Prostate-Specific Antigen Nadir and Cancer-Specific Mortality Following Hormonal Therapy for Prostate-Specific Antigen Failure
http://www.100md.com
《临床肿瘤学》
the Brigham and Women's Hospital, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA
Memorial Sloan-Kettering Cancer Center, New York, NY
University of Connecticut, Storrs, CT
Walter Reed Hospital, Bethesda, MD
University of California San Francisco, San Francisco, CA
Duke University, Durham, NC
ABSTRACT
PURPOSE: For men receiving androgen-suppression therapy (AST) for a rising postoperative or postradiation prostate-specific antigen (PSA), we evaluated whether a PSA nadir of more than 0.2 ng/mL was significantly associated with prostate cancer–specific mortality (PCSM).
PATIENTS AND METHODS: The study cohort comprised 747 men with rising PSA and negative bone scan after surgery (n = 486) or radiation therapy (n = 261) who were treated with AST. Cox regression was used to evaluate whether a significant association existed between the PSA nadir level after 8 months of AST and the time to PCSM, controlling for treatment and known prognostic factors.
RESULTS: The post-AST PSA nadir (pCox < .0001), the pre-AST PSA doubling time (DT) (pCox = .002), PSA level (P = .0001), and Gleason eight to 10 cancers (pCox = .01) were significantly associated with time to PCSM. The adjusted hazard ratio for PCSM was 20 (95% CI, 7 to 61; pCox < .0001), for men with a PSA nadir of more than 0.2 ng/mL as compared with all others. A PSA DT of less than 3 months was observed in 30% (224 of 747) of the study cohort. Of the 28 observed prostate cancer deaths, 21 (75%) occurred in men whose PSA nadir was more than 0.2 ng/mL and who had a PSA DT of less than 3 months.
CONCLUSION: A PSA nadir of more than 0.2 ng/mL after 8 months of AST given for postoperative or postradiation PSA failure is significantly associated with PCSM and is clinically significant because it accounted for 75% of the cancer deaths observed in this study.
INTRODUCTION
Prostate-specific antigen (PSA) failure after radical prostatectomy (RP) or external-beam radiation therapy (EBRT) for patients with clinically localized prostate cancer occurs in up to 30% of patients within 10 years after treatment.1,2 Yet, not all patients with PSA failure after RP or EBRT will die from prostate cancer. Specifically, for patients with a PSA doubling time (DT) of less than 3 months, the median overall survival from the time of PSA relapse is relatively short (approximately 6 years), whereas patients with a PSA DT of more than 12 months have a median survival that is greater than a decade.3
Although the value of the PSA DT before the institution of hormonal therapy is closely associated with prostate cancer–specific mortality (PCSM),3 the post–androgen-suppression therapy (AST) PSA nadir has also been shown to be associated with PCSM.4,5 In particular, in the setting of localized disease treated with primary AST, studies have shown that patients whose PSA does not have a nadir below 0.2 ng/mL are at a significantly higher risk of death from prostate cancer.4,5
Whether PSA nadir after AST given for PSA failure after primary local therapy provides clinically important information regarding the subsequent time to PCSM has not been studied. Therefore, we evaluated whether a PSA nadir level of more than 0.2 ng/mL compared with 0.2 ng/mL was significantly associated with PCSM for men receiving AST for a rising postoperative or postradiation PSA state.
PATIENTS AND METHODS
Patient Selection, Staging, and Treatment
The study cohort consisted of patients from two multi-institutional databases, Cancer of the Prostate Strategic Urologic Research Endeavor6 and the Center for Prostate Disease Research,7 which contain baseline, treatment, and follow-up information. The patients were treated with RP or EBRT between January 1, 1988, and January 1, 2002, for clinical stage T1C-4, NX or N0, M0 prostate cancer.8 All patients who received neoadjuvant or adjuvant AST were excluded, including 12 and 26 men treated with RP and EBRT, respectively. This left 747 men with a rising PSA after RP (n = 486) or EBRT (n = 261) and negative bone scan obtained within 1 month of the initiation of AST for study. An approved and signed internal review board informed consent form was obtained for each patient before study entry. Ten RP-managed patients (2%) did not achieve an undetectable PSA (median, 0.1 ng/mL; range, 0.1 to 0.2 ng/mL). The median PSA nadir after EBRT was 0.8 ng/mL (range, 0.1 to 2.8 ng/mL). Table 1 lists the patient characteristics of the study cohort.
At the time of PSA relapse, defined by using the American Society for Therapeutic Radiology and Oncology consensus definition for all study patients,9 all patients underwent a history and physical examination including a digital rectal examination, bone scan, and computed tomography scan of the pelvis. All patients with radiographically visible metastases were excluded from the analysis.
The AST treatments included orchiectomy (30 patients [4%]), luteinizing hormone-releasing hormone (LHRH) agonists (456 patients [61%]), antiandrogens (15 patients [2%]), and an LHRH agonist with an antiandrogen (246 patients [33%]).
Follow-Up
The median overall follow-up for the study cohort was 2.2 years (range, 0.8 to 9.5 years), defining 8 months after the initiation of AST as time zero. No patient died during the initial 8 months of AST. At each follow-up, a PSA level was obtained, and history and physical examination including digital rectal examination were performed. Treatment at the time of rising PSA on AST was left to the discretion of the treating physician. Overall, there were 53 deaths, 28 (53%) of which resulted from prostate cancer.
Calculation of the PSA DT
PSA DT was calculated by using first-order kinetics and a minimum of three PSA values separated by a minimum of 3 months, each with an increase of more than 0.2 ng/mL. Patients treated with EBRT alone may not have achieved an undetectable PSA level (PSA < 0.2 ng/mL). To determine the PSA DT of the EBRT-managed patients, the nadir PSA level was subtracted from subsequent postradiation PSA values as described previously.3
Statistical Methods
Cox regression10 was used to evaluate whether a significant association existed between the value of the PSA nadir level achieved within 8 months after the initiation of AST and the time to PCSM after 8 months of AST, controlling for treatment and known prognostic factors. Known prognostic factors evaluated included the pre-AST PSA DT, PSA level when AST was initiated, interval to PSA failure, Gleason score, and age at the time of AST initiation. The prostatectomy Gleason score was used in the Cox model for RP-managed patients, and the biopsy Gleason score was used for patients receiving EBRT as an initial therapy.
The PSA DT before AST, interval to PSA failure, and PSA nadir after AST were analyzed first as continuous variables and then as categoric variables (PSA DT < 3 v 3 months; PSA nadir > 0.2 v 0.2 ng/mL; time interval to PSA failure 2 v > 2 years). The Gleason score was analyzed as a categoric variable, 6 (baseline), 7, or 8, in both Cox analyses.
The end point of a PSA DT of less than 3 months was selected for study because evidence exists to support it as a surrogate end point for PCSM.3 A PSA nadir that did not fall below 0.2 ng/mL within 8 months after AST was selected, because this PSA level is generally accepted to be a detectable level on AST.11 In addition, 8 months should be adequate time to observe the PSA nadir.12 A PSA cut point was chosen at 15 ng/mL to show the impact of PSA level at the start of AST on time to PCSM. No pre-existing standard exists for this variable; therefore, the value of 15 ng/mL was chosen for the purpose of illustration. Finally, a time interval to PSA failure of less than 2 years was selected based on its ability to predict time to metastasis in men observed after a rising PSA postoperatively.11 Adjusted hazard ratios (AHRs) for PCSM with 95% CIs were calculated by using the Cox model.13 For the purpose of illustration, cumulative incidence estimates of PCSM were calculated and stratified by the pre-AST PSA DT and the post-AST PSA nadir. These mortality estimates were compared by using a two-sided log-rank test.
RESULTS
Predictors of PCSM
The median time to PSA nadir after AST was 3 months (range, 1 to 6.5 months). Therefore, all patients had at least 1.5 months of observation to confirm the PSA nadir on AST. Of the 28 observed prostate cancer deaths, 21 (75%) and 3 (11%) occurred in men whose PSA nadir was more than 0.2 ng/mL and who had a PSA DT less than 3 months or between 3 and 8.99 months, respectively. The four remaining prostate cancer–specific deaths occurred in 3 men and 1 man whose PSA nadirs were 0.2 ng/mL and who had pre-AST DTs of less than 3 months or between 3 and 5.9 months, respectively.
As shown in Table 2, when analyzed as continuous variables, both the post-AST PSA nadir (P < .0001; AHR, 1.1; 95% CI, 1.0 to 1.1) and the pre-AST PSA DT (P < .0001; AHR, 0.60; 95% CI, 0.47 to 0.76) were significantly associated with the time to PCSM. In addition, the categoric variables of PSA nadir more than 0.2 ng/mL (P < .0001; AHR, 25.3; 95% CI, 8.8 to 73.1) and PSA DT less than 3 months (P < .0001; AHR, 12.1; 95% CI, 4.2 to 34.9) similarly were significantly associated with time to PCSM. In both of these Cox models, Gleason scores 8 to 10 and PSA level at the initiation of AST remained significantly associated with time to PCSM, whereas initial primary treatment and age at the time of AST initiation were not.
Estimates of PCSM
Figures 1 and 2 illustrate the PCSM estimates stratified by using the pre-AST PSA DT and post-AST PSA nadir, respectively. Seven-year cumulative incidence estimates of PCSM were 26% (95% CI, 13% to 39%) for a DT of less than 3 months and 3% (95% CI, 0% to 6%) for a DT of 3 months. These values were 49% (95% CI, 24% to 73%) for a post-AST PSA nadir value of 0.2 ng/mL and 2% (95% CI, 0% to 4%) for those with a PSA nadir value of more than 0.2 ng/mL.
As noted in Figure 3, 68 (30%) of the 224 men whose pre-AST PSA DT was less than 3 months did not have a drop in their PSA level to 0.2 ng/mL on AST. For these men, the 7-year cumulative incidence estimate of PCSM rates was 72% (95% CI, 45% to 99%) compared with 4% (95% CI, 0% to 9%) for patients whose PSA nadir on AST was 0.2 ng/mL. These values were 11% (95% CI, 0% to 27%) and 1% (95% CI, 1% to 2%), respectively, for patients with a pre-AST PSA DT of 3 months.
DISCUSSION
PSA failure after RP or EBRT1,2 occurs, but not all men with PSA failure will die as a result of prostate cancer.3 Achieving a PSA nadir after AST has been shown to be important in patients treated with primary AST for clinically localized prostate cancer.4,5 However, whether the PSA nadir after AST given for PSA failure after RP or EBRT can be used to predict PCSM had not been studied previously and was the subject of this study.
The results of this study showed that a PSA nadir of more than 0.2 ng/mL combined with a pre-AST PSA DT of less than 3 months is closely associated with PCSM and is independent of the initial prostate cancer treatment. Seventy-five percent (21 of 28) of the prostate cancer deaths observed in this study occurred in this group.
It has been reported previously that achieving a PSA nadir under 4.0 ng/mL after orchiectomy with or without flutamide was not significantly associated with PCSM in patients with hormone-na?ve and bone scan–positive prostate cancer.14 However, an update of this study now finds a significant difference in overall survival when stratified by PSA response.15 Therefore, PSA nadir after AST in men with or without radiographic evidence of metastatic disease seems to be significantly associated with PCSM.
Given the recent findings of a survival benefit for the use of docetaxel in patients with metastatic and hormone-refractory prostate cancer,16,17 the clinical significance of the finding in this study is the potential to design a phase III study of AST ± docetaxel in patients with a PSA DT of less than 3 months after RP or EBRT and use a PSA nadir of more than 0.2 ng/mL as one end point. This end point may have clinical relevance, given that the estimate of PCSM in this group reached 72% by 7 years despite the use of AST. The question remaining, however, is whether survival will be prolonged if men receiving AST and docetaxel experience a PSA nadir of 0.2 ng/mL significantly more frequently than those men receiving AST alone.
One potential confounding point regarding the AST used was that some patients (61%) received LHRH agonists, and others (33%) received LHRH agonists and an antiandrogen. If an LHRH agonist plus an antiandrogen were shown to be superior to an LHRH agonist alone, this study would need to be reanalyzed by using a type of AST as a covariate in the Cox model.
Nevertheless, the observation that cancer-related death occurs often and relatively quickly in patients with both a pre-AST PSA DT of less than 3 months and a post-AST PSA nadir of more than 0.2 ng/mL is clinically relevant. Therefore, pending validation, its use as an end point in future randomized clinical trials is justified. In conclusion, a PSA nadir of more than 0.2 ng/mL after AST given for a postoperative or postradiation PSA DT of less than 3 months is closely associated with PCSM and is clinically significant, because it accounted for 75% of the cancer deaths observed in this study.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Presented in part at the 41st Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 13-17, 2005 (poster discussion). The authors state that this is an original work.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Zeitman AL, Chung CS, Coen JJ, et al: 10-year outcome for men with localised prostate cancer treated with external radiotherapy: Results of a cohort study. J Urol 171:210-214, 2004
Han M, Partin AW, Zahurak M, et al: Biochemical (prostate specific antigen) recurrence probability following radical prostatectomy for clinically localized prostate cancer. J Urol 169:517-523, 2003
D'Amico AV, Moul JW, Carroll PR, et al: Surrogate end point for prostate cancer-specific mortality after radical prostatectomy or radiation therapy. J Natl Cancer Inst 95:1376-1383, 2003
Kwak C, Jeong SJ, Park MS, et al: Prognostic significance of the nadir prostate specific antigen level after hormone therapy for prostate cancer. J Urol 168:995-1000, 2002
Shulman MJ, Benaim EA: The natural history of androgen independent prostate cancer. J Urol 172:141-145, 2004
Lubeck DP, Litwin MS, Henning JM, et al: The CaPSURE database: A methodology for clinical practice and research in prostate cancer: CaPSURE Research Panel—Cancer of the Prostate Strategic Urologic Research Endeavor. Urology 48:773-777, 1996
Sun L, Gancarczyk K, Paquette EL, et al: Introduction to the Department of Defense Center for Prostate Disease Research Multicenter National Prostate Cancer Database, and analysis of changes in the PSA-era. Urol Oncol 6:203-209, 2001
American Joint Committee on Cancer–Prostate, in Greene FL (ed): AJCC Cancer Staging Manual (6th ed). New York, NY, Springer, 2002, pp 309-316
American Society for Therapeutic Radiology and Oncology Consensus Panel: Consensus statement—Guidelines for PSA following radiation therapy. Int J Radiat Oncol Biol Phys 37:1035-1041, 1997
Neter J, Kutner MH, Wasserman W, et al: Simultaneous inferences and other topics in regression analysis–1, in Applied Linear Regression Models. Homewood, IL, McGraw-Hill, 1983, pp 150-153
Pound CR, Partin AW, Eisenberger MA, et al: Natural history of progression after PSA elevation following radical prostatectomy. JAMA 281:1591-1597, 1999
Gleave ME, Goldenberg SL, Chin JL, et al: Randomized comparative study of 3 versus 8-month neoadjuvant hormonal therapy before radical prostatectomy: Biochemical and pathological effects. J Urol 166:500-506; discussion 506-507, 2001
Gaynor JJ, Feuer EJ, Tan CC, et al: On the use of cause-specific failure and conditional failure probabilities: Examples from clinical oncology data. J Am Stat Assoc 88:400-409, 1993
Eisenberger MA, Blumenstein BA, Crawford ED, et al: Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med 339:1036-1042, 1998
Hussain M: PSA as an outcome marker in de novo metastatic prostate cancer. Presented at Am Soc Clin Oncol Prostate Cancer Symposium, Orlando, FL, February 17-19, 2005
Tannock IF, de Wit R, Berry WR, et al: Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 351:1502-1512, 2004
Petrylak DP, Tangen CM, Hussain MHA, et al: Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory prostate cancer. N Engl J Med 351:1513-1520, 2004(Alexandra J. Stewart, How)
Memorial Sloan-Kettering Cancer Center, New York, NY
University of Connecticut, Storrs, CT
Walter Reed Hospital, Bethesda, MD
University of California San Francisco, San Francisco, CA
Duke University, Durham, NC
ABSTRACT
PURPOSE: For men receiving androgen-suppression therapy (AST) for a rising postoperative or postradiation prostate-specific antigen (PSA), we evaluated whether a PSA nadir of more than 0.2 ng/mL was significantly associated with prostate cancer–specific mortality (PCSM).
PATIENTS AND METHODS: The study cohort comprised 747 men with rising PSA and negative bone scan after surgery (n = 486) or radiation therapy (n = 261) who were treated with AST. Cox regression was used to evaluate whether a significant association existed between the PSA nadir level after 8 months of AST and the time to PCSM, controlling for treatment and known prognostic factors.
RESULTS: The post-AST PSA nadir (pCox < .0001), the pre-AST PSA doubling time (DT) (pCox = .002), PSA level (P = .0001), and Gleason eight to 10 cancers (pCox = .01) were significantly associated with time to PCSM. The adjusted hazard ratio for PCSM was 20 (95% CI, 7 to 61; pCox < .0001), for men with a PSA nadir of more than 0.2 ng/mL as compared with all others. A PSA DT of less than 3 months was observed in 30% (224 of 747) of the study cohort. Of the 28 observed prostate cancer deaths, 21 (75%) occurred in men whose PSA nadir was more than 0.2 ng/mL and who had a PSA DT of less than 3 months.
CONCLUSION: A PSA nadir of more than 0.2 ng/mL after 8 months of AST given for postoperative or postradiation PSA failure is significantly associated with PCSM and is clinically significant because it accounted for 75% of the cancer deaths observed in this study.
INTRODUCTION
Prostate-specific antigen (PSA) failure after radical prostatectomy (RP) or external-beam radiation therapy (EBRT) for patients with clinically localized prostate cancer occurs in up to 30% of patients within 10 years after treatment.1,2 Yet, not all patients with PSA failure after RP or EBRT will die from prostate cancer. Specifically, for patients with a PSA doubling time (DT) of less than 3 months, the median overall survival from the time of PSA relapse is relatively short (approximately 6 years), whereas patients with a PSA DT of more than 12 months have a median survival that is greater than a decade.3
Although the value of the PSA DT before the institution of hormonal therapy is closely associated with prostate cancer–specific mortality (PCSM),3 the post–androgen-suppression therapy (AST) PSA nadir has also been shown to be associated with PCSM.4,5 In particular, in the setting of localized disease treated with primary AST, studies have shown that patients whose PSA does not have a nadir below 0.2 ng/mL are at a significantly higher risk of death from prostate cancer.4,5
Whether PSA nadir after AST given for PSA failure after primary local therapy provides clinically important information regarding the subsequent time to PCSM has not been studied. Therefore, we evaluated whether a PSA nadir level of more than 0.2 ng/mL compared with 0.2 ng/mL was significantly associated with PCSM for men receiving AST for a rising postoperative or postradiation PSA state.
PATIENTS AND METHODS
Patient Selection, Staging, and Treatment
The study cohort consisted of patients from two multi-institutional databases, Cancer of the Prostate Strategic Urologic Research Endeavor6 and the Center for Prostate Disease Research,7 which contain baseline, treatment, and follow-up information. The patients were treated with RP or EBRT between January 1, 1988, and January 1, 2002, for clinical stage T1C-4, NX or N0, M0 prostate cancer.8 All patients who received neoadjuvant or adjuvant AST were excluded, including 12 and 26 men treated with RP and EBRT, respectively. This left 747 men with a rising PSA after RP (n = 486) or EBRT (n = 261) and negative bone scan obtained within 1 month of the initiation of AST for study. An approved and signed internal review board informed consent form was obtained for each patient before study entry. Ten RP-managed patients (2%) did not achieve an undetectable PSA (median, 0.1 ng/mL; range, 0.1 to 0.2 ng/mL). The median PSA nadir after EBRT was 0.8 ng/mL (range, 0.1 to 2.8 ng/mL). Table 1 lists the patient characteristics of the study cohort.
At the time of PSA relapse, defined by using the American Society for Therapeutic Radiology and Oncology consensus definition for all study patients,9 all patients underwent a history and physical examination including a digital rectal examination, bone scan, and computed tomography scan of the pelvis. All patients with radiographically visible metastases were excluded from the analysis.
The AST treatments included orchiectomy (30 patients [4%]), luteinizing hormone-releasing hormone (LHRH) agonists (456 patients [61%]), antiandrogens (15 patients [2%]), and an LHRH agonist with an antiandrogen (246 patients [33%]).
Follow-Up
The median overall follow-up for the study cohort was 2.2 years (range, 0.8 to 9.5 years), defining 8 months after the initiation of AST as time zero. No patient died during the initial 8 months of AST. At each follow-up, a PSA level was obtained, and history and physical examination including digital rectal examination were performed. Treatment at the time of rising PSA on AST was left to the discretion of the treating physician. Overall, there were 53 deaths, 28 (53%) of which resulted from prostate cancer.
Calculation of the PSA DT
PSA DT was calculated by using first-order kinetics and a minimum of three PSA values separated by a minimum of 3 months, each with an increase of more than 0.2 ng/mL. Patients treated with EBRT alone may not have achieved an undetectable PSA level (PSA < 0.2 ng/mL). To determine the PSA DT of the EBRT-managed patients, the nadir PSA level was subtracted from subsequent postradiation PSA values as described previously.3
Statistical Methods
Cox regression10 was used to evaluate whether a significant association existed between the value of the PSA nadir level achieved within 8 months after the initiation of AST and the time to PCSM after 8 months of AST, controlling for treatment and known prognostic factors. Known prognostic factors evaluated included the pre-AST PSA DT, PSA level when AST was initiated, interval to PSA failure, Gleason score, and age at the time of AST initiation. The prostatectomy Gleason score was used in the Cox model for RP-managed patients, and the biopsy Gleason score was used for patients receiving EBRT as an initial therapy.
The PSA DT before AST, interval to PSA failure, and PSA nadir after AST were analyzed first as continuous variables and then as categoric variables (PSA DT < 3 v 3 months; PSA nadir > 0.2 v 0.2 ng/mL; time interval to PSA failure 2 v > 2 years). The Gleason score was analyzed as a categoric variable, 6 (baseline), 7, or 8, in both Cox analyses.
The end point of a PSA DT of less than 3 months was selected for study because evidence exists to support it as a surrogate end point for PCSM.3 A PSA nadir that did not fall below 0.2 ng/mL within 8 months after AST was selected, because this PSA level is generally accepted to be a detectable level on AST.11 In addition, 8 months should be adequate time to observe the PSA nadir.12 A PSA cut point was chosen at 15 ng/mL to show the impact of PSA level at the start of AST on time to PCSM. No pre-existing standard exists for this variable; therefore, the value of 15 ng/mL was chosen for the purpose of illustration. Finally, a time interval to PSA failure of less than 2 years was selected based on its ability to predict time to metastasis in men observed after a rising PSA postoperatively.11 Adjusted hazard ratios (AHRs) for PCSM with 95% CIs were calculated by using the Cox model.13 For the purpose of illustration, cumulative incidence estimates of PCSM were calculated and stratified by the pre-AST PSA DT and the post-AST PSA nadir. These mortality estimates were compared by using a two-sided log-rank test.
RESULTS
Predictors of PCSM
The median time to PSA nadir after AST was 3 months (range, 1 to 6.5 months). Therefore, all patients had at least 1.5 months of observation to confirm the PSA nadir on AST. Of the 28 observed prostate cancer deaths, 21 (75%) and 3 (11%) occurred in men whose PSA nadir was more than 0.2 ng/mL and who had a PSA DT less than 3 months or between 3 and 8.99 months, respectively. The four remaining prostate cancer–specific deaths occurred in 3 men and 1 man whose PSA nadirs were 0.2 ng/mL and who had pre-AST DTs of less than 3 months or between 3 and 5.9 months, respectively.
As shown in Table 2, when analyzed as continuous variables, both the post-AST PSA nadir (P < .0001; AHR, 1.1; 95% CI, 1.0 to 1.1) and the pre-AST PSA DT (P < .0001; AHR, 0.60; 95% CI, 0.47 to 0.76) were significantly associated with the time to PCSM. In addition, the categoric variables of PSA nadir more than 0.2 ng/mL (P < .0001; AHR, 25.3; 95% CI, 8.8 to 73.1) and PSA DT less than 3 months (P < .0001; AHR, 12.1; 95% CI, 4.2 to 34.9) similarly were significantly associated with time to PCSM. In both of these Cox models, Gleason scores 8 to 10 and PSA level at the initiation of AST remained significantly associated with time to PCSM, whereas initial primary treatment and age at the time of AST initiation were not.
Estimates of PCSM
Figures 1 and 2 illustrate the PCSM estimates stratified by using the pre-AST PSA DT and post-AST PSA nadir, respectively. Seven-year cumulative incidence estimates of PCSM were 26% (95% CI, 13% to 39%) for a DT of less than 3 months and 3% (95% CI, 0% to 6%) for a DT of 3 months. These values were 49% (95% CI, 24% to 73%) for a post-AST PSA nadir value of 0.2 ng/mL and 2% (95% CI, 0% to 4%) for those with a PSA nadir value of more than 0.2 ng/mL.
As noted in Figure 3, 68 (30%) of the 224 men whose pre-AST PSA DT was less than 3 months did not have a drop in their PSA level to 0.2 ng/mL on AST. For these men, the 7-year cumulative incidence estimate of PCSM rates was 72% (95% CI, 45% to 99%) compared with 4% (95% CI, 0% to 9%) for patients whose PSA nadir on AST was 0.2 ng/mL. These values were 11% (95% CI, 0% to 27%) and 1% (95% CI, 1% to 2%), respectively, for patients with a pre-AST PSA DT of 3 months.
DISCUSSION
PSA failure after RP or EBRT1,2 occurs, but not all men with PSA failure will die as a result of prostate cancer.3 Achieving a PSA nadir after AST has been shown to be important in patients treated with primary AST for clinically localized prostate cancer.4,5 However, whether the PSA nadir after AST given for PSA failure after RP or EBRT can be used to predict PCSM had not been studied previously and was the subject of this study.
The results of this study showed that a PSA nadir of more than 0.2 ng/mL combined with a pre-AST PSA DT of less than 3 months is closely associated with PCSM and is independent of the initial prostate cancer treatment. Seventy-five percent (21 of 28) of the prostate cancer deaths observed in this study occurred in this group.
It has been reported previously that achieving a PSA nadir under 4.0 ng/mL after orchiectomy with or without flutamide was not significantly associated with PCSM in patients with hormone-na?ve and bone scan–positive prostate cancer.14 However, an update of this study now finds a significant difference in overall survival when stratified by PSA response.15 Therefore, PSA nadir after AST in men with or without radiographic evidence of metastatic disease seems to be significantly associated with PCSM.
Given the recent findings of a survival benefit for the use of docetaxel in patients with metastatic and hormone-refractory prostate cancer,16,17 the clinical significance of the finding in this study is the potential to design a phase III study of AST ± docetaxel in patients with a PSA DT of less than 3 months after RP or EBRT and use a PSA nadir of more than 0.2 ng/mL as one end point. This end point may have clinical relevance, given that the estimate of PCSM in this group reached 72% by 7 years despite the use of AST. The question remaining, however, is whether survival will be prolonged if men receiving AST and docetaxel experience a PSA nadir of 0.2 ng/mL significantly more frequently than those men receiving AST alone.
One potential confounding point regarding the AST used was that some patients (61%) received LHRH agonists, and others (33%) received LHRH agonists and an antiandrogen. If an LHRH agonist plus an antiandrogen were shown to be superior to an LHRH agonist alone, this study would need to be reanalyzed by using a type of AST as a covariate in the Cox model.
Nevertheless, the observation that cancer-related death occurs often and relatively quickly in patients with both a pre-AST PSA DT of less than 3 months and a post-AST PSA nadir of more than 0.2 ng/mL is clinically relevant. Therefore, pending validation, its use as an end point in future randomized clinical trials is justified. In conclusion, a PSA nadir of more than 0.2 ng/mL after AST given for a postoperative or postradiation PSA DT of less than 3 months is closely associated with PCSM and is clinically significant, because it accounted for 75% of the cancer deaths observed in this study.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Presented in part at the 41st Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 13-17, 2005 (poster discussion). The authors state that this is an original work.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Zeitman AL, Chung CS, Coen JJ, et al: 10-year outcome for men with localised prostate cancer treated with external radiotherapy: Results of a cohort study. J Urol 171:210-214, 2004
Han M, Partin AW, Zahurak M, et al: Biochemical (prostate specific antigen) recurrence probability following radical prostatectomy for clinically localized prostate cancer. J Urol 169:517-523, 2003
D'Amico AV, Moul JW, Carroll PR, et al: Surrogate end point for prostate cancer-specific mortality after radical prostatectomy or radiation therapy. J Natl Cancer Inst 95:1376-1383, 2003
Kwak C, Jeong SJ, Park MS, et al: Prognostic significance of the nadir prostate specific antigen level after hormone therapy for prostate cancer. J Urol 168:995-1000, 2002
Shulman MJ, Benaim EA: The natural history of androgen independent prostate cancer. J Urol 172:141-145, 2004
Lubeck DP, Litwin MS, Henning JM, et al: The CaPSURE database: A methodology for clinical practice and research in prostate cancer: CaPSURE Research Panel—Cancer of the Prostate Strategic Urologic Research Endeavor. Urology 48:773-777, 1996
Sun L, Gancarczyk K, Paquette EL, et al: Introduction to the Department of Defense Center for Prostate Disease Research Multicenter National Prostate Cancer Database, and analysis of changes in the PSA-era. Urol Oncol 6:203-209, 2001
American Joint Committee on Cancer–Prostate, in Greene FL (ed): AJCC Cancer Staging Manual (6th ed). New York, NY, Springer, 2002, pp 309-316
American Society for Therapeutic Radiology and Oncology Consensus Panel: Consensus statement—Guidelines for PSA following radiation therapy. Int J Radiat Oncol Biol Phys 37:1035-1041, 1997
Neter J, Kutner MH, Wasserman W, et al: Simultaneous inferences and other topics in regression analysis–1, in Applied Linear Regression Models. Homewood, IL, McGraw-Hill, 1983, pp 150-153
Pound CR, Partin AW, Eisenberger MA, et al: Natural history of progression after PSA elevation following radical prostatectomy. JAMA 281:1591-1597, 1999
Gleave ME, Goldenberg SL, Chin JL, et al: Randomized comparative study of 3 versus 8-month neoadjuvant hormonal therapy before radical prostatectomy: Biochemical and pathological effects. J Urol 166:500-506; discussion 506-507, 2001
Gaynor JJ, Feuer EJ, Tan CC, et al: On the use of cause-specific failure and conditional failure probabilities: Examples from clinical oncology data. J Am Stat Assoc 88:400-409, 1993
Eisenberger MA, Blumenstein BA, Crawford ED, et al: Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med 339:1036-1042, 1998
Hussain M: PSA as an outcome marker in de novo metastatic prostate cancer. Presented at Am Soc Clin Oncol Prostate Cancer Symposium, Orlando, FL, February 17-19, 2005
Tannock IF, de Wit R, Berry WR, et al: Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 351:1502-1512, 2004
Petrylak DP, Tangen CM, Hussain MHA, et al: Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory prostate cancer. N Engl J Med 351:1513-1520, 2004(Alexandra J. Stewart, How)