|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 24 (August 20), 2006: pp. 3984-3990 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.06.4246 Absolute Prostate-Specific Antigen Value After Androgen Deprivation Is a Strong Independent Predictor of Survival in New Metastatic Prostate Cancer: Data From Southwest Oncology Group Trial 9346 (INT-0162)
From the University of Michigan, Ann Arbor, MI; Southwest Oncology Group Statistical Center; Puget Sound Oncology Consortium, Seattle, WA; Devine-Tidewater Urology, Norfolk, VA; University of Colorado Health Science Center, Denver, CO; University of Wisconsin Hospital and Clinics, Madison, WI; Marmara University Hospital, Istanbul, Turkey; University of California, San Francisco, San Francisco, CA; Prostate Cancer Institute, Calgary, Canada; Northwestern University, Feinberg School of Medicine, Chicago, IL; and the Cleveland Clinic Foundation, Cleveland, OH Address reprint requests to Southwest Oncology Group (SWOG-9346), Operations Office, 14980 Omicron Drive, San Antonio, TX 78245-3217; e-mail: pubs{at}swog.org
PURPOSE: To establish whether absolute prostate-specific antigen (PSA) value after androgen deprivation (AD) is prognostic in metastatic (D2) prostate cancer (PCa). PATIENTS AND METHODS: D2 PCa patients with baseline PSA of at least 5 ng/mL received 7 months induction AD. Patients achieving PSA of 4.0 ng/mL or less on months 6 and 7 are randomly assigned to continuous versus intermittent AD on month 8. Eligibility for this analysis required a prestudy PSA with at least two subsequent PSAs and that patients be registered at least 1 year before analysis date. Survival was defined as time to death after 7 months of AD. Associations were evaluated by proportional hazards regression models. RESULTS: One thousand one hundred thirty four of 1,345 eligible patients achieved a PSA of 4 ng/mL or less. At end of induction, 965 patients maintained PSA of 4 or less and 604 had a PSA of 0.2 ng/mL or less. After controlling for prognostic factors, patients with a PSA of 4 or less to more than 0.2 ng/mL had less than one third the risk of death (ROD) as those with a PSA of more than 4 ng/mL (P < .001). Patients with PSA of 0.2 ng/mL or less had less than one fifth the ROD as patients with a PSA of more than 4 ng/mL (P < .001) and had significantly better survival than those with PSA of more than 0.2 to 4 ng/mL or less (P < .001). Median survival was 13 months for patients with a PSA of more than 4 ng/mL, 44 months for patients with PSA of more than 0.2 to 4 ng/mL or less, and 75 months for patients with PSA of 0.2 ng/mL or less. CONCLUSION: A PSA of 4 ng/mL or less after 7 months of AD is a strong predictor of survival. This data should be used to tailor future trial design for D2 prostate cancer.
Androgen-deprivation therapy (ADT) is standard for metastatic (D2) prostate cancer; however, response duration and survival are variable in subgroups of patients.1 To date there is no clear variable or set of variables that identify patients with worse prognosis with a high level of confidence, nor is there a validated surrogate for survival. The ability to assess prognosis individually will improve study design, expedite completion of trials and appropriately maximize clinical benefit while minimizing exposure to unnecessary toxicities. Since its identification, prostate-specific antigen (PSA) has been a biomarker for diagnosis, risk prediction, and monitoring of disease. In D2 disease, data from a Southwest Oncology Group (SWOG) trial2 suggested that PSA progression heralds clinical progression in patients treated with ADT by a median of 6 months. However, it is not well established whether the absolute PSA value after ADT is prognostic in this setting. To address this, we evaluated data from a SWOG phase III trial in patients with new stage D2 prostate cancer. The results of this analysis are the subject of this article.
Study Design SWOG 9346 is a prospective intergroup (SWOG, Eastern Cooperative Oncology Group [ECOG], Cancer and Leukemia Group B [CALGB], European Organisation for Research and Treatment of Cancer [EORTC], National Cancer Institute of Canada [NCIC]) phase III trial (Fig 1). Its primary objective is to determine whether survival with intermittent ADT is equivalent to survival with continuous ADT. The accrual goal is 1,512 eligible, randomly assigned patients. Key eligibility requirements include a new stage D2 prostate cancer, a minimum pretreatment PSA of 5 ng/mL by Hybritech (San Diego, CA) or Abbott (Abbott Park, IL) assays (or similarly calibrated assays), and a SWOG performance status of 0 to 2.
Patients are treated for a 7-month "induction" course (step 1) with Goserelin and bicalutamide. At the end of the induction period responding patients, as judged by a stable or declining PSA level of 4 ng/mL or less at months 6 and 7 of treatment, are randomly assigned to intermittent or continuous ADT (step 2). To maximize enrollment, patients who had started ADT before registration were allowed into the study provided that they were otherwise eligible and therapy was initiated no longer than 6 months before registration. Therefore, at registration (referred to as induction registration) there were two subsets of patients: a group that initiated therapy after registration and is referred to as early-registration induction and a group that had already started ADT before registration, referred to as late-registration induction. These groups are identified for the purposes of rational statistical analysis and to avoid inadvertent selection biases, but the nature and timing of random assignment are such that no biologic impact from these groupings would be anticipated in the overall result of the trial. Patients who failed to achieve PSA of 4 ng/mL or less or who did not maintain PSA of 4 mg/mL or less during the induction phase of the trial were removed from protocol but followed for progression and survival. These patients were treated off-study as deemed medically appropriate. Details on subsequent nonprotocol treatment were not collected. Only those who achieved and maintained a confirmed PSA of 4 mg/mL or less at the end of this induction period were randomly assigned to continuous or intermittent ADT. At the time the study was designed (early 1990s), there was little available experience with intermittent ADT. The rationale for the PSA criteria for random assignment was based on available preliminary reported data, which based a decision regarding proceeding with intermittent ADT on achieving a PSA threshold of 4 ng/mL or less with a declining trend, irrespective of prior local therapy. This value and trend were felt to reflect clinically an androgen-dependent tumor.3 While on study, patients were seen and examined every 3 months and PSA levels were obtained at months 1, 4, 6, and 7 of the induction period. After random assignment, PSA levels were assessed monthly for both arms.
Statistical Analysis Methods To be eligible for this analysis, patients had to be eligible for S9346 and had to have a prestudy PSA and at least two subsequent PSA values collected during induction. Early- and late-induction patients were included in the analysis. Late-induction patients with no recorded treatment start date were excluded. To allow the appropriate duration of follow-up and data collection, patients included in this analysis had to have been registered at least 1 year before the date of analysis (for the early induction patients) or started combined therapy at least 1 year prior (for the late induction patients). The start date of treatment was the registration date for early-induction patients, and the start date of combined therapy for late-induction patients. Patients without a PSA value in the month-6 to -7 time window as required by study were assumed to not have achieved a PSA of 4 ng/mL or less at that time point. If a patient had a PSA of 4 ng/mL or less before month 6 to 7 that was followed by a PSA of more than 4 ng/mL during month 6 or 7, then he was counted as having PSA progression at the end of induction. Undetectable PSA was any PSA of 0.2 ng/mL or less. Patients with any undetectable PSA at months 6 and 7, regardless of other values before or after the undetectable value during that period, were counted as having achieved an undetectable value. Patients with last follow-up less than 7 months after registration (early induction) or after treatment start (late induction) were excluded from the survival analysis. Overall survival was measured starting after 7 months of induction ADT until death or last follow-up. Univariate logistic and proportional hazards regression models were developed for each covariate predicting PSA levels of 4 ng/mL or less and postinduction survival, respectively. Multivariate models were constructed using the significant covariates from the univariate models, for which there was less than 15% missing data, and then variables associated with a PSA levels of 4 ng/mL or less were also included in the model. Because S9346 is an ongoing trial, the study's data and safety monitoring committee has given permission for these results to be reported before study closure since no information about randomized treatment assignment and corresponding survival is being reported and the data generated will not prospectively influence trial accrual or assessment of outcomes.
Patients Characteristics This analysis was performed on the first 1,395 patients registered to the study who had adequate PSA assessments. Patient characteristics stratified by PSA status at the end of induction are presented in Table 1.
Of the 1,395 registered patients 1,345 (96.4%) were included for the survival analyses and 50 were excluded because of lack of postinduction follow-up. Three of these patients had a PSA 4 ng/mL or less, and 47 did not. The median number of on-study PSA values during induction was 5 (range, 2 to 18). The primary difference between the 1,345 included and the 50 excluded from the survival analyses was calendar year of study entry. Those excluded tended to be registered later in the study (on average, half a year later than those included) so less time was available for collection of follow-up data.
Achieving PSA of 4.0 ng/mL or Less
Undetectable PSA Of the 1,395 patients, 675 (48%) achieved an undetectable PSA (  0.2 ng/mL) during the induction period (Table 2), and 604 (89%) maintained undetectable levels at the end of induction period.
Predictors for Failing to Achieve a PSA Less Than 4.0 ng/mL
Table 4 provides the multivariate associations with lack of achieving a PSA of 4 ng/mL or less at any time and at months 6 and 7. Independent predictors of lack achieving of a PSA of 4 ng/mL or less at months 6 and 7 include higher PSA, Gleason score  8, younger age, worse performance status, and presence of bone pain (all P < .05). The results for lack of a PSA of 4 ng/mL or less at any time during induction are also similar to predictors at months 6 and 7 except that age at study entry and performance status were not significant (Table 4).
PSA Values and Survival Of the 1,395 registered patients 1,345 were included for the postinduction survival analyses. Figure 2 demonstrates the Kaplan-Meier curves for postinduction survival by PSA status at months 6 and 7 of induction.
Because the vast majority of those achieving PSA levels of 4.0 or less during induction still maintained that at months 6 and 7 (965 of 1,134, 85%), the survival curves for PSA status at any time during induction looked very similar to Figure 2 and are not shown here. The median follow-up time for the 965 patients who achieved a PSA of 4.0 ng/mL or less was 38 months, and it was 35 months for those who did not achieve a PSA level of 4.0 ng/mL or less. To evaluate potential differential loss to follow-up, we defined a patient as lost if his last known contact date was at least 15 months before the analysis. Of those who failed to achieve a PSA of 4.0 ng/mL or less, 2.9% were lost to follow-up, and of those who did achieve a PSA of 4.0 ng/mL or less, 7.3% were lost. The median overall survival was 13 months for the 383 with a PSA of more than 4 ng/mL at the end of induction (95% CI, 11 to 16 months), 44 months for the 360 patients who had a PSA more than 0.2 ng/mL but 4.0 ng/mL or less at months 6 and 7 (95% CI, 39 to 55 months), and 75 months for the 602 patients with undetectable PSA ( 0.2 ng/mL) at months 6 and 7 (95% CI, 62, 91 months). Predictors of postinduction survival are shown in Table 5. Patients with lower PSA at study entry, better performance status, no significant weight change, no bone pain, a Gleason sum of 7 or less, no visceral metastases, or no distant lymphadenopathy had better survival (all P < .05). Patients with a PSA of 4.0 ng/mL or less during induction had one quarter the risk of dying relative to those who did not have a PSA of 4.0 ng/mL or less (hazard ratio [HR], 0.26; 95% CI, 0.22 to 0.31; P < .0001). An undetectable PSA was also a significant predictor of longer survival (HR, 0.34; 95% CI, 0.29 to 0.40; P < .0001).
Four statistically significant (P < .05) independent risk factors were included in the multivariate risk model predicting postinduction survival: performance status, Gleason sum, bone pain, and PSA at study entry. After adjustment for these factors, patients who had a PSA of 4 or less but more than 0.2 ng/mL during induction had less than one third the risk of death of those who had a PSA more than 4 ng/mL (HR, 0.30; 95% CI, 0.24 to 0.38; P < .0001). Patients with an undetectable PSA had less than one fifth the risk of death of those with PSA more than 4 ng/mL (HR, 0.17; 95% CI, 0.13 to 0.21; P < .0001) and had significantly better survival than those with PSA of 4 ng/mL or less but who did not achieve undetectable PSA at months 6 and 7 (Wald  2, P < .0001), after adjustment for the other covariates.
We assessed the impact of achieving a PSA of 4 ng/mL or less or an undetectable PSA on survival in hormone-naïve D2 prostate cancer patients treated with ADT on S9346. The data were generated from more than 1,000 patients all treated uniformly and followed similarly in the context of a phase III trial. Its strength also stems from the fact that PSA was collected prospectively using similarly calibrated assays. Furthermore, the threshold established for the assessments were prospectively set; as a result very clear survival outcome differences were detected using a simply defined PSA end point. Specifically, this analysis demonstrates that a PSA of 4 ng/mL or less and undetectable PSA ( 0.2 ng/mL) at the end of a 7-month ADT period is the most powerful predictor of risk of death in patients with D2 prostate cancer. Oosterlinck et al4 found that a PSA of 4ng/mL or less in 546 men with M0 and M1 disease treated with ADT occurred in 51% of patients after 3 months and in 66% after 6 months. PSA values rarely decreased further after 6 months. Others attempted to correlate PSA with outcome. Collette et al5 recently reported one of the largest analyses evaluating PSA end points as possible surrogates for overall survival in men with metastatic disease. PSA normalization was one of the PSA end points, but unlike our study, they applied a meta-analytic approach, utilizing data from three randomized trials. Although their analysis included 2,161 men with new D2 disease, these men were not treated uniformly (therapy ranged from single-agent bicalutamide with differing dose levels to goserelin with or without bicalutamide or flutamide and orchiectomy), not all included trials monitored PSA similarly, and it is not clear whether the same PSA assay or calibrated assays were required in these trials. Patients with a baseline PSA of more than 20 ng/mL were eligible for the PSA normalization analysis. PSA was evaluated over 18 months, and adjustments for other patient or disease covariates were not described. The authors' conclusion was that PSA end points could not be statistically validated as a surrogate for overall survival. However, PSA normalization was a strong prognostic factor for survival. Other published reports have investigated the prognostic value of PSA response in hormone-naïve metastatic patients.4,6-15 Although a variety of definitions of PSA response are used, these studies consistently reveal a significant correlation between PSA response and time to progression and overall survival. In the study by Oosterlinck et al, patients who normalized their PSA had a median time to progression of more than 36.5 months, whereas those who did not normalize after 6 months experienced a median time to progression of 15.5 months. Furthermore, they reported that patients with PSA of 4 ng/mL or less at 6 months had the best prognosis for a complete remission, and patients with a level over 4 ng/mL had a high probability of early treatment failure.4 Although the study by Collette et al contained more than 2,000 men,5 a few of these earlier PSA end point studies included only a few hundred patients,4,7,9 and several of these analysesincluded fewer than 100 men with newly D2 disease.6,8,10,13,15 Additionally, the hormonal therapy in several cases would not be considered standard therapy today. The information provided by this survival analysis as a function of PSA value in response to ADT is the most powerful prognosticator yet, at an individual patient level, using one easily and reproducibly measurable factor. The results are robust enough to allow the identification of patients who are unlikely to do well with standard ADT long before developing overt androgen independence, thus providing a window of opportunity to investigate newer therapies. At the time the study was designed, the relevance of pretherapy PSA velocity was not known. Because the primary objective of the study was to assess survival outcome on the basis of the schedule of ADT, pretreatment PSA data were not collected. Additional potential confounding factors are that the 50 patients not included in this survival analysis disproportionately had a PSA of more than 4 ng/mL. They tended to be registered more recently to the study, but there was also a suggestion of worse performance status (other risk factor distributions were comparable to those included in the analysis). It is unlikely that the exclusion of these individuals had much effect on the results except that one might hypothesize that had they been included in the survival analysis, those who failed to achieve a PSA of 4 ng/mL or less may have had even worse survival. Another potential confounding factor is that patients who failed to achieve a PSA of 4 ng/mL or less might have received less effective subsequent treatment. However, with prevailing patterns of practice that, at a minimum, maintain gonadal suppression, it is unlikely that the difference in outcome observed is based on inferior treatment, but rather disease biology. Furthermore, it is established that response to ADT is limited, and regardless of initial response to induction ADT, all patients will receive a variety of subsequent treatments including chemotherapy regimens at treatment failure. However, the treatment course is not expected to be different depending on initial PSA response to induction ADT. The precise factors associated with better survival after ADT in this setting are not well characterized. The observed variability in survival is striking with median overall survivals ranging from 13 to 75 months on the basis of the absolute PSA achieved, thus reflecting the biologic heterogeneity of this disease. In an effort to gain a better understanding of the determinants of this heterogeneity, S9346 will explore the impact of circulating hormone levels, growth factors, and genetic variability of candidate genes involved in the synthesis and metabolism of steroid hormones and the insulin-like growth factor. In conclusion, achieving a PSA of 4 ng/mL or less after 7 months ADT is a strong and specific predictor of risk of death. Patients with a PSA of 0.2 ng/mL or less have the greatest survival advantage. This is the first trial in the setting of new D2 prostate cancer to demonstrate unequivocally the survival advantage associated with absolute PSA value in response to ADT. These findings should be used to tailor future trial design for D2 prostate cancer.
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Dollar Amount Codes (A) < $10,000 (B) $10,000-99,999 (C)
Presented in part at the 42nd Annual Meeting of the American Society of Clinical Oncology, Atlanta, GA, June 2-6, 2006. Supported in part by the following Public Health Service Cooperative Agreement Grants No. awarded by the National Cancer Institute (National Institutes of Health, Bethesda, MD), Department of Health and Human Services: CA38926, CA32102, CA14028, CA58882, CA42777, CA46441, CA58658, CA46282, CA35192, CA58416, CA46113, CA76132, CA45807, CA20319, CA27057, CA04919, CA12644, CA35431, CA22433, CA35261, CA46368, CA67575, CA58861, CA76447, CA63848, CA37981, CA63844, CA46136, CA45560, CA11083, CA67663, CA16385, CA12213, CA35119, CA35090, CA35178, CA63845, CA86780, CA74647, CA45461, CA95860, CA45377, CA35281, CA35996, CA45808, CA35128, and CA35262. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Tangen CM, Faulkner JR, Crawford ED, et al: Ten-year survival in patients with metastatic prostate cancer. Clin Prostate Cancer 2: 41-45, 2003[Medline] 2. Eisenberger M, Crawford ED, McLeod D, et al: The prognostic significance of prostate specific antigen (PSA) in stage D2 prostate cancer (PC) interim evaluation of intergroup study 0105. Proc Am Soc Clin Oncol 14: 235, 1995 (abstr 613) 3. Goldenberg SL, Bruchovsky N, Gleave ME, et al: Intermittent androgen suppression in the treatment of prostate cancer: A preliminary report. Urology 45: 839-845, 1995[CrossRef][Medline] 4. Oosterlinck W, Mattelaer J, Casselman J, et al: PSA evolution: A prognostic factor during treatment of advanced prostatic carcinoma with total androgen blockade—Data from a Belgian multicentric study of 546 patients. Acta Urol Belg 65: 63-71, 1997[Medline] 5. Collette L, Burzykowski T, Carroll KJ, et al: Is prostate-specific antigen a valid surrogate end point for survival in hormonally treated patients with metastatic prostate cancer? Joint research of the European Organisation for Research and Treatment of Cancer, the Limburgs Universitair Centrum, and AstraZeneca Pharmaceuticals. J Clin Oncol 23: 6139-6148, 2005 6. Arai Y, Yoshiki T, Yoshida O: Prognostic significance of prostate specific antigen in endocrine treatment for prostatic cancer. J Urol 144: 1415-1419, 1990[Medline] 7. Collette L, de Reijke TM, Schroder FH: Prostate specific antigen: A prognostic marker of survival in good prognosis metastatic prostate cancer? (EORTC 30892). Eur Urol 44: 182-189, 2003[CrossRef][Medline] 8. Cooper EH, Armitage TG, Robinson MR, et al: Prostatic specific antigen and the prediction of prognosis in metastatic prostatic cancer. Cancer 66: 1025-1028, 1990[Medline] 9. Dijkman GA, Janknegt RA, De Reijke TM, et al: Long-term efficacy and safety of nilutamide plus castration in advanced prostate cancer, and the significance of early prostate specific antigen normalization. International Anandron Study Group. J Urol 158: 160-163, 1997[CrossRef][Medline] 10. Ercole CJ, Lange PH, Mathisen M, et al: Prostatic specific antigen and prostatic acid phosphatase in the monitoring and staging of patients with prostatic cancer. J Urol 138: 1181-1184, 1987[Medline] 11. Fowler JE Jr, Pandey P, Seaver LE, et al: Prostate specific antigen regression and progression after androgen deprivation for localized and metastatic prostate cancer. J Urol 153: 1860-1865, 1995[CrossRef][Medline] 12. Matzkin H, Soloway MS, Schellhammer PF, et al: Prognostic factors in stage D2 prostate cancer treated with a pure nonsteroidal antiandrogen. Cancer 72: 1286-1290, 1993[CrossRef][Medline] 13. Miller JI, Ahmann FR, Drach GW, et al: The clinical usefulness of serum prostate specific antigen after hormonal therapy of metastatic prostate cancer. J Urol 147: 956-961, 1992[Medline] 14. Reynard JM, Peters TJ, Gillatt D: Prostate-specific antigen and prognosis in patients with metastatic prostate cancer–A multivariable analysis of prostate cancer mortality. Br J Urol 75: 507-515, 1995[Medline] 15. Stamey TA, Kabalin JN, Ferrari M, et al: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate—IV, Anti-androgen treated patients. J Urol 141: 1088-1090, 1989[Medline] Submitted March 7, 2006; accepted June 19, 2006.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
