|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 25 (September 1), 2005: pp. 6157-6162 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.01.2336 Preoperative PSA Velocity Is an Independent Prognostic Factor for Relapse After Radical ProstatectomyFrom the Departments of Radiation Oncology and Urology, Division of Urologic Oncology, Stanford University School of Medicine, Stanford, CA Address reprint requests to Christopher R. King, PhD, MD, Department of Radiation Oncology, Stanford University School of Medicine, 875 Blake Wilbur Dr, Stanford, CA 94305-5847; e-mail: christopher.king{at}stanford.edu
PURPOSE: Preoperative prostate-specific antigen (PSA) velocity (PSAV), or the rate of PSA rise before diagnosis, predicts for risk of cancer death after radical prostatectomy (RP). We evaluated the relative merit of established preoperative factors, including biopsy indices and preoperative PSAV, for their impact on relapse after RP. PATIENTS AND METHODS: The outcomes of 202 men who underwent RP were reviewed. Biopsies were characterized for grade, percentage positive cores, and total linear tumor length. Surgical specimens were characterized for cancer volume, relative percentage by grade, extracapsular extension, and margin status. Univariate and multivariate analyses were performed with respect to relapse-free survival after RP.
RESULTS: Thirty-one patients relapsed after RP (defined as PSA CONCLUSION: The preoperative PSAV is a significant independent clinical factor predicting for relapse after RP and also predicts for larger, more aggressive, and more locally advanced tumors. Its inclusion will be useful in risk stratification, evaluation for alternatives to surgery, and patient selection for neoadjuvant or adjuvant therapies as part of randomized clinical trials.
A recent study showed that the rate of rise in prostate-specific antigen (PSA) during the year before the diagnosis of prostate cancer (the annual PSA velocity [PSAV]) predicted the risk of recurrence and the risk of prostate cancer death after radical prostatectomy (RP).1 After adjusting for known clinical factors (clinical T stage, initial PSA level, and biopsy Gleason score), the PSAV remained an independent factor. In the current era of PSA screening and extended biopsy schemes, more men are diagnosed with lower PSA levels, nonpalpable tumors, and smaller tumors; thus, the relative usefulness of the previously established clinical factors such as T stage and PSA level will diminish. As a separate measure of biologic aggressiveness, the PSAV should have an increasing role in guiding treatment recommendations and in patient selection for adjuvant therapies. To further evaluate the usefulness of PSAV, we examined the relapse rate after RP with a multivariate analysis that included clinical T stage, initial PSA level, biopsy Gleason score, and preoperative PSAV. What distinguishes the current study is the inclusion of biopsy indices that have been shown to independently predict for relapse after RP, the percent of positive cores2 and the amount of tumor within positive cores,3 and the examination of the pathologic implications for patients with elevated preoperative PSAV.
Patient Selection and Pathology Of 1,180 men treated with RP at Stanford University Medical Center (Stanford, CA) between December 1989 and October 2001 for clinically localized prostate cancer, 202 men had preoperative annual PSAV data available and form the basis of this study. Because the majority (82%) of these patients were treated between 1996 and 2001, the results are not susceptible to "era effect." No patient received neoadjuvant androgen suppression before RP. Staging evaluation included a history and physical examination, serum PSA level, and transrectal ultrasound-guided needle biopsy of the prostate. The majority of patients had sextant biopsies (55% of patients had 6 to 7 cores, 32% had 8 to 10 cores, and the remaining 13% had 11 to 16 cores). Imaging with computed tomography, magnetic resonance imaging, or bone scan for staging was at the discretion of the physicians, depending on the pretreatment clinical parameters. Clinical T stage was based on digital rectal examination only and recorded by using the 2002 American Joint Committee on Cancer (AJCC) staging system.4 Biopsies were characterized for grade, percent of positive cores, and total linear tumor length. Surgical specimens were characterized for volume of cancer based on tumor mapping from serial step-sectioning at 3-mm intervals, relative percentage by grade, extracapsular extension, and margin status. Gleason grading5 and characterization of biopsy and prostatectomy specimens were made by a single genitourinary pathologist (J.E.M.). Bilateral pelvic lymph node sampling was performed on all patients. Patients with pathologically positive lymph node(s) were not included in this study.
In general, patients were followed up (with a digital rectal examination and serum PSA level) every 3 months for the first 2 years and then every 6 months thereafter. Relapse was defined as biochemical failure when a detectable PSA (
Statistical Methods The annual PSAV (the rate of change in PSA per year) was calculated by simple linear regression from two or more PSA measurements before the positive biopsy diagnosis. Two PSA measurements were available among 56% of the patients, three among 25% of the patients, four among 15% of the patients, and five among 4% of the patients. The median time interval between PSA measurements was 12.0 months (range, 3.4 to 34 months). No PSA subsequent to the prostate needle biopsy was used. Relapse-free survival (RFS) was calculated by using the Kaplan-Meier method.6 RFS was defined as the time from RP to relapse (event) or to date of the last follow-up (censored).
Univariate and multivariate Cox proportional-hazards regression analysis7 of preoperative variables that might predict for relapse after RP was performed. The preoperative covariates examined were age, clinical tumor stage (T1c or T2), preoperative PSA (
Thirty-one patients relapsed after RP, with a median time to failure of 16 months. All but age and clinical T stage were significant on univariate analyses in predicting for relapse after RP; appropriate clinical variables were tested as both continuous and categoric variables. These results are summarized in Table 2, which also includes the actuarial RFS rate at 5 years. It is worth pointing out the statistical significance of the percent of positive cores and the total tumor length. It is also interesting to note the persistence of the preoperative PSAV in predicting for RFS for the different cutoff levels of  1, 2, and 3 ng/mL/year. Kaplan-Meier estimates of RFS at 5 years were 97% for the group with an annual PSAV of 1.0 ng/mL, compared with 75% for the group with an annual PSAV of > 1.0 ng/mL (P = .01), 89% compared to 73% for PSAV of 2 v > 2 ng/mL (P = .003), and 88% compared to 68% for PSAV of 3 v > 3 ng/mL (P = .005).
The Cox multivariate regression analyses are presented in Table 3. Two separate analyses were performed, one with the preoperative PSAV 1 cutoff and the other with the PSAV 2 cutoff. A separate analysis was also performed with the PSAV 3 cutoff but is not shown because it failed to reach significance. Only two clinical factors retained their independent significance in predicting RFS after RP: the biopsy Gleason sum (P < .008; relative risk [RR], > 4.8) and the preoperative PSAV, whether dichotomized at 1 ng/mL/year (P = .04; RR, 4.7) or 2 ng/mL/year (P = .02; RR, 3.0). For these two factors, Kaplan-Meier curves illustrating the unadjusted relationship of RFS and time since surgery are shown in Figure 1 for the preoperative PSAV and Figure 2 for biopsy Gleason sum.
We examined the relationship between the preoperative PSAV and surgical pathologic findings, which are summarized in Table 4. Patients with a PSAV of > 2 ng/mL/year were more likely be pT3 (P = .007), have positive margins (P = .01), have tumors > 1 mL (P = .05), possess > 10% grade 4/5 tumors (P = .04), and have a volume of grade 4/5 > 0.5 mL (P = .03).
We have shown in this study that the preoperative PSAV predicts for relapse after RP and is an independent significant factor after adjusting for clinical T stage, initial PSA level, biopsy Gleason score, percent of positive cores, and amount of tumor within positive cores. The only other preoperative factor that is also independently significant was the biopsy Gleason score. Men with an annual preoperative PSAV of > 2 ng/mL have a three-fold increase in their risk of relapse after prostatectomy (95% CI, 1.2 to 7.7). We confirm recently published findings that demonstrate that preoperative PSAV predicts for relapse and cancer death after RP.1 We have also shown, at least within the context of these data, that although the biopsy cancer volume indices (percentage of positive cores and total linear tumor length) were individually predictive on univariate analysis, they no longer are significant on multivariate analysis when PSAV is included among the pretreatment parameters. Our data show a progressive rate of relapse with higher PSAV threshold levels (Fig 1). Although the specific PSAV threshold level one chooses is arbitrary, a level of 2 ng/mL/year seems reasonable for clinical usage. A higher threshold of 3 ng/mL/year no longer separated out groups significantly on multivariate analysis when other clinical factors were considered. A lower threshold of 1 ng/mL/year did retain the ability to independently distinguish patients at risk, but such a low threshold might be too susceptible to alternate causes of PSA fluctuations and to measure error if the time interval between measurements is too short. A threshold level of 2 ng/mL/year is also consistent with that observed in the previously cited study.1 The biology behind the power of preoperative PSAV in predicting for relapse after RP is probably complex. Some insight is suggested by considering the time to relapse. Based on our understanding of the increased likelihood of persistent local disease associated with a late time to failure,8 a rough estimate can be made that half of the failures in this series are local and half are distant, because the median time to relapse is 16 months. Therefore preoperative PSAV is not likely a simple predictor of either local failure or distant failure. We do not possess PSA doubling time after biochemical relapse for these patients, but such information might help to further distinguish between local or systemic relapse. These data also show that patients with a preoperative PSAV of > 2 ng/mL/year are more likely to have pT3 stage, larger tumors, a larger component of high-grade tumors, and a higher chance for positive surgical margins. Several prior studies have failed to show a relationship between pretreatment PSAV and adverse pathologic features.9,10 These studies had modest patient numbers (86 and 82, respectively), had separate pathologists for specimen interpretations, and were possibly underpowered. On the other hand, two other modest-sized studies (56 and 62 patients, respectively) using a related metric, the PSA doubling time, have shown a relationship between the PSA doubling time and adverse pathologic features11,12 and biochemical relapse rates.12 The previously mentioned study1 did find a relationship between PSAV and advanced pathologic stage, higher tumor grade, and positive lymph nodes. In conclusion, this study shows that the preoperative PSAV is a significant independent pretreatment clinical factor predicting for relapse after RP. It can also serve to predict which patients are more likely to possess larger, more aggressive, and more locally advanced tumors. The inclusion of preoperative PSAV in risk stratification, evaluation for alternatives to surgery, and patient selection for neoadjuvant or adjuvant therapy as part of randomized clinical trials seems justified.
The authors indicated no potential conflicts of interest.
1. D’Amico AV, Chen M-H, Roehl KA, et al: Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med 351:121-135, 2004
2. D’Amico AV, Whittington R, Malkowicz SB, et al: Clinical utility of the percentage of positive prostate biopsies in defining biochem-ical outcome after radical prostatectomy for patients with clinically localized prostate cancer. J Clin Oncol 18:1164-1172, 2000 3. Freedland SJ, Aronson WJ, Csathy GS, et al: Comparison of percentage of total prostate needle biopsy tissue with cancer to percentage of cores with cancer for predicting PSA recurrence after radical prostatectomy: Results from the SEARCH database. Urology 61:742-747, 2003[Medline] 4. Greene FL, Page DL, Flemming ID, et al: AJCC Manual for Staging Cancer (ed 6). New York, NY, Springer Verlag, 2002, pp 309-313 5. Gleason DF: Veterans Administration Cooperative Urological Research Group.Histological grading and staging of prostatic carcinoma, in Tannenbaum M (ed): Urologic Pathology: The Prostate. Philadelphia, PA, Lea & Febiger, 1977, pp171-187 6. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958[CrossRef] 7. Cox DR: Regression models and life tables (with discussion). J R Stat Soc B 34:187-220, 1972
8. Pound CR, Partin AW, Eisenberger MA, et al: Natural history of progression of PSA elevation following radical prostatectomy. JAMA 281:1591-1597, 1999 9. Freedland SJ, Dorey F, Aronson WJ: Preoperative PSA velocity and doubling time do not predict adverse pathologic features or biochemical recurrence after radical prostatectomy. Urology 57:476-480, 2001[Medline] 10. Thiel R, Pearson JD, Epstein JI, et al: Role of prostate-specific antigen velocity in prediction of final pathologic stage in men with localized prostate cancer. Urology 49:716-720, 1997[CrossRef][Medline] 11. Goluboff ET, Heitjan DF, DeVries GM, et al: Pretreatment prostate specific antigen doubling times: Use in patients before radical prostatectomy. J Urol 158:1876-1878,1997[CrossRef][Medline] 12. Egawa S, Arai Y, Tobisu K, et al: Use of pretreatment prostate-specific antigen doubling time to predict outcome after radical prostatectomy. Prostate Cancer Prostatic Dis 3:269-274, 2000[CrossRef][Medline] Submitted January 19, 2005; accepted May 18, 2005.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
0.2 ng/mL), with a median time to failure of 16 months. Median follow-up was 48 months. Kaplan-Meier relapse-free survival at 5 years was 89%, compared with 73% for PSAV 
