therapy and permanent source brachytherapy is worthy of further study and is currently under evaluation in RTOG 0232, a phase III trial that compares permanent source brachytherapy as monotherapy to the combination of external beam treatment followed by brachytherapy.

4569

Prostate Cancer Treated With Radiotherapy With or Without Androgen Deprivation: The Importance of the PSA Nadir Within 12 Months

A. Pollack, A. L. Hanlon, E. M. Horwitz, S. J. Feigenberg, R. G. Uzzo; Fox Chase Cancer Center, Philadelphia, PA

Background: The nadir in PSA after prostate cancer treatment with 3D conformal radiotherapy (RT) may take years to occur; it is these kinetics which limit the clinical application of this highly significant predictor of outcome. In this report the value of the nadir PSA within 12 months of completing RT (nadir PSA12) was determined for patients treated with RT alone or RT plus androgen deprivation (AD).

Methods: Between 5/1989-11/1999, 997 men received RT alone and 153 received RT+AD (median duration of AD 3 mo; range 1-23 mo; 127 men received ≤ 6 mo). The median follow-up from the completion of all treatment was 59 mo and 67 mo for the RT alone and RT+AD groups. The median RT doses were 76 Gy (range 63-84 Gy) and 77 Gy (range 64-82 Gy) for the RT alone and RT+AD groups. Multivariate analysis (MVA) via Cox proportional hazards regression was used to determine the association of nadir PSA12 to biochemical failure (BF; ASTRO definition), distant metastasis (DM) and overall mortality (OM).

Results: The median nadir PSA12s were 1.3 (range < 0.1-177.0) and 0.3 ng/ml (range < 0.1-9.1) in the RT alone and RT+AD groups. For those who received RT alone, nadir PSA12 was significantly related in the MVAs to BF (p=0.0005), DM (p < 0.0001) and OM (p < 0.0001), independently of the other significant covariates of RT dose (continuous), T-stage (T1/T2ab vs T2c/T3) , Gleason score (2-6 vs 7 vs 8-10) and initial pre-RT PSA (iPSA; continuous) for BF and all but iPSA for DM and OM. For those who received RT+AD, nadir PSA12 was significantly related to BF (p=0.0003) and DM (p < 0.0001), independently of the other significant covariates of RT dose and T-stage for BF and Gleason score for DM; there were not enough events to run an MVA for OM. A nadir PSA12 of >2 ng/ml was associated with 5 yr BF, and DM rates of 36% and 8% for RT alone and 100% and 50% for RT+AD. The 5 yr OM rate for those with a nadir PSA12 >2 ng/ml after RT alone was 8%.

Conclusions: Nadir PSA12 was a strong independent predictor of eventual BF, DM and OM, providing meaningful early information that may prove useful in directing additional treatment, especially for those who received RT+AD.

4570

Immediate Hormonal Therapy Versus Observation After Radical Prostatectomy and Pelvic Lymphadenectomy for Node Positive Prostate Cancer: At 10 Years Results of EST3886

E. M. Messing, J. Manola, M. Sarosdy, G. Wilding, D. Crawford, M. Kiernan, D. Trump; University of Rochester, Rochester, NY; Dana Farber Cancer Institute, Boston, MA; South Texas Urology, San Antonio, NY; University of Wisconsin, Madison, NY; University of Colorado, Denver, NY; Roswell Park Cancer Institute, Buffalo, NY

Purpose: We previously reported 7 year follow-up of men with clinically localized prostate cancer, who underwent radical prostatectomy (RP) and pelvic lymphadenectomy (PLD) and were found to have nodal metastases (N+), had significantly improved survival and disease-specific survival with immediate (within 3 months of surgery) and continuous hormonal monotherapy (HT) compared with those who had androgen ablation withheld until distant metastases were identified (NEJM 341:1781,1999). To determine the durability of this effect, we update results at 10 years median follow-up.

Methods: 98 men were randomized to receive Goserelin acetate or bilateral orchiectomy (N=47) vs. observation (N=51) until distant disease was documented. Patients were followed with semi-annual exams and laboratory tests, and annual bone scans.

Results: Patients were well-matched for age (mean 65.9 yrs), prostate size, local extent, and Gleason score. Only 20% of men had detectable PSA's following their surgery at the time of randomization. At median follow-up of 10 years, 13 of 47 men who received immediate HT died (6 of prostate cancer, 7 of non-prostate cancer causes with undetectable PSA's at the time of death) vs 26 in the deferred arm (22 of prostate cancer, 4 of non-prostate cancer). The differences in both overall (immediate 72.4%; deferred 49.0%, p = .025), and cause specific (immediate: 87.2%; deferred: 56.9%, p = .001) survivals, between the two arms were significant. Toxicities of treatment were tolerable, and at last follow-up, no immediately treated man had experienced an osteoporotic fracture or had discontinued HT.

Conclusion: 10 years, early HT in patients with N+ disease following RP+PLD continues to be associated with highly significant improvements in overall and disease-specific survival. The treatment has been generally well tolerated. Early HT's role in other disease scenarios warrants testing in prospective randomized studies.

4575

Continuing Benefit of Zoledronic Acid for the Prevention of Skeletal Complications in Men with Advanced Prostate Cancer

F. Saad, D. M. Gleason, R. Murray, N. S. S. Tchekmedyian, P. Venner, L. Lacombe, J. Chin, J. J. Vinholes, M. Zheng, Y.-J. Hei; Centre Hospitalier de l'Universitè de Montrèal, Montrèal, PQ, Canada; Advanced Clinical Therapeutics, Tuscon, AZ; Peter MacCallum Cancer Centre, East Melbourne, Victoria, Australia; Pacific Shores Medical Group, Long Beach, CA; Cross Cancer Institute, Edmonton, AB, Canada; LeCentre de Recherche en Cancerologie del'Univ Lav, Quebec, PQ, Canada; London Health Sciences Centre, East London, ON, Canada; Irmandade da Santa Casa de Misericordia de Porto, Rio Grande do Sul, Brazil; Novartis Pharmaceuticals Corporation, East Hanover, NJ

Background: We previously reported the efficacy and safety of 4 mg zoledronic acid (Zol) at 15 months compared with placebo in patients with bone metastases from prostate cancer (Saad et al. J Natl Cancer Inst. 2002;94:1458-1468). We report here the results of an exploratory analysis of the 15- to 24-month extension phase of this study to evaluate the benefit of Zol treatment beyond 15 months.

Methods: Data were collected during months 15 to 24 for patients who chose to enter the 9-month extension phase. Endpoints included the percentage of patients with a skeletal-related event (SRE), time to first SRE, skeletal morbidity rate (SMR), and multiple event analysis using the method of Andersen and Gill.

Results: Of 422 patients randomized to the 4-mg or placebo treatment arms, 147 completed the 15-month core phase, and 133 elected to enter the extension phase. Among these patients, multiple event analysis revealed a 53% reduction in the risk of developing an SRE for patients treated with 4 mg Zol compared with placebo (risk ratio = 0.467; 95% CI = 0.243, 0.897; P = .022). Additionally, only 19% of patients treated with 4 mg Zol had an SRE during the extension phase versus 38% of patients treated with placebo (P = .017). Time to first SRE was significantly extended in the 4-mg Zol group compared with placebo (P = .036), and mean SMR was reduced approximately 2-fold (0.42 for 4 mg Zol versus 0.88 for placebo; P = .016).

Conclusions: This exploratory analysis suggests that long-term treatment with Zol provides significant and ongoing clinical benefit in patients with bone metastases from advanced prostate cancer. Considering that patients who have an SRE are also at higher risk for a subsequent SRE, this analysis suggests that patients should continue on bisphosphonate therapy even after they have an SRE.

4576

Clinical Benefit of Zoledronic Acid for the Prevention of Skeletal Complications in Patients with Prostate Cancer Based on History of Skeletal Complications

J. L. Chin, F. Saad, D. M. Gleason, R. Murray, S. Tchekmedyian, P. Venner, L. Lacombe, J. J. Vinholes, M. Zheng, Y.-J. J. Hei; London Health Sciences Centre, East London, ON, Canada; Centre Hospitalier de l'Universitè de Montrèal, Montrèal, PQ, Canada; Advanced Clinical Therapeutics, Tucson, AZ; Peter MacCallum Cancer Institute, East Melbourne, Australia; Pacific Shores Medical Group, Long Beach, CA; Cross Cancer Institute, University of Alberta, Edmonton, AB, Canada; LeCentre de Recherche en Cancerologie del'UnivLaval, Quebec, PQ, Canada; Irmandade de Santa Casa de Misericórdia de Porto, Rio Grande do Sul, Brazil; Novartis Pharmaceuticals Corporation, East Hanover, NJ

Background: We previously reported the efficacy and safety of 4 mg zoledronic acid (Zol) at 15 mo compared with placebo in patients with bone metastases from prostate cancer (Saad et al. J Natl Cancer Inst. 2002;94:1458-1468). We report here the results of a retrospective analysis performed to determine the clinical benefit of Zol in men with prostate cancer based on patients' history of skeletal complications before study entry.

Methods: Data were analyzed for patients treated with 4 mg Zol or placebo every 3-4 weeks for 24 months. Patients were retrospectively stratified according to history of a skeletal-related event (SRE) before study entry.

Results: Of 422 patients randomized to 4 mg Zol or placebo, 147 completed the 15-month study and 133 elected to enter 9-month extension. Approximately one third of patients had an SRE before study entry. In the placebo group, 51% of patients with a history of an SRE developed an on-study SRE vs 47% of patients without an SRE before study entry. Regardless of SRE history, Zol treatment reduced skeletal morbidity compared with placebo. Multiple event analysis demonstrated that among patients with an SRE before study entry, the risk of developing an SRE was reduced by 40% for patients treated with Zol 4 mg vs placebo (hazard ratio = 0.603; P = .028). For patients without an SRE before study entry, the risk of developing an on-study SRE was also reduced (33%) relative to placebo (hazard ratio = 0.670; P = .027). In the placebo group, median time to first SRE appeared to be shorter for patients with a prior SRE vs no prior SRE. However, in both stratification groups Zol 4 mg extended median time to first SRE relative to placebo (361 days vs 258 days for placebo in patients with a prior SRE and 499 days vs 337 days for placebo in patients without a prior SRE; P = .066 and .065, respectively).

Conclusions: This exploratory analysis suggests that Zol provides clinical benefit in patients with advanced prostate cancer and bone metastases independent of their prior history of skeletal complications.

4577

Need for Awareness and Monitoring of Ocular Toxicities (OT) Due to Weekly Docetaxel Administration: Experience During a Trial of Neoadjuvant Docetaxel (D) and Mitoxantrone (M) for Patients with High-risk Prostate Cancer (PC)

C. S. Higano, T. M. Beer, M. Garzotto, C. W. Ryan, M. Pitzel, C. R. Works, S. P. Hall; University of Washington, Seattle, WA; Oregon Heath and Sciences University, Portland, OR

Background: Epiphora (E), excessive tearing (ET) due to insufficient drainage, is toxicity unique to 5FU or D. It is more common with q wk versus q 3 wk D (Esmaeli, Ophthalmology 2002). D is secreted in tears and causes inflammation and canalicular stenosis resulting in E, which can cause significant problems with reading, driving, and other activities as well as present the false impression of crying. While it has been reported mainly in pts with breast cancer, E has also been seen in other tumors treated with weekly D. Our preliminary experience in PC pts is presented.

Methods: Pts with high-risk PC were treated with neoadjuvant D 35 mg/M2 and M 4 mg/M2 q wk x 3, with 1 wk off, for a total of 12 doses prior to radical prostatectomy. Pts who developed ET were instructed to use artificial tears; if symptoms (sx) did not improve, 1% prednisolone acetate eye drops were prescribed on the day before, the day of, and the day after chemotherapy.

Results: Of 41 pts on protocol, 21 had ocular sx including blurry vision (1), eye irritation (1), ET (15), dry eyes (2-1 had ET later), photophobia (1), ingrown eyelashes (1), or infection of the upper eyelid (1). ET occurred at a median of 7.5 weeks (range 1-17) after starting chemotherapy. None of 6 pts treated with steroid eye drops improved and 1 with pre-existing glaucoma experienced an increase in intra-ocular pressure (IOP) requiring addition of a beta-blocker eye drop. In five pts who saw an ophthalmologist, ET was due to canalicular stenosis (2), ingrown eyelashes (1), "floppy eyelid syndrome"(1), normal (1).

Conclusions: Weekly D and M can cause significant OT. Given these findings, baseline ophthalmologic evaluation and monthly follow-up are recommended. As weekly D administration is widespread and the impact of OT on quality of life can be significant, prophylactic use of artificial tears and/or steroid drops, and early intervention with silicone intubation of the canaliculi should be studied. Better documentation of the range and grading of OT due to D is needed.

4578

The Effect of Hormonal Therapy for Prostate Cancer on the Electrocardiographic QT Interval: Phase 3 Results Following Treatment with Leuprolide and Goserelin, Alone or with Bicalutamide, and the GnRH Antagonist Abarelix

M. B. Garnick, C. M. Pratt, M. Campion, J. Shipley; Praecis Pharmaceuticals Incorporated, Waltham, MA; Baylor College of Medicine, Houston, TX

Background: The electrocardiographic QT interval measures ventricular repolarization, and, if prolonged, may predispose to arrhythmias and sudden death. Preclinically, testosterone (T) deficiency may prolong the QT interval. However, the effect of hormone treatments (HTs) for prostate cancer which induce androgen deficiency (AD) has not been well studied.

Methods: Three randomized Phase III studies were conducted that compared the GnRH antagonist abarelix to either goserelin plus bicalutamide (G+B) (Study 1); leuprolide (L) monotherapy (Study 2; Urology 58:756, 2001); or L plus B (L+B) (Study 3; J. Urol 167:1670, 2002) in patients (pts) with prostate cancer undergoing initial HT and included pts with Stage D1/D2 disease, rising PSA, intermittent HT or neoadjuvant HT. Study 1 prospectively evaluated data from 177 pts; Studies 2 and 3 retrospectively evaluated 299 of 520 treated pts. Barzett and Fredericia QT intervals (QTcB, QTcF) were measured at baseline (BL), days(D) 84 and 336 of treatment in Study 1; and at BL and D 169 in Studies 2 and 3.

Results: Median T levels (ng/mL) at BL, D 85 and 336 in Study 1: G+B: 347; 23; 26, respectively; abarelix: 316, 23, 35. In Studies 2 and 3, T levels at BL and D169: L: 342, 9; L+B: 358, 13; abarelix (pooled): 361,14, respectively. Median QTcF (msec) values at BL and median change from BL on study are: Study 1: G+B: 402, +18; abarelix: 407, +13 (p=0.018 t-test; 0.015 Wilcoxon rank sum). Study 2 and 3: L: 413, +21; L+B: 417, +9; abarelix: 410, +12 (p values from pairwise comparisons, all NS). Percentage of pts with maximum change in QTc of >30 to < 60 msec: Study 1: G+B: 46%; abarelix: 39%; Studies 2 and 3: L: 27%; L+B: 26%; abarelix: 17%. Percentage of pts with maximum change in QTc of = 60 msec: Study 1: G+B: 8%; abarelix: 2%; Studies 2 and 3: L: 6%; L+B: 0%; abarelix: 4%.

Conclusions: HTs which induce AD are all associated with prolongation of the QT interval. These findings should be considered in assessing the risk and benefit of HT, especially in those with BL QT values of >450 msec and who are taking Class IA or III antiarrhythmics.

4582

Impact of Atrasentan on Prostate-Specific Outcomes with Hormone Refractory Prostate Cancer Patients

Background: For men with advanced cancer facing reduced life expectancy and no possibility of cure, symptom relief and maintenance of function are primary objectives of treatment. The Functional Assessment of Cancer Therapy-Prostate (FACT-P) quality of life (QOL) instrument incorporates a disease-specific, 12-item prostate cancer (PCa) subscale (PCS) that includes assessment of PCa-related symptoms such as pain and fatigue. The recently developed FACT Advanced Prostate Symptom Index (FAPSI), an 8-item index derived from the FACT-P, assesses weight loss, urinary problems, pain and fatigue. Scores range from 0-48 for PCS and 0-32 for FAPSI with higher scores indicating better QOL and fewer symptoms. The impact of atrasentan on PCS and FAPSI was evaluated in a phase 3, randomized, placebo-controlled study in men with metastatic hormone refractory PCa (HRPC) (N=809).

Methods: All randomized patients with a baseline and final QOL assessment were analyzed. Mean change from baseline to final PCS and FAPSI scores was analyzed using analysis of covariance to compare treatments. The same analyses were performed on a subset of patients (N = 474) with HRPC metastastic only to bone.

Results: Patients treated with placebo recorded a greater deterioration in both the PCS and FAPSI than those receiving atrasentan (p=0.032 and p=0.081). For both indices, the treatment difference was greater and statistically significant in men with bone metastases. The difference between groups was less than the minimally important difference (MID) of 3 points. However, as a group, the placebo-treated patients deteriorated beyond the MID, whereas the atrasentan-treated patients did not.

Conclusion: These data indicate that atrasentan modestly improves PCa-specific outcomes relative to placebo. The QOL benefit observed with atrasentan therapy was most apparent in HRPC patients with metastases only to bone.

4589

Predictors of Prostate Cancer in Men Undergoing Serial Prostate Biopsies: Results from 371 Patients

M. Garzotto, S. Mongoue-Tchokote, J. Bledsoe, R. O. Parra, L. Peters, Y. Park, T. M. Beer, M. Mori; Portland VA Medical Center, Portland, OR; Oregon Health & Science University, Portland, OR

Background: Due to a high rate of false negative prostate biopsies, a repeat biopsy is often recommended to men if there is a suspicion of occult carcinoma. We sought to determine the rate of prostate cancer (PC) detection and identify risk factors for cancer detection in men who despite suspicion of PC had a negative biopsy initially and then underwent serial prostate biopsies. Methods: Records of all men undergoing at least one repeat biopsy of the prostate from 1992 to 2003 were reviewed. Variables recorded were: age, race, family history of PC (FH), body mass index, biopsy indication, all PSA values, digital rectal exam (DRE), PSA density (PSAD; i.e. PSA/prostate volume), ultrasound findings, total number of cores obtained at all procedures, and the presence of high-grade PIN on initial biopsy. Outcome was defined as time from the initial biopsy to cancer detection or last follow-up. The data were analyzed by Kaplan-Meier method, log-rank test (univariate analysis), stepwise Cox regression (multivariate analysis) and binary recursive partitioning for censored data (survival tree).

Results: 373 patients underwent 977 biopsy procedures. Median FU was 3.1 years for all patients and 3.6 years for patients without PC. PC was detected in 105 of 373 (28.7%) patients during follow-up. The estimated median time to PC detection for the entire cohort was 10.3 years. Factors significantly associated (p < 0.05) with a positive biopsy in the univariate analysis were PSADT, PSAD, biopsy indication, and PIN. The survival tree analysis identified 4 distinct risk groups with estimated 5-year PC detection rates (±S.E.) of 21 ± 4%, 40 ± 7%, 58 ± 8% and 100%. The Cox regression analysis showed:

Conclusions: Patients who are referred for suspicion of PC but have a negative biopsy can be stratified into clinically meaningful risk groups. Prospective validation of this model is warranted.

4593

A Phase II Trial of Celecoxib in PSA Recurrent Prostate Cancer After Definitive Radiation Therapy or Radical Prostatectomy

R. S. Pruthi, E. Derksen; University of North Carolina at Chapel Hill, Chapel Hill, NC

Purpose: COX-2 inhibitors have also been shown to have anti-neoplastic effects in a variety of tumor types including prostate cancer. This trial evaluated the efficacy of COX-2 inhibitor celecoxib in PSA-recurrent prostate cancer after radiation therapy (XRT) or radical prostatectomy (RP).

Methods: 24 patients with rising PSA after definitive XRT (n=4) or RP (n=20) were treated with celecoxib at 400 mg/day (n=12) or 800 mg/day (n=12) (bid dosing). PSA levels were obtained at 3, 6, and 12 months after initiation of treatment. Data was evaluated by calculating 1) PSA doubling times (PSADT) and 2) slope of logPSA vs. time before and after celecoxib treatment at each time point. Testosterone levels were also obtained.

Results: 22 of 24 (92%) patients had an inhibitory effect on their serum PSA after 3 mos of treatment: 8/24 had a decline and 3/24 had stabilization of their PSA. Of the remaining 13 patients, 11 had slowing of their PSADT - mean increase (i.e. slowing) in PSADT of 4.5-fold from pre-treatment. Only 2 patients (12%) had no initial change in their PSADT at 3 mos., yet both patients eventually showed increase of PSADT (2.2-fold and 4.0-fold) by 12 mos. Short-term responses at 3 mos. continued at 6 and 12 mos. of treatment. There was a significant shift of patients with rapid towards slower/stable PSADT at all time points (p = 0.008) (table). Comparison of rate of PSA rise before vs. after celecoxib treatment showed significant flattening of mean slope of logPSA vs. time from pre-treatment (1.48) vs. 3 mos. (0.09), 6 mos. (0.35) and 12 mos. (0.46) (p < 0.05 for each time point), There was no significant change in testosterone levels suggesting an androgen-independent mechanism.

Conclusions: COX-2 inhibitors may decrease serum PSA levels in patients with biochemical progression after XRT or RP. These results suggest that COX-2 inhibitors may help delay disease progression in these patients.

4595

Prior Estrogen Therapy as a Predictor of Response to Subsequent Estramustine-based Therapy in Androgen-independent Prostate Cancer Patients

J. A. Garcia, V. Weinberg, B. I. Rini, J. E. Rosenberg, C. J. Ryan, E. J. Small; University of California, San Francisco, San Francisco, CA

Background: Secondary hormonal therapy with estrogens results in PSA responses in 30-60% of androgen-independent prostate cancer (AiPCa) patients. Diethylstilbestrol (DES) and other estrogens share a similar chemical structure with estramustine phosphate, which is commonly used together with taxanes as chemotherapy for AiPCa patients. It is unknown if prior estrogen therapy impacts on subsequent responses to estramustine-based chemotherapy.

Methods: A retrospective chart review of 76 consecutive AiPCa patients who received secondary hormonal therapy with estrogens (DES or PC-SPES) was conducted. Patients who went on to receive estramustine-based chemotherapy were identified and their medical records abstracted. PSA response proportions (using Consensus Criteria), duration of response and overall survival was calculated.

Results: Between 1998 and 2003, 76 AiPCa patients were treated with an estrogen (DES or PC-SPES). 48/76 (63%) had a PSA response. Of these, 24 subsequently were treated with estramustine-based chemotherapy. All but 1 had failed to respond to estrogen or developed progressive disease while on estrogen. 23/24 (96%) had metastatic disease. Median PSA was 69.1ng/mL. A PSA response to estramustine-based chemotherapy was observed in 13/23 (57%; CI = 34%-77%). Prior response to DES or PC SPES did not predict the likelihood of subsequent response to estramustine-based therapy (63% response rate in prior estrogen responders, 53% response rate in prior estrogen non-responders, p = 0.67).

Conclusions: The use of prior estrogen therapy in AiPCa patients does not have an obvious impact on the likelihood of subsequent response to estramustine-based therapy, with 57% of patients responding to estramustine-based therapy following estrogen therapy. Prior response to estrogen was not predictive of response to estramustine-based therapy. These data suggest that estrogen and estramustine have different mechanisms of action, and that prior estrogen therapy need not exclude AiPCa patients from estramustine-based regimens.

4596

Statin Use and Prostate Cancer Risk

J. Shannon, M. Garzotto, A. J. Palma; Oregon Health & Science University, Portland, OR

Background: One in seven men over age 60 will be diagnosed with prostate cancer. Improvements in screening and diagnostic procedures allow clinicians to better identify men at high risk of having prostate disease, but no established means of preventing prostate cancer are currently available. Statins are a group of drugs used to control high cholesterol through inhibiting HMG CoA reductase and interfering in the mevalonate pathway. Recent research has demonstrated that in a number of tumor types, statins also induce apoptosis and growth arrest. The purpose of this pilot study was to evaluate the association between statin use and prostate cancer risk in a population of older veterans enrolled in a case-control study of diet and prostate cancer risk.

Methods: Cases were recruited upon referral for prostate biopsy and frequency age-matched; PSA normal controls were identified from the VA primary care clinic. Statin use was determined from medical record review. Years of use, type of statin, dose and any changes were recorded. Years of use were categorized as > or < = 3 years of use of any single or combined statin. Dose was categorized as > or < = 40mg/day, based on the highest dose prescribed.

Results: Records from 47 incident prostate cancer cases and 142 PSA normal clinic controls were reviewed. After adjusting for age and body mass index, men with any recorded statin use had a 56% lower risk of prostate cancer as opposed to non-users (odds ratio (OR) 0.44, 95% confidence interval (CI) (0.22-0.88)). When statin use was grouped by duration and dose, those with the longer duration of use (>3 years) and higher dose (>40mg/day) were at lower risk of prostate cancer as compared to non-users (OR=0.35, 95% CI=0.12-1.00).

Conclusions: Results of this pilot study suggest that any statin use is associated with a reduced risk of prostate cancer and that use for a longer period of time at a higher dose may confer the most risk reduction. This study is limited by a small sample size and lack of control for additional comorbid factors. If these results are confirmed in a larger prospective study they may provide necessary evidence to consider the use of statin drugs in prostate cancer prevention.

4597

Toremifene Citrate Versus Placebo for Treatment of Bone Loss and Other Complications of Androgen Deprivation Therapy in Patients with Prostate Cancer

M. S. Steiner, A. Patterson, R. Israeli, K. G. Barnette, R. Boger, D. Price; GTx, Inc, Memphis, TN; University of Tennessee Urology Research, Memphis, TN; Staten Island Urological Research, PC, Staten Island, NY; Regional Urology LLC, Shreveport, LA

Background: Patients with prostate cancer receiving androgen deprivation therapy (ADT) may experience side effects including bone loss (osteoporosis), hot flashes, and gynecomastia. Bone loss is of particular significance as a bone fracture in this population has been associated with a significant decrease in length of survival. Toremifene citrate (Acapodene™), a selective estrogen receptor modulator (SERM), has the potential to ameliorate many of these side effects of ADT and to improve quality of life.

Methods: 46 consenting patients who had been on ADT for at least 12 months were randomized to 20 mg, 40 mg, or 60 mg of toremifene citrate or placebo administered orally once daily for 6 months. Assessments included bone mineral density (BMD) determination by dual energy X-ray absorptiometry (DEXA), bone turnover markers, and hot flashes.

Results: Toremifene was well tolerated at all doses studied. Based on our own observations and recent reports in the literature, the antiandrogen bicalutamide can affect bone metabolism. To avoid confounding effects from bicalutamide, patients on bicalutamide were excluded from the analysis. Administration of 60 mg/day of toremifene (but not 20 mg/day or 40 mg/day) resulted in an increase of BMD after 6 months of therapy compared to a decrease in the placebo group (p < 0.01 ), as well as beneficial effects on other side effects of ADT.

Conclusion: Toremifene administration resulted in a statistically significant increase in BMD after just 6 months of therapy in patients with prostate cancer, suggesting that toremifene, a SERM, is a promising agent to improve bone mineral density and quality of life in patients with prostate cancer receiving ADT. A long-term large scale trial is in progress to demonstrate the long term benefits of toremifene in prostate cancer patients receiving chronic ADT.

4598

Previous History of Bladder Cancer as a Prognostic Factor for Patients Diagnosed with Transitional Cell Carcinoma of the Upper Urinary Tract

M. Mullerad, D. Golijanin, H. N. Chen, S. Donat, B. Bochner, H. W. Herr, P. C. Sogani, P. Russo, M. W. Kattan, G. Dalbagni; Memorial Sloan-Kettering Cancer Center, New York, NY

Background: Upper tract transitional cell carcinoma (UTTCC) is a relatively rare tumor. Tumor stage, grade, carcinoma in situ and multifocality have been previously described as adverse prognostic factors. This study evaluate whether prior history of bladder TCC (BTCC) is a prognostic factor for patients diagnosed with UTTCC.

Methods: We report on 129 patients, median age 68 years, who underwent 129 nephroureterectomies for UTTCC from July 1989 to June 2002. Sixty seven patients had primary UTTCC and 62 had either previous (n=52) or synchronous (n=10) BTCC. Medical records were reviewed and analyzed for possible prognostic predictors (primary tumor stage, grade, multifocality, carcinoma in situ, symptoms and signs at presentation, gender, and history of smoking ). Disease specific survival and bladder recurrence free survival was analyzed by the Kaplan-Meier curves, and differences between the groups were compared by the log-rank test. The Cox Proportional Hazards regression model was used to assess the significance of each predictor.

Results: Disease specific death was reported for 44 patients. In a multivariate analysis, using previous BTCC as a predictor (categorized to superficial, invasive or none), both primary disease stage and history of BTCC (invasive and superficial) were significantly associated with disease specific survival (P=0.0013, P=0.0124). Kaplan Meier curves demonstrate that history of BTCC grouped the patients into distinct populations in terms of disease specific survival and bladder free recurrence (P< 0.05).

Conclusions: This study demonstrates that prior history of BTCC (either invasive or superficial) has an adverse effect on the prognosis of patients diagnosed with UTTCC independent of primary tumor stage.

4601

Intravesical Gemcitabine (G) Single Agent as Adjuvant Chemotherapy in Superficial Transitional Cell Carcinoma (TCC) of Bladder

K. Bouzid, R. Ferhat, A. Bounedjar, H. Mahfouf, F. Smaili, K. Adjali; EHS Pierre & Marie Curie Center, Algiers, Algeria; CAC Blida, Blida, Algeria; CHU Bab el Oued, Algiers, Algeria

Background: Intravesical G (2000 mg twice weekly) has been shown in a phase I study to be a safe regimen in bacillus Calmette-Guerin (BCG) refractory TCC (Dalbagni G et al. J Clin Oncol. 2002; 20: 3193-3198). The objectives of this trial were to evaluate the toxicity and efficacy of intravesical G as adjuvant chemotherapy in patients (pts) with superficial TCC of the bladder.

Methods: Chemonaive pts with histological diagnosis of TCC confirmed by total (ultrasonographic control) transurethral resection (TUR) received intravesical G 2000 mg/week diluted (inbuffered) in 100 ml of saline solution once weekly for 6 weeks. After Total TUR and three weeks of rest, they received 2000 mg G once a week for 6 weeks. Additional inclusion criteria were: age > 18 yrs, PS = 1, adequate hematologic, hepatic and renal function, and written informed consent. Toxicity was evaluated at every intravesical administartion. Efficacy was evaluated three months after the end of treatment by ultrasonography, cytology, and TUR with bladder biopsy.

Results: Twenty-nine pts (28 male, 1 female), recruited in two centers were enrolled in this study. There were 9 pts with carcinoma in situ and 20 pts with pT1 lesions. The median age was 60 yrs (range 24-83). After 348 total instillations (6 months of treatment), 29 pts were evaluable for toxicity in WHO scale. Nonhematological toxicities were grade 1: irritative bladder reaction (7,4%) , asthenia (4,8%) , nausea (1%) hot flashes (4%). Grade 1 hematologic toxicities were anemia (9%), leucopenia (8%) and thrombocytopenia ( 0,5% ) 2 pts. after a median follow-up time of 12 months ( range 3-15 months), all 29 pts were in complete remission without superficial bladder carcinoma-related events.

Conclusions: G as adjuvant intravesical chemotherapy in superficial TCC of the bladder is safe , with minimal toxicity even after repeated instillations. Preliminary results of efficacy appears to be favorable.

4603

Phase II Randomized Trial of Docetaxel Plus Estramustine (DE) Versus Docetaxel (D) in Patients (pts) with Hormone-refractory Prostate Cancer (HRPC): A Final Report

J.-C. Eymard, F. Joly, F. Priou, A. Zannetti, A. Ravaud, P. Kerbrat, M. Mousseau, B. Paule, F. Touze, E. Ecstein-Fraisse; Institut Jean Godinot, Reims, France; Centre Jean François Baclesse, Caen, France; Centre Hospitalier Départemental, La Roche/Yon, France; Clinique du Parc, Cholet, France; Institut Bergonié, Bordeaux, France; Centre Eugène Marquis, Rennes, France; CHU Michallon, Grenoble, France; Hospital Henri Mondor, Créteil, France; Laboratoire Aventis, Paris, France

Background: Single-agent docetaxel has proven efficacy in pts with HRPC, but the D and E combination is synergistic in vitro. This study evaluated PSA decline (PSA?), safety, and quality of life (QoL) for DE vs D.

Methods: Major eligibility criteria were: WHO PS < 2, appropriate renal, hepatic, and hematologic function, no prior chemotherapy, and withdrawal of antiandrogen therapy. Pts were randomized to Arm A (D 70 mg/m2 IV over 1 hour d1 q3w + E 560 mg/d PO starting 1 day prior to D, for 5 consecutive days) or Arm B (D 75 mg/m2 IV over 1 hour d1 q3w) for a maximum of 6 cycles. Prophylactic warfarin (1mg/d PO) was given continuously in Arm A. Corticosteroid was given before and after D infusion in both arms.

Results: Between 02/2001 and 10/2002, 92 pts were randomized (Arm A: 48, Arm B: 44) but 1 pt did not receive treatment. Median age was 68 years (range: 46-86), PS 0/1/2 (32/50/9 pts), median PSA was 115 ng/ml (range: 0.3-1585), and 40 pts (Arm A: 22, Arm B: 18) had measurable disease (MD). With a median number of 6 treatment cycles in both arms, cycle delays >7 days were more frequent in Arm A (15% pts) vs Arm B (11% pts); dose reduction was similar, 4.3% vs 4.5% pts, respectively. Median follow-up was 12.8 months. Main grade 3-4 hematologic toxicities among pts in Arms A vs B, respectively, were: neutropenia 25.5% vs 27.3%; anemia 10.6% vs 2.3%. Main grade 3-4 treatment toxicities (pts) were: thrombophlebitis (1) and allergic reaction (1) in Arm A; febrile neutropenia (1) per arm; fatal acute pulmonary edema (1) in Arm B. Response in Arm A vs B, respectively, was: PSA? >50%: 68% vs 29%; PSA? >75%: 36% vs 16%; median PSA response duration was 6 months in both arms. Of 40 pts with MD, PR was observed in 18.2% (Arm A) vs 16.7% (Arm B). Median time to progression in Arm A was 5.7 months (range: 4.7-5.8) vs 2.8 (2-6.9) in Arm B. There was no worsening in QoL using the FACT-P instrument and the pain score was stable throughout treatment in both arms.

Conclusion: Docetaxel-based regimens are active in HRPC with predictable and manageable toxicity profiles.

4611

Clinical and Biologic Activity of Soy Protein Powder (SPP) in Healthy Male Volunteers: Effect on Testosterone and Luteinizing Hormone (LH)

S. Goodin, F. Shen, W. Shih, N. Dave', M. Kane, G. Lambert, M. Gallo, R. Dipaola; Cancer Institute of New Jersey, New Brunswick, NJ; UMDNJ/Robert Wood Johnson Medical School, Piscataway, NJ

Background: Epidemiologic studies and a recent prospective trial have revealed a reduced risk of prostate cancer associated with consumption of soy foods, possibly due to the estrogenic properties of soy. To determine if a commonly used SPP can have estrogenic activity in men, we specifically evaluated supplementation with SPP in healthy male volunteers and the effect on testosterone and LH.

Methods: Healthy volunteers recruited to this study were men 18 years of age or older that were in good health with no known allergy to soy or phenylalanine. Treatment consisted of consuming two scoops (56g) of pure SPP (Puritan's Pride, Oakdale, New York) daily for 28 days. Serum testosterone and LH levels were collected prior to initiation of supplementation with soy protein powder, throughout the 28 days, and 14 days after the completion of therapy. Patients maintained a food diary throughout the 28-day period.

Results: Twelve healthy volunteers were enrolled with a mean age of 32 years (range 25 to 47). Serum testosterone concentrations decreased during the 4-week use of SPP and increased within two weeks after SPP was discontinued, with a statistically significant percent change in testosterone over the four-week period (P=0.021). At no time did the mean serum testosterone fall below the normal reference range. Serum LH concentrations decreased during the 4 week use of soy protein powder then increased within two weeks after soy protein powder was discontinued but the changes did not reach statistical significance (P =0.20). There were no side effects reported by any participants. Using a yeast based assay transfected with the alpha and beta estrogen receptor (ER), we found that the soy protein selectively activated ER-beta.

Conclusions: Soy protein powder decreases testosterone levels in healthy patients, which is reversible upon removal of SPP from the diet. These data support further study of these hormonal effects as a mechanism in prostate cancer prevention.

4612

Extended Pelvic Lymphadenectomy in Men Undergoing Radical Prostatectomy - An Update On > 300 Cases

A. Heidenreich, R. Von Knobloch, Z. Varga, R. Hofmann, U. Engelmann; University of Köln, Köln, Germany; Philipps - University, Marburg, Germany

Background: Lymph node dissection (LND) in PCA is considered as a mere staging procedure and limited to obturator and external iliac nodes. Since the prostate drains into the external iliac, hypogastric, lateral and subaortic sacral nodes, metastases might be missed by the standard approach. We performed a radical pelvic lymphadectomy (RLND) in pts with PCA to assess frequency and location of lymph node metastases and to evaluate the reliability of current prediction models for lymph node involvement.

Methods: 311 pts underwent RLND at time of radical retropubic prostatectomy via an extraperitoneal approach. RLND comprised 8 selective fields: external, internal and common iliac nodes, obturator nodes bilaterally. Nodes from each location and side were submitted separately for pathohistologic evaluation. Pathohistology of lymph nodes was compared to serum PSA, preoperative clinical stage and biopsy Gleason sum, pathohistological stage and Gleason sum of the prostatectomy specimen. Reliability of Partin tables, CART analysis and clinical parameters in terms of accurate prediction of lymph nodes metastases was evaluated.

Results: Lymph node metastases were diagnosed in 19.8% ; mean number of dissected lymph nodes was 28 (21-42). 22% of lymph node metastases were located in the internal and common iliac region only. 23/58 pts had 1 positive lymph node, 18 pts had 2, 7 pts had 3 and 8 pts had > 3 lymph nodes involved. Preoperative PSA serum levels were 26.8 (3.5-133)ng/ml and 16.5 (2.5-125)ng/ml in pts with and without lymph node metastases, resp. (p< 0.05). CART nomograms did not accurately predict lymph node metastases with 58% of pN+ having been understaged. 27% and 58% of pN+ pts had serum PSA levels < 10ng/ml and a biopsy Gleason sum < 7, resp. The Partin tables understaged 33.5% of pts with positive lymph nodes.

Conclusions: 47% of pts demonstrate positive lymph nodes around the internal and common iliac artery making this area essential for adequate staging LND. Current clinical staging systems (Partin, CART) miss up to 50% of all pN+ pts. Radical pelvic lymphadenectomy represents the standard lymphadenectomy technique of choice in surgical management of PCA.

4616

Phase II Trial of Docetaxel (D) and High-dose Celecoxib (C) in Patients (Pts) with Hormone Resistant Prostate Cancer (HRPC)

B. Kasimis, J. Cogswell, S. Hwang, V. Chang, M. Llorente, I. Boholli, S. Srinivas, E. Morales, C. Davis, M. Blumenfrucht; VA NJ Health Care System, East Orange, NJ

Background: COX-2 is an inducible enzyme, which converts arachidonic acid to prostaglandins. COX-2 is over expressed in 89.3% of our pts with prostate cancer and is an independent predictor of survival (Cogswell et al; Proc ASCO 2003, #1675). Both D and C are angiogenesis inhibitors and induce apoptosis; their combination could be synergistic. In an ongoing pilot trial we determined the effects on PSA, Overall Response(OR), toxicity, and survival.

Methods: Eighteen (18) pts with HRPC progressing by rising PSA and radiographic manifestations were treated with D 30mg/m2 IV over 30min,weekly for 3 wks and C 400mg po bid daily of each 4wk cycle. All pts were assessed by PSA and radiological tests every 2 cycles. Dose modifications for hematologic, hepatic and renal toxicity were made. Detailed renal function tests were assessed by a nephrologist. The RECIST criteria and PSA decline by >50% were used. Sample size was determined by Simon's Optimal Two-Stage design.

Results: The median age was 73.5 years(55-94), ECOG PS 1(0-1),LDH 180.5 U/L(125-899 ), Hgb 11.9g/dL (8.6-14.6), PSA 215.8 ng/dL (28.5-2675). The median number of cycles was 3(2-6).The median follow up was 6 mos (2-11+). Soft tissue metastases were present in 15 pts (83.3%) and bone metastases in 17 pts (94.4 %). Seventeen(17)pts are evaluable for response. Of these, 11 pts(64.7%) had PSA responses and the PSA normalized in 4pts (23.5%).Soft tissue responses were seen in 5(2CR's + 3PR's)pts(29.4%)and SD in 5pts(29.4%). One pt(5.9%) had an impressive bone scan response. The OR for 11pts was 64.7%. One pt withdrew due to abdominal pain, 2pts(11.7%) had grade III diarrhea and one had grade III nail changes. There was excellent renal tolerance (Yudd et al; Proc ASCO 2004 subm ).

Conclusions: The combination of D and C is well tolerated. The OR rate was 64.7%. PSA decline by>50%, PSA normalization, soft tissue , and bone scan responses were demonstrated. Further accrual and longer follow up will determine its full impact in pts with HRPC. (Supported by Aventis and Pfizer).

4621

Bone Alkaline Phosphatase Doubling Time and Outcome in Metastatic Hormone-refractory Prostate Cancer Patients.

D. J. Sleep, S. M. Hulting, J. D. Isaacson, A. R. Allen; Abbott Laboratories, Abbott Park, IL

Background: Bone alkaline phosphatase (BAP) has been validated as a potential predictor of survival in hormone-refractory prostate cancer (HRPC) patients. Phase 2 results indicate that atrasentan, a potent, selective, oral endothelin A receptor antagonist, significantly attenuates BAP rise in HRPC patients. We tested the hypothesis that BAP response to treatment is predictive of time to disease progression (TTP) and survival.

Methods: Data from a randomized, double-blind, placebo-controlled phase 3 trial of HRPC patients given either placebo or atrasentan 10 mg QD were combined. Values from 681 patients with at least 3 BAP values on treatment were analyzed. Data from treatment arms were pooled and stratified by the time over which BAP doubled (BAPDT) < 12 mo vs =12 mo. Kaplan-Meier methodology was used to compare TTP (a composite of radiographic and clinical measures) and survival between the BAPDT < 12 mo and =12 mo groups. In addition, frequency tables from the pooled data based on quartiles of time to BAP progression (TTBAP, =50% increase from nadir BAP value) vs quartiles of survival and TTP were analyzed, excluding subjects censored for both survival/TTP and TTBAP events.

Results: 304 patients had BAPDT < 12 mo; 377 had BAPDT =12 mo. Patients with BAPDT =12 mo had a significantly longer median TTP and a significantly longer time to death than those with BAPDT < 12 mo (table). Frequency tables of TTBAP vs TTP and TTBAP vs survival also indicate that early TTBAP was associated with a faster TTP and a shorter survival duration.

Conclusions: These data show that a BAPDT < 12 mo is predictive of early disease progression and reduced survival of HRPC patients, independent of therapy. This association warrants further exploration to determine whether BAP kinetics is a suitable surrogate endpoint for disease progression or survival.

4622

Randomized Phase IIb Trial Comparing Two Schedules of Docetaxel (D) Plus Estramustine (E) for Metastatic Hormone Refractory Prostate Cancer (HRPC)

R. Birch, L. Kalman, L. Holt, B. Graham, B. Wheeler, L. Schwartzberg; Online Community Oncology Group, Memphis, TN; Oncology Hematology Group of South Florida, Miami, FL; Coastal Cancer Center, Myrtle Beach, SC; Jackson Oncology Associate, Jackson, MS; The West Clinic, Memphis, TN

Background: Combination chemotherapy with D + E is an active treatment for HRPC. However, the optimal schedule of these agents is not well established. Since many patients have comorbid conditions which can increase toxicity from these drugs, use of weekly schedules may be preferable. Methods: Between 8/98 and 6/03, we randomized 62 chemotherapy naive HRPC patients to E 280 mg/m2 day 1-5 and day 22-26 PO with D 36 mg/m2 IV day 2, 9, 16, 23, 30, 37 q 8 wks (arm A) or E 280 mg/m2 day 1-5 PO with D, 70 mg/m2 IV day 2 q 3 weeks (arm B). Dexamethasone 16 mg IV was given as premed with D without routine oral steroids. Endpoints were PSA response rate (RR), clinical RR, Grade 3/4 hematologic and nonhematologic toxicity, dose delivery, TTP, and OS.

Results: Baseline characteristics were well balanced between arms A/B as follows: Median age 72/71.5, median PSA 99.9/110.3, ECOG 1/1. Median follow up is 33.4 months. PSA RR, defined as >50% reduction sustained for >4 weeks, was seen in 11/21 evaluable patients (52%) in arm A and 19/29 evaluable patients (66%) in arm B, p=0.35. The median overall survival was 16.4 months in arm A vs.13.2 months in arm B, p=0.42. Hematologic toxicity differed between arms. Grade 3/4 ANC toxicity was observed and G-CSF required for 4% of patients on arm A versus 40% of patients on arm B (p=0.0022). Median ANC nadir was 2.27 x 109/L (+ 1.3) in arm A vs. 1.55 x 109/L mm (+ 2.8) in arm B. Median nadir platelet count was 180 x 103 in arm A vs.167 x 103 in arm B and median nadir hemoglobin was 10.0 gm/dl in arm A vs.10.3 in arm B. Grade 3/4 nonhematologic toxicity was similar in both arms. Overall, 30% arm A vs.27% arm B had grade 3/4 toxicity (p=0.81), with neuropathy, nausea and vomiting and edema being most common. There were no deaths attributable to study drugs. The median number of 8 week cycles delivered on arm A was 3 (range 1-6) and a median of 5 (range 1-12) 3 week cycles were administered on Arm B. Dose was delayed or reduced in 48% of patients on arm A and 37% of patients on Arm B (p=0.38).

Conclusions: Both weekly and every-three-week D+E have substantial activity in HRPC. The therapeutic index of weekly D+E is better with less hematologic toxicity.

4627

Outcomes of Androgen Deprivation as Sole Therapy for Localized Prostate Cancer

D. Janoff, S. Mongoue-Tchokote, C. Peterson, R. O. Parra, L. Peters, Y. Park, T. M. Beer, M. Mori, M. Garzotto; Oregon Health & Science University, Portland, OR; Portland VA Medical Center, Portland, OR

Background: Recent studies indicate increasing use of androgen deprivation (AD) as primary therapy for localized prostate cancer. The utility of this approach is not well known; thus, we sought to characterize the outcomes and determine independent predictors of disease progression with this approach.

Methods: The records of all patients with cT1-2, N0, M0 adenocarcinoma of the prostate treated with AD as sole initial therapy at the Portland VA between 1993 and 2000 were reviewed. Age, race, Charlson health index (CHI), family history, PSA, PSA density (PSAD, PSA/prostate volume), digital rectal exam findings, Gleason score, and % of positive biopsy cores at diagnosis were recorded. Outcomes included PSA failure (confirmed with 2 rising PSAs), PSA nadir, bone fractures, local failure, distant failure, and overall survival. Kaplan-Meier method and log rank test (univariate analysis) and stepwise Cox regression (multivariate analysis) were used to estimate time to each end point.

Results: 81 patients with a median age of 73 (58 to 84), median CHI of 1, and median PSA of 13.8 ng/ml were followed for a median of 54 months (6 to 115). Rates of endpoints were: local failure = 9.9%, distant failure = 7.4%, bone fractures= 24.7% (18 osteopenic, 2 pathologic), PSA failure = 21.0%, and death = 40.7%. Factors associated with PSA failure by univariate analysis were % positive biopsy cores > 83% (p=0.0008), age < 70 (p= 0.006), Gleason score > 7 (p= 0.015), abnormal DRE (p= 0.025), and PSA nadir > 0.2 ng/ml (p=0.044). The multivariate analysis identified age < 70 (HR= 6.52, 95% CI = 2.29-18.55) and Gleason score > 6 (HR = 4.00, 95% CI = 2.00-12.00) as independent risk factors of PSA failure after AD therapy. Kaplan-Meier estimate of PSA failure at 5 years was 25%±6%. None of the variables were predictive of overall survival.

Conclusions: AD therapy resulted in reasonable control of localized prostate cancer; however, younger patients and those with Gleason > 6 cancers were at higher risk for treatment failure. Bone fracture were relatively frequent, perhaps due to prolonged treatment. The efficacy and toxicity of AD for localized prostate cancer requires further study before this modality can be recommended.

4634

Bicalutamide Combination Therapy Versus Castration Alone: A Combined Analysis of Historical Data

L. H. Klotz, P. F. Schellhammer, K. J. Carroll; Sunnybrook & Women's College Health Sciences Centre, Toronto, ON, Canada; Eastern Virginia Medical School, Norfolk, VA; AstraZeneca, Macclesfield, United Kingdom

Background: Meta-analyses of randomized trials using flutamide or nilutamide + castration compared with castration alone in D2 prostate cancer have shown a modest overall survival benefit. No direct comparisons have been undertaken of combination therapy using bicalutamide ('Casodex') 50 mg with castration alone; here we combine historical data to try to delineate the likely benefit of bicalutamide in this setting. This methodology has been well accepted in other oncology areas: eg in colon cancer, studies comparing the efficacy of 5-fluorouracil (5-FU) + leucovorin to 5-FU alone, and 5-FU + leucovorin to capecitabine, were integrated on the principle that equivalent groups were being compared in otherwise disparate studies (Stat Med 2003;22:239-64). This analysis resulted in FDA approval for capecitabine based on this approach.

Methods: A double-blind, randomized, multicenter trial comparing bicalutamide and flutamide combination therapy (Urology 1997;50:330-6) was combined with the Prostate Cancer Trialists' Collaborative Group (PCTCG) meta-analysis for flutamide + castration vs castration alone (Lancet 2000;355:1491-8). The HR was estimated by multiplying the combination therapy HR for bicalutamide vs flutamide by the HR for flutamide combination therapy vs castration alone. The statistical uncertainty was calculated as the sum of the variances in the comparison of bicalutamide vs flutamide combination therapy, and the effect of flutamide in the PCTCG meta-analysis. The key assumption is that the effect of flutamide in both populations was the same. The lack of heterogeneity in the effect of flutamide in the PCTCG analysis supports this assumption.

Results: The analysis showed that there is a 98.5% probability that bicalutamide, as combination therapy with an LHRHa, provides a survival advantage over castration alone. The estimated HR is 0.80 (95% CI 0.66, 0.98) (ie 20% risk reduction).

Conclusions: There is a high probability that bicalutamide (given once daily) + castration provides a survival advantage over castration alone. The estimated reduction in risk of death is 20%. 'Casodex' is a trademark of the AstraZeneca group of companies

4637

Changing Demographics and the Contemporary Surgical Management of Renal Cortical Tumors

P. Russo, M. E. Snyder, F. Rabbani, M. W. Kattan, R. Motzer, V. Reuter; Memorial Sloan Kettering Cancer Center, New York, NY

Background: To study the demographics and clinical outcomes in the surgical management of renal cortical tumors in a large 14-year contemporary single institutional experience.

Methods: All patients presenting to our center for the surgical management of renal cortical tumors between 7/89 and 7/03 were prospectively entered and followed in a comprehensive database. Demographic features studied include: clinical presentation, tumor size, stage, histologic sub-type, and operation performed. Differences between groups were analyzed using Kaplan Meier survival distributions and the log rank test. The 2002 UICC staging system was employed (P1a< 4cm, P1b< 7cm).

Results: 1548 patients with renal cortical tumors, including 78 patients (5.0%) with bilateral tumors, were studied. 63.6% of patients were male, 36.4% female, 88.0% white, 5.0% black, and 3.1% Hispanic. The median age is 63 (15-94) and mean follow-up is 46.2 months (0.1-164.9). Significant trends were observed in both modes of incidental presentation (Year 1: 55.2%; Year 14: 72.8%, p< 0.01) and median tumor size (Year 1: 6.0 cm; Year 14: 3.5 cm, p< 0.001). The percentage of patients with P1a (2002 UICC staging) tumors rose from 20.9% in Year 1 to 53.9% in Year 14 (p< 0.01) while patients presenting with M1 or N1 disease declined from 17.9% in Year 1 to 4.4% in Year 14 (p< 0.01). Partial nephrectomy increased from 14.9% in Year 1 to 58.3% of nephrectomies in Year 14 (p< 0.01). Progression Free Probability is 90.1% at 3 years and 87.5% at 5 years. Histologic subtype was the most critical determinant of prognosis with 5-year Progression Free Probability of 84.7% for conventional clear cell, 92.6% for papillary, 90.0% for chromophobe, and 100% oncocytic (p=0.02).

Conclusions: Incidental tumor detection, smaller primary tumors, appreciation of histologic subtype, and increased application of kidney-sparing surgery are the current characteristics of the surgical management of renal cortical tumors at our center.

4645

Residual Tumor Resection (RTR) in Advanced Testicular Cancer: Indication, Outcome and Prediction of Histology

M. Seger, M. Roessler, J. Beyer, A. Heidenreich; Philipps - University, Marburg, Germany; Division of Oncological Urology, Köln, Germany

Background: RTR following inductive chemotherapy for advanced testicular cancer is an essential part of the interdisciplinary multimodality treatment. Since RTR is a challenging surgical procedure, aim of our study was to develop a clinically reliable model for identification of a low risk group in which RTR could be safely omitted.

Methods: RTR was performed in 85 pts after first-line chemotherapy with all pts showing normalized tumor markers. In a retrospective analysis > 15 pre-RTR parameters including immunohistochemical expression of mdr-1 in the orchiectomy specimens were assessed to predict necrosis/fibrosis or teratoma in residual tumors. Statistical analysis was performed by uni- and multivariate analysis, parameters were statistically significant if p< 0.05.

Results: 56 pts (67.5%) had necrosis/fibrosis, 17 (20.5%) and 10 pts (12%) had mature teratoma and vital cancer in the RTR specimens, resp. After multivariate analysis only pre-chemotherapeutic AFP levels and tumor size before RTR were independent predictors of necrosis. However, positive predictive values were only 74% and 76%, resp., sensitivity was 81% and 52.8%, resp., specificity was 59% and 79.3%, resp. Of pts with < 2cm residual masses 6 had vital cancer /mature teratoma, of pts with < 15ng/ml AFP 9 had vital cancer/mature teratoma. Mature teratoma in the orchiectomy specimen was associated with a 86% risk for teratoma in the residual mass whereas non-teratomatous components in the primary cancer had a 47% risk to harbour teratoma in the RTR specimen p< 0.001). High mdr-1 expression in the orchiectomy specimen correlated with vital cancer/teratoma in the RTR specimen or with recurrences in pts with necrosis in the RTR specimen.

Conclusions: Although pre-chemotherapy AFP levels < 15ng/ml and a residual mass < 2cm are associated with a favourable histology in the RTR specimen, an individual histology cannot be predicted. All pts with nonseminomatous germ cell tumors and residual masses need to undergo RTR. Mdr-1 expression in the orchiectomy specimen might turn out to be a valid clinical marker for unfavourable histology or early recurrences.

4651

Screening with Prostate Specific Antigen and Metastatic Prostate Cancer Risk: A Population-based Case-control Study

J. A. Kopec, V. Goel, P. S. Bunting, J. Neuman, E. C. Sayre, P. Warde, P. E. Levers, N. Fleshner; University of British Columbia, Vancouver, BC, Canada; University of Toronto, Toronto, ON, Canada; University of Ottawa, Ottawa, ON, Canada; Mt Sinai Hospital, Toronto, ON, Canada; Arthritis Research Centre of Canada, Vancouver, BC, Canada; Southlake Regional Health Centre, Newmarket, ON, Canada

Background: Screening of asymptomatic men with the prostate specific antigen (PSA) is a highly controversial issue. There is little evidence to date that such screening is effective in reducing mortality from prostate cancer. We conducted a population-based case-control study to determine if screening with PSA reduces the risk of metastatic prostate cancer.

Methods: The study was carried out in Metropolitan Toronto and five surrounding counties in Ontario, Canada. Cases were men diagnosed with metastatic prostate cancer between August 1, 1999 and May 31, 2002. Controls were randomly sampled from the source population and matched to cases for age, area of residence, and observation time. Medical records and a self-administered questionnaire were used to collect the data. We obtained information on PSA testing, digital rectal exams, symptoms, comorbidity, use of health services, as well as known and suspected risk factors for prostate cancer. Analyses were performed on 236 cases and 462 controls.

Results: After adjustment for observation time, the frequency of screening among asymptomatic men was 24.6% in the cases and 33.1% in the controls (odds ratio 0.66, 95% confidence interval 0.47 to 0.92). Self-reported screening was also significantly less frequent among the cases (odds ratio 0.48, 0.34 to 0.67). No significant confounding was found in multiple logistic regression analyses. Secondary analyses including symptomatic and asymptomatic men showed that the cases generally had fewer symptoms and fewer PSA tests than the controls, except for the last year before the diagnosis of prostate cancer. In symptomatic men, the association between PSA testing and metastatic prostate cancer was not significant.

Conclusions: The results suggest that screening of asymptomatic men with PSA may substantially reduce the risk of metastatic prostate cancer. Further research is needed to confirm these results in a randomized trial and to assess the economic and quality of life aspects of PSA screening.

4652

Is Prostate-specific Antigen Progression a Surrogate for Objective Clinical Progression in Early Prostate Cancer?

D. Newling, K. Carroll, T. Morris; AstraZeneca, Macclesfield, United Kingdom

Background: It is a common misconception that a correlation between endpoints is enough to demonstrate surrogacy. To show true surrogacy, the effect of an intervention on an intermediate endpoint relative to a control treatment needs to reliably predict the effect of the intervention on the clinical outcome of interest. Valid surrogate endpoints are needed to accelerate availability of information about new therapies for early prostate cancer where clinical progression and survival times are prolonged. The usefulness of prostate-specific antigen (PSA) progression as a surrogate for objective clinical progression has therefore been assessed in the bicalutamide ('Casodex') Early Prostate Cancer (EPC) Program, the world's largest prostate cancer treatment program.

Methods: Individual data from all 8113 patients across 21 countries in the EPC Program were examined. Time to PSA progression and time to objective clinical progression were the endpoints evaluated. The effect of treatment on time to PSA progression was compared with the effect of treatment on time to objective clinical progression.

Results: Analyses suggest that time to PSA progression is a modest surrogate endpoint for the effect of a hormonal intervention on objectively confirmed progression in patients with early prostate cancer (r2 = 0.52-0.65, p< 0.001). An intervention that produces around a 50% reduction in the risk of PSA progression would likely result in around a 20% reduction in the risk of objective clinical progression.

Conclusions: The effect of treatment on PSA progression is moderately predictive for the effect of hormonal treatment on objective clinical progression. A large effect on PSA progression predicts for a smaller, but nonetheless clinically important effect on objective clinical progression. These data suggest that a large positive treatment effect on time to PSA progression is reasonably likely to reflect a clinically important delay in objective clinical progression, making PSA progression a valid endpoint for the evaluation of hormonal medicines in early prostate cancer. 'Casodex' is a trademark of the AstraZeneca group of companies

4653

Nonrandomized Single Institutional Comparison of Radical Prostatectomy, External Beam Radiotherapy, or Permanent Seed Implantation in the Management of Localized Prostate Cancer

P. Kupelian, J. Ciezki, A. Mahadevan, A. Reuther, C. Reddy, E. Klein; MD Anderson Orlando, Orlando, FL; Cleveland Clinic, Cleveland, OH

Background: We review biochemical relapse-free survival (bRFS) rates after 125I permanent seed implantation (PI), external beam radiotherapy (EBRT), or radical prostatectomy (RP) for clinical stage T1-T3 prostate cancer in a large single center cohort.

Methods: The study population was selected from 3074 consecutively treated patients with clinically localized disease treated between 1996-2001. Biochemical failure was defined as a confirmed PSA >0.2 ng/ml after RP and 3 consecutive rising PSA levels (ASTRO consensus definition) after EBRT or PI. RP patients receiving adjuvant radiotherapy and all patients receiving neoadjuvant or adjuvant androgen deprivation were excluded from analysis. To eliminate the time dependency of the ASTRO definition for failure, RP patients with < 24 month follow-up and EBRT and PI patients with < 48 months of follow-up were also excluded, leaving 1056 patients in the study cohort. Clinical characteristics of the study population are summarized in the table.

Results: The 5-year bRFS rates for PI vs EBRT vs RP were 86%, 81%, and 81%, respectively (p=0.56). Multivariate analysis identified PSA (p< 0.001) and biopsy Gleason sum (p< 0.001) as independent predictors of relapse, while treatment modality (p=0.42), clinical T stage (p=0.10), age (p=0.56), and race (p=0.11) were not predictors of failure.

Conclusions: Even when allowing for longer follow-up times for patients treated with either EBRT or PI to manifest biochemical failure, bRFS rates were similar between PI, EBRT and RP as monotherapy for clinically localized prostate cancer in this contemporary cohort. Outcome in these predominantly low risk patients is determined by pretreatment PSA levels and biopsy Gleason score.

4673

The Prognostic Importance of Biopsy Proven Unilateral Versus Bilateral Disease in Patients Treated with I-125 Seed Implant for Localized Prostate Cancer: A Single Institution Retrospective Study.

N. Dunlap, R. Shirazi, A. Naghavi, W. Barrett; University of Cincinnati, Department of Rad/Onc, Cincinnati, OH

Background: The prognostic value of prostate disease laterality has rarely been assessed. The objective of this retrospective study was to determine the significance of biopsy proven unilateral versus bilateral prostate cancer as a prognostic indicator of biochemical control in patients treated with I-125 seed implant for localized prostate cancer.

Methods: Between 1996 and 2001, 159 consecutive patients with clinically localized prostate cancer were treated with I-125 seed implant at our institution. As part of the initial workup, all patients underwent bilateral prostate biopsies, pretreatment PSA, and Gleason score determination. Based on biopsy results, the patients were divided into two groups according to positive biopsies from one versus both lobes of the prostate. Biochemical control was assessed according to the ASTRO consensus definition of biochemical failure and by determining the number of patients achieving and maintaining a PSA nadir of < 1.0, < 0.5, or < 0.2.

Results: Mean pretreatment PSA level for patients exhibiting unilateral disease was 7.32 versus 8.20 in patients with bilateral disease (p=.1151). Mean Gleason score was 6.03 for patients with unilateral disease versus 6.27 for patients with bilateral disease (p=.0778). Biochemical control according to the ASTRO consensus definition was achieved in 91% of the patients with unilateral disease compared to 77% of patients with bilateral disease (p=.0069). PSA nadirs of < 1.0, < 0.5, or < 0.2 were achieved in 94% versus 79% (p=.0026), 89% versus 68% (p=.0006) and 72% versus 53% (p=.0096) of patients with unilateral disease versus bilateral disease respectively.

Conclusions: Unilateral versus bilateral disease is an important prognosticator for biochemical control in patients with early stage localized prostate cancer treated with I-125 seed implant. Patients with positive biopsies from both lobes of the prostate have worse prognoses compared to patients with only unilateral biopsy proven disease.

4674

Neoadjuvant Hormonal Therapy Plus Brachytherapy vs. Brachytherapy Alone for Patients with Prostate Cancer: A Single Institution Retrospective Study.

R. Shirazi, M. V. Mishra, A. Kleimeyer, W. M. Kassing, W. L. Barrett; University of Cincinnati/Barrett Cancer Center, Cincinnati, OH

Background: Neoadjuvant hormonal therapy(NHT) has sometimes been used in patients with prostate cancer opting for brachytherapy. We compared the biochemical response to NHT plus prostate brachytherapy(PBT) versus PBT alone in a consecutive series of patients treated at our institution.

Methods: From 1996-2000, 125 consecutive patients with adenocarcinoma of the prostate with clinical stage T1c-T2a,b; Gleason score (GS) 3-8; PSA of 1.1 - 30 were treated either with NHT plus I-125 seed implants (Group1; n=36) vs. seed implant alone (Group2; n=89). None of these patients recieved external beam radiation. The indications for hormone therapy were most often intermediate risk disease (GS 7-8 or PSA 10-30) or significant prostatic enlargement. Prescribed radiation dose was 16,000 cGy to the periphery of the prostate. NHT was a 3 or 4 month leuprolide injection before PBT. Patients were evaluated for biochemical recurrence according to the ASTRO consensus statement, and for achieving and maintaining a PSA nadir =1.0, =0.5 or =0.2. The mean follow up was 43 months (Group1= 43 vs. Group2= 44), with a range of 24-80 months.

Results: please see the summary in table below.

Conclusions: There was no significant difference appreciated in biochemical failure rate or PSA nadir between NHT plus seed implant vs. seed implant alone. The only significant difference noted between the groups was the mean initial PSA being higher in the group who received NHT. A beneficial effect of NHT in reducing the volume of disease and improving the prognosis is possible.

4680

Effect of Zoledronic Acid on Clinically Meaningful Changes in Pain Associated with Metastatic Prostate Cancer.

K. P. Weinfurt, K. J. Anstrom, L. D. Castel, J. Brandman, K. A. Schulman; Duke University Medical Center, Durham, NC; Novartis Pharmaceuticals Corporation, East Hanover, NJ

Background: Zoledronic acid (ZA) has been shown to significantly decrease skeletal-related events (SREs) in patients with prostate cancer by 11% relative to placebo. Trends were seen in pain improvement with ZA vs placebo, but there were no reliable treatment group differences. The objective of this study was to reexamine the pain outcomes using an alternative analytic framework that focuses on the frequency of clinically meaningful changes in pain scores.

Methods: Patients were randomized to ZA 4mg (n=214) or placebo (n=208). The Brief Pain Inventory Composite Score (BPI) was used to assess pain at baseline, 3 weeks, 6 weeks, and every 6 weeks for a total of 60 weeks. We determined whether clinically meaningful pain changes from baseline (+/- 2 BPI points) were related to treatment assignment using a modified non-parametric rank test.

Results: The rapid progression of the prostate cancer led to high rates of dropout as only 76 of 214 (35.5%) patients randomized to ZA and 62 of 208 (29.8%) patients randomized to PL completed their 60 week visit (p=0.21). At 10 of 11 assessment times, ZA patients reported more favorable clinically meaningful changes in pain scores (see table). Overall, a typical ZA patient had a 27% chance of having a favorable response (ZA>PL) compared with a typical PL patient who had an 19% chance of a favorable response (PL>ZA)-a difference that was statistically significant (95% CI: 1.1% - 15.5%).

Conclusions: Patients receiving ZA were 8% more likely to have clinically beneficial changes in pain scores compared to patients receiving placebo. This is roughly consistent with the 11% difference between groups in the incidence of SREs as reported in the primary clinical analysis. Thus, ZA 4 mg significantly reduces the occurrence of skeletal-related events and the associated changes in pain they produce.

4687

Benefit of Atrasentan in Men with Hormone Refractory Prostate Cancer Metastatic to Bone.

A. Lipton, D. J. Sleep, S. M. Hulting, J. D. Isaacson, A. R. Allen; Pennsylvania State U, Milton S. Hershey Medical Center, Hershey, PA; Abbott Laboratories, Abbott Park, IL

Background: Bone alkaline phosphatase (BAP) predicts time to progression (TTP) and survival in patients with hormone refractory prostate cancer (HRPC). We tested 2 hypotheses: a) BAP at baseline is predictive of both TTP and survival in HRPC patients with bony metastases only and b) BAP predicts response to atrasentan therapy in these men.

Methods: Data were analyzed from 474 men with HRPC metastastic only to bone from a randomized, double-blind, placebo-controlled phase 3 trial of atrasentan 10 mg QD. Kaplan-Meier and Cox proportional hazard methodologies were used to compare TTP (a composite of radiological and clinical outcomes) between treatments. Values from 451 patients with baseline BAP values were analyzed. Data from treatment arms were pooled and stratified by baseline BAP < and >/= the median value. TTP and survival and the effect of treatment within the 2 groups were compared using Kaplan-Meier methodology.

Results: 253 patients received atrasentan, 221 received placebo; atrasentan resulted in a significant delay in TTP versus placebo (p=0.026). The median baseline BAP value was 31.5 ng/mL. Median TTP for patients whose baseline BAP < median was 168 days versus 85 days for patients with baseline BAP >/= median (p<0.001) and median time to death was 741 days vs 461 days for patients with BAP < median and >/= median, respectively (p<0.001). In patients with a baseline BAP > /= median, atrasentan significantly delayed TTP versus placebo, while in patients with baseline BAP < median, treatment effect did not reach significance (table).

Conclusions: These data confirm that high baseline BAP is predictive of early disease progression and reduced survival in HRPC patients with only bone metastases, independent of therapy. Atrasentan therapy significantly delayed TTP versus placebo in subjects with bone metastases only. It also significantly increased TTP in bony metastatic HRPC patients with a high baseline BAP.

4698

ZD1839 (Gefitinib) and Hormone Resistant (HR) Prostate Cancer - Final Results of a Double Blind Randomized Placebo-Controlled Phase II Study.

F. H. Schröder, M. F. Wildhagen, Uro-Oncology Group Rotterdam; Erasmus MC and University, Rotterdam, Netherlands

Background: The epidermal growth factor receptor (EGFR) is expressed in most prostate cancers. We investigated the effect of gefitinib (ZD1839), an orally active EGFR tyrosine kinase inhibitor, on the slope of changing prostate specific antigen (PSA) levels.

Methods: We conducted a placebo-controlled, double blind, randomized Phase II study of gefitinib (500 mg/day) in patients with prostate cancer and hormone-independent PSA progression. To calculate sample size, the method of Schlesselman (J Chron Dis 1973; 26: 561-70) was applied; 24 patients per group were necessary to detect a change in slope of 0.25 (ie to zero slope) with 90% power using a two-sided significance level of alpha = 0.05. Asymptomatic patients with PSA progression after castration or under luteinizing hormone-releasing hormone treatment (excluding antiandrogens and recent radiotherapy) were eligible. Unblinding took place after 6 months or in the case of medical emergencies. Secondary endpoints included PSA response rate at 6 months, time to response, PSA progression, time to progression (TTP), overall survival (OS), and safety. The study was approved by regional and institutional ethical review boards.

Results: All 58 patients randomized (28 to placebo and 30 to gefitinib) were evaluable for comparison of PSA slopes. Both slopes differed significantly from zero (placebo: p< 0.0001; gefitinib: p< 0.0023). The slopes amounted to 0.2569 for placebo and 0.2001 for gefitinib per month, translating into doubling times of 3.9 and 5.0 months (p=0.25), respectively. This study was not powered to compare changes in slope between treatment arms however, no relevant differences in progression rates, TTP, and OS were seen between arms in intention-to-treat and treatment-received analysis. Adverse events (AEs) were usually mild; CTC grade 3/4 AEs occurred in 2 patients receiving placebo and 5 receiving gefitinib (liver toxicity, diarrhea).

Conclusions: Gefitinib showed modest activity in delaying PSA rise. Differences in clinical progression rates were not seen. Gefitinib was generally well tolerated.

4711

Limited Success of Durasphere for the Treatment of Urinary Incontinence Following Radical Prostatectomy (RP)

F. P. Secin, K. S. Eilber; Memorial Sloan Kettering Cancer Center, New York, NY

Background: Durasphere is a permanent bulking agent used for the treatment of stress urinary incontinence (SUI). To our knowledge there is no data available regarding its efficacy for the treatment of SUI after RP. The purpose of this study is to describe our initial experience with Durasphere in the treatment of SUI following RP.

Methods: The medical records of eight patients who were treated with Durasphere for SUI after RP between 1/1/03 to 1/10/03 were retrospectively reviewed according to IRB guidelines. Patient age, preoperative urinary continence, time from RP to treatment, number of pads, operative reports and responses to treatment were reviewed.

Results: Eight patients aged 50 to 71 (mean age 63.2) with SUI after RP were treated with Durasphere. The average time from RP to treatment was 4 years (range 1.3 to 11). No patient had urinary incontinence preoperatively. The average number of pads used preoperatively was 3 (range 1 to 6). The median number of syringes of Durasphere used intraoperatively was 5 (range 4 to 7). Optimal coaptation following injection of Durasphere was obtained in only three patients. At a mean follow-up of 5.3 months (range 2 to 10 months), no patient was completely dry. Five patients (62.5%) opted for other surigcal treatment of their incontinence The remaining three patients were satisfied with the postoperative result. Two patients decreased daily pad use from 6 to 3 while the third patient continued to use 1 pad per day.

Conclusions: Durasphere does not consistently provide successful outcomes for SUI after RP. Its poor success rate is most likely secondary to decreased tissue compliance of the bladder neck after RP which limits its bulking ability.

4718

Biochemical Disease-free Survival and Toxicity Associated with Permanent Prostate Brachytherapy for Men < 60 Years of Age

R. J. Patel, D. J. Perry, J. Maier, J. A. Fugazzi, C. Zuniga, A. J. Frazier; Wayne State University / Detroit Medical Center, Detroit, MI

Background: To evaluate the biochemical disease-free survival (bDFS) and GI/GU toxicity for men 60 years of age or less who underwent transperineal ultrasound-guided I-125 permanent prostate brachytherapy (PPB).

Methods: A total of 71 patients, age < 60 underwent transperineal ultrasound guided PPB using I-125 for clinical stage T1c - T2a, Nx, M0 (2002 AJCC), Gleason score 6, adenocarcinoma of the prostate from November 1995 to December 2002. No patients received external beam radiation therapy. 18 patients (25%) received neoadjuvant hormonal manipulation with a median duration of 5 months (range, 2-10). The median patient age was 57 years (range, 42-60). 77% (55/71) reported erectile function sufficient for vaginal penetration prior to therapy. Patients were followed for PSA response/relapse, and GI/GU morbidity. Median follow-up was 34 months (range, 7 - 81).

Results: The actuarial bDFS rate was 93.0 % at 5 years. The median post-treatment prostate-specific antigen (PSA) nadir was 0.5 ng/mL. A total of 7% (5/71) developed acute urinary retention, and 3% (2/71) reported rectal bleeding in follow-up. One patient required a TURP as a complication for obstruction, and developed chronic incontinence thereafter. Chronic incontinence was not reported by any other patients. Of the patients potent prior to treatment, 65% (36/55) reported potency without the aid of medications, and 20% (11/55) with the assistance of medication, at the date of last follow-up. The remainder 15% (8/55) reported impotence, or no interest regaining erectile function.

Conclusions: Early stage, good prognosis, prostate cancer patients < 60 years of age have a high probability of bDFS following I-125 PPB. Risk of incontinence and rectal injury is low. Although potency is affected by therapy, the overall risk of erectile dysfunction is low (15%) when medical therapy is initiated.

4760

Effect of Zoledronic Acid (ZA) Versus Placebo on Bone Mineral Density (BMD) in Androgen-deprived Prostate Cancer Patients with and without Bone Metastases (mets)

C. W. Ryan, D. Huo, M. Hanson, W. M. Stadler, N. J. Vogelzang; Oregon Health & Science University, Portland, OR; University of Chicago, Chicago, IL

Background: Bisphosphonates can prevent BMD loss in prostate cancer patients receiving androgen deprivation. Previous studies of ZA and BMD in this population excluded patients with bone mets because of concern that mets may interfere with DEXA results.

Methods: We performed a double-blind, placebo-controlled, randomized study of ZA in prostate cancer patients with and without bone mets on androgen deprivation for < 12 months. Patients were randomized to receive either 4 mg ZA or normal saline placebo IV every 3 months for 4 treatments. BMD of the femoral neck (FN) and lumbar spine (LS) were measured by DEXA at 0, 3, 6, 9, and 12 months.

Results: From 1/00 through 12/02, 42 of a planned 80 patients were randomized. Accrual was stopped due to initiation of a multi-center study that is currently ongoing. BMD data are available for 18 patients who received placebo and 21 who received ZA. Baseline characteristics were well-balanced between the 2 groups (placebo vs. ZA): mean FN BMD 0.998 g/cm2 (SD 0.126) vs. 0.979 g/cm2 (SD 0.196); mean LS BMD 1.385 g/cm2 (SD 0.274) vs. 1.344 g/cm2 (SD 0.207). 9 patients in each arm had bone mets detected by bone scan. A random effect model was used to examine change in BMD over one year. The mean change in BMD per year (slope, g/cm2/year) is shown in the table below. Results are shown for all patients and for patients excluded based on bone scan abnormalities detected in the region of BMD measurement.

Conclusions: When all patients were analyzed, no statistically significant difference in the mean change per year of BMD was detected. When patients with scintigraphic evidence of bone mets at the site of DEXA measurement were excluded, the expected benefit of ZA therapy became apparent. Mets detected in the region of DEXA measurement can interfere with interpretation of ZA effect. Bone marker analysis for this study is pending.

*excluding 1 outlier with BMD that increased 50% from baseline

4767

Intermediate Risk Localized Prostate Cancer: Which Radiotherapeutic Approach?

V. Gondi, I. Deutsch, D. Vecchio, E. Goluboff, R. D. Ennis; Columbia University Medical Center, New York, NY

Background: Standard dose ( < =70 Gy) external beam radiotherapy (SD-EBRT) is known to be modestly effective treatment for intermediate risk prostate cancer. Over the last decade, several approaches have emerged to improved on those results. To address the uncertainty of which of these treatments is most effective, we performed a retrospective cohort study of such patients treated at our institution.

Methods: Analyzed all intermediate-risk patients as defined by D'Amico. These 346 patients had received either SD-EBRT (N=141), SD-EBRT + androgen deprivation (AD) (N=84; AD 3-40 mos., median 6), high-dose (HD) EBRT (>=72 Gy) +/- AD (N=49, AD 3-24 mos., median 4.5), brachytherapy (BRACHY) (N=10), BRACHY+EBRT (N=39), BRACHY+AD (N=11, AD 1-13 mos., median 3.5), and BRACHY+EBRT+AD (N=12, AD 3-9 mos., median 5). Outcome was biochemical disease-free survival (BDFS) as defined by ASTRO, estimated using Kaplan Meyer method and compared using log rank statistic. Because of uncertainty regarding the appropriateness of the ASTRO definition in AD patients, a second analysis using the definition of PSA > 1.5 ng/ml (as per Bolla et al.) was used. Multivariate analysis: Cox regression analysis. To adjust for stage migration over the study period, years of treatment were grouped and included as a variable in analysis.

Results: Using ASTRO: Within SD-EBRT and HD-EBRT outcome did not differ +/- AD. Outcome did not differ between different BRACHY-based treatments. BRACHY-based (p=0.0003) is superior to SD-EBRT +/- AD. HD-EBRT +/- AD was borderline significantly better (p=0.07). Multivariate, including year of treatment, confirmed BRACHY-based (p=0.01, RR 3.1) to be superior to SD-EBRT +/- AD. HD-EBRT +/- AD was not (p=0.4). Using PSA > 1.5 definition: SD-EBRT+AD (p=0.0001), BRACHY-based (p < 0.00009), HD-EBRT +/- AD (p=0.01) were superior to SD-EBRT. Multivariate: SD-EBRT+AD (p=0.01, RR 2.6), BRACHY-based (p=0.01, RR 4.5) remained superior to SD-EBRT.

Conclusions: Brachytherapy-based treatments and possibly the addition of AD to SD-EBRT improve the outcome of intermediate-risk prostate cancer over SD-EBRT. Determination of BDFS of AD treatments is definition dependent.

Please log-in or register in order to submit comments.

Powered by AkoComment!