|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Comparison of the Efficacy of Local Therapies for Localized Prostate Cancer in the Prostate-Specific Antigen Era: A Large Single-Institution Experience With Radical Prostatectomy and External-Beam RadiotherapyByFrom the Departments of Radiation Oncology and Urology, Cleveland Clinic Foundation, Cleveland, OH. Address reprint requests to Patrick Kupelian, MD, Department of Radiation Oncology, Desk T28, Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195; email: kupelip{at}ccf.org
PURPOSE: To review biochemical relapse-free survival (bRFS) rates after either external-beam radiotherapy (RT) or radical prostatectomy (RP) for localized prostate cancer.
PATIENTS AND METHODS: All 1,682 patients had pretreatment prostate-specific antigen (PSA) levels and biopsy Gleason scores (bGS) assigned. No adjuvant therapy was administered after local treatment. RP was the treatment in 1,054 patients (63%) and RT in 628 patients (37%). Median follow-up was 51 months (range, 1 to 134). The median follow-up for RP versus RT patients was 50.5 v 51.0 months. Biochemical relapse was considered detectable PSA levels (> 0.2 ng/mL) in RP patients and three consecutive rising PSA levels in RT patients. The analysis was repeated with a more stringent definition of biochemical control after either RP or RT—namely, reaching and maintaining a PSA level
RESULTS: Eight-year bRFS rates for RP versus RT were 72% and 70%, respectively (P = .010). Multivariate analysis indicated T stage (P < .001), pretreatment PSA (P < .001), bGS (P < .001), year of therapy (P < .001), and neoadjuvant AD (P = .019) to be the only independent predictors of relapse. Age (P = .78), race (P = .29), prior transurethral resection of prostate (P = .81), and treatment modality (P = .96) were not independent predictors of treatment failure. Fifty-one percent of RP patients had favorable tumors (T1 to T2A, pretreatment PSA CONCLUSION: Eight-year biochemical failure rates were identical between RT and RP in any subgroup. Outcome is determined mainly by pretreatment PSA levels, bGS, clinical T stage, and, for RT patients, radiation dose.
THE TREATMENT OF localized prostate cancer remains controversial because of the lack of conclusive well-controlled or randomized studies comparing outcomes with radiotherapy (RT) to radical prostatectomy (RP). A randomized trial published in 1982 demonstrating an advantage to RP was never widely accepted, because of randomization artifacts and worse than previously reported RT results.1,2 The Southwest Oncology Group closed a randomized study in the mid-1980s because of poor accrual. In 1993, Stamey3 reported a 20% 5-year cure rate with RT and suggested that radiation accelerates prostate cancer growth.4 Subsequently, large RT series were published with outcome results stratified by biopsy grade, T stage, and serum prostate-specific antigen (PSA) demonstrating similar outcomes for RT and RP.5-8 The previous observation of a 20% "cure" rate with RT can largely be explained by patient selection factors.8 Furthermore, Leibman et al9 demonstrated that PSA velocity is similar in those whose disease fails to respond to radiation or surgery. None of these studies clearly answers the question of which is the best local therapy for localized prostate cancer. The unsettled nature of this issue is further complicated by the marked polarization of radiation oncologists and urologists in their counseling of newly diagnosed patients with localized prostate cancer.10 Biochemical failure rates after radiation or surgery in a nonrandomized comparative cohort of RP and RT patients treated at our institution in the early PSA era (1987 to 1993) were similar at 5 years.11 In this report, we expand and update our initial observations to a larger cohort of patients who had similar staging evaluations, more uniform selection criteria, and longer follow-up, and who are representative of contemporarily treated patients.
Between January 1990 and December 1998, 1,865 consecutive patients with clinical stage T1 and T2 adenocarcinoma of the prostate were treated with either RP or RT to a dose of  68 to 78 Gy. The following were excluded from the analysis: no available pretreatment PSA (iPSA) levels (n = 8), no available biopsy Gleason scores (bGS; n = 4), androgen deprivation (AD) given after the local therapy (n = 75), RT to the prostatic fossa postoperatively in an adjuvant setting (n = 27), neoadjuvant AD given for a duration exceeding 6 months (n = 51), and no follow-up PSA levels available for analysis (n = 18). A total of 1,682 patients were included in the present study. Only 320 (19%) had neoadjuvant AD, all for 6 months or less. RP was the primary treatment in 1,054 patients (63%), and RT in 628 patients (37%). Seventeen percent of RP patients received neoadjuvant AD, compared with 23% of RT patients. The initial evaluation included determination of the 1992 American Joint Committee on Cancer clinical stage, iPSA level, and bGS determination. Further work-up with transrectal ultrasound, bone scan, chest x-ray, and computed tomography scans of the abdomen and pelvis was obtained according to the individual physician preference. Similar proportions of RP versus RT patients had bone scans (61% and 60%, respectively), and pelvic computed tomography scans (37% and 34%, respectively). None of these evaluations revealed definite bony or lymphatic metastases. The major discrepancy was in the surgical evaluation of pelvic lymph nodes; only 16 RT patients (3%) had pelvic lymph node dissection (PLND), whereas 793 patients (75%) managed surgically had bilateral PLND. Two percent of RP patients (n = 17) had positive nodes versus none of the 16 RT patients who also underwent PLND. The prostatectomies were completed in the 17 RP patients with microscopic lymph node involvement, no adjuvant AD was given, and all 17 patients were included in this series. On the basis of clinical stage, iPSA levels, bGS, and PLND findings in RP patients, we estimate that approximately 4% of the RT patients would have had surgically detectable lymph node metastases. Only a total of 69 patients (4%) had undergone prior transurethral resection of prostate (TURP) for benign prostatic hypertrophy.
A unilateral or bilateral nerve-sparing procedure was performed in RP 578 patients (55%). Megavoltage equipment was used in the delivery of radiation. The median total dose was 70.2 Gy (range, 68.0 to 78.0 Gy). A conformal technique was used in 321 patients (51% of RT patients).12,13 Higher radiation doses, especially more than 68 Gy, have been demonstrated to result in decreased biochemical failure rates.14 Therefore, we defined adequate RT as
Follow-up information always included PSA levels; 11,678 follow-up PSA levels were available for analysis. The average number of follow-up PSA levels was 8.7 for RT patients versus 5.9 for RP patients. The median follow-up time for all 1,682 patients was 51 months (range, 1 to 134 months; mean, 54 months). The frequency of follow-up visits was based on physician preference. Typically, the follow-up visits were obtained every 6 months. The median follow-time up for RP versus RT patients was 50.5 v 51.0 months. The median follow-time up for RT patients receiving less than 72 Gy v The analysis end point was biochemical relapse-free survival (bRFS). For RT patients, the American Society of Therapeutic Radiology and Oncology consensus definition for biochemical failure was used: three consecutive rising PSA levels after a nadir. The time to failure was calculated to be midway between the time of nadir and the first PSA increase. For RP patients, failure of disease to respond to therapy was defined as two consecutive detectable PSA levels (> 0.2 ng/mL). The time to failure was considered as the time of the initial detectable level. Because all clinical relapses were associated with or preceded by PSA level elevations, biochemical failures included both PSA increases and clinical failures.
To study the effect of local modality (RP v RT) on bRFS, Kaplan-Meier curves were generated and the log rank statistic was used to determine if there was a significant difference between the curves. To examine the possibility of confounding, a Cox proportional hazards multivariate analysis was performed using the following variables; T stage, bGS, iPSA, treatment modality (surgery v radiation), race, age, neoadjuvant AD, prior TURP, and year of therapy. The RT group was further analyzed by comparing outcomes in those treated with 68.0 to 71.9 Gy (the < 72 Gy group) to those treated with 72.0 to 78.0 Gy (the
Finally, to address the issue of the adequacy of using different definitions of biochemical failure for the two treatment subgroups, the entire analysis was repeated with the more stringent definition of biochemical control of reaching and maintaining a PSA level
Pretreatment/Treatment Characteristics Table 1 summarizes the pretreatment clinical characteristics of the 1,682 patients by treatment modality. As expected, the patients treated with RP were significantly younger, a higher proportion was white, and they had more favorable tumor characteristics.
Outcomes in RP Patients Of the 1,054 RP patients, no residual cancer could be found in four patients. A surgical specimen Gleason score could not be determined because of prior neoadjuvant AD in 176 patients. Of the remaining 874 RP patients, 398 (46%) had a Gleason score of  6, and 476 (54%) had a Gleason score of 7. Subgrouping patients by exclusive pathologic categories, 607 (58%) had organ-confined disease, 175 (16%) had extracapsular extension and negative surgical margins, 165 (16%) had extracapsular extension and positive surgical margins, 90 (8%) had seminal vesicle invasion, and 17 (2%) had positive lymph nodes. Table 2 compares the bRFS rates between the prostatectomy patients in the present series and patients from contemporary prostatectomy series from Johns Hopkins University, Washington University, and Baylor.15-17 The results are stratified by different variables.
Treatment Results For the 1,682 patients, the 5- and 8-year bRFS rates were 77% (95% confidence interval [CI], 75% to 80%) and 71% (95% CI, 67% to 75%), respectively (Fig 1). The 5-year bRFS rates for RP versus RT were 80% (95% CI, 77% to 83%) v 73% (95% CI, 69% to 77%), respectively (Fig 2). The 8-year bRFS rates for RP versus RT were 72% (95% CI, 68% to 77%) v 70% (95% CI, 64% to 75%), respectively. The difference was statistically significant between the two treatment arms in favor of prostatectomy (P = .010). Table 3 lists the results of the multivariate analysis for all patients; T stage, iPSA, bGS, year of therapy, and neoadjuvant AD were the only independent predictors of relapse. Age, race, prior TURP, and treatment modality were not.
To control for the effect of tumor parameters on bRFS, two risk groups were created. Patients with a favorable risk were defined as having stage T1 to T2A lesions, bGS 6, and iPSA levels 10 ng/mL. All other patients were classified as having disease with an unfavorable prognosis. Fifty-one percent of RP patients had tumors with a favorable prognosis, compared with only 34% of RT patients (P < .001,  2). No outcome differences between RP and RT could be observed with favorable tumors (Fig 3); the 8-year bRFS rates for RP versus RT were 86% (95% CI, 77% to 90%) v 90% (95% CI, 85% to 96%), respectively (P = .53). There was also no statistically significant difference between RP and RT with unfavorable tumors (Fig 4); the 8-year bRFS rates for RP versus RT were 62% (95% CI, 55% to 69%) v 59% (95% CI, 51% to 67%), respectively (P = .21). For patients with an unfavorable outcome, this difference was particularly striking when comparing standard-dose RT (< 72 Gy) to RP (Fig 5); the 5-year bRFS rates for RP versus RT less than 72 Gy were 70% (95% CI, 65% to 75%) v 50% (95% CI, 42% to 57%), respectively (P < .001). On the other hand, the 5-year bRFS rates for unfavorable patients with RP versus RT 72 Gy were 70% (95% CI, 65% to 75%) v 82% (95% CI, 76% to 88%), respectively (P = .004).
To better evaluate parameters affecting bRFS, two separate multivariate analyses were performed for RT and RP patients (Table 4). For RP patients, surgical margin status was the most significant parameter predicting failure. Nerve-sparing surgery did not affect bRFS rates. For RT patients, radiation dose rather than radiation technique was the important factor predictive of outcome. For RT patients, neoadjuvant AD for 6 months was associated with significant improvement of bRFS. The effect of neoadjuvant AD was minimal for RP patients.
Outcome Analysis Using the Alternative Definition of Biochemical Failure For both RP and RT patients, biochemical failure was defined as a follow-up PSA level of more than 0.5 ng/mL. RT patients in whom the PSA levels were still clearly decreasing were censored at the time of last follow-up. The 5-year bRFS rates for RP versus RT were 81% v 61%, respectively (P < .001). Figure 6A illustrates the bRFS rates for all patients, with clearly worse outcome for RT less than 72-Gy patients. The 8-year bRFS rates for RP, RT 72 Gy, and RT less than 72 Gy were 72%, 68%, and 34%, respectively. The same finding is observed in patients with favorable (Fig 6B) and unfavorable tumors (Fig 6C). For favorable patients, the 8-year bRFS rates for RP, RT 72 Gy, and RT less than 72 Gy were 86%, 86%, and 48%, respectively. For unfavorable patients, the 8-year bRFS rates for RP, RT 72 Gy, and RT less than 72 Gy were 62%, 61%, and 28%, respectively. Table 5 displays the results of the multivariate analysis of factors affecting bRFS rates, with the alternate RT failure definition. Treatment modality was not an independent predictor of outcome (P = .10).
Repeat Analysis Excluding Patients Receiving Neoadjuvant AD To avoid any effect of the use of neoadjuvant AD for 6 months on the follow-up PSA profile, the analysis was repeated with the exclusion of patients receiving any AD. Figure 7 illustrates the bRFS curves of the 878 RP patients and 484 RT patients treated with local therapy alone, broken down by radiation dose groups, with either failure definition used. The same observations as in Fig 6 are made again, with no differences detected between RP and RT 72 Gy. Outcomes were worse with RT less than 72 Gy. Table 6 displays the results of the multivariate analysis of factors affecting bRFS rates, similar to Table 3, in this group of patients treated with surgery or RT alone with no AD. Treatment modality was again not an independent predictor of outcome (P = .23).
Because the initial report of our nonrandomized large single-institutional experience of RT versus RP at the Cleveland Clinic Foundation in patients treated in the early PSA era (1987 to 1993),11 similar observations were made from other institutions. These largely demonstrated no significant difference between the two modalities when outcomes are stratified by pretreatment parameters such as T stage, bGS, and most importantly, by pretreatment PSA levels.18,19 In our present study, there was significant improvement in outcomes observed at 5 years, compared with our previous observations made for patients treated earlier in the PSA era.11 This is the results of several factors: stage migration,20 better patient selection,13 and improvement in surgical and radiotherapeutic techniques.12,21 For example, in our present study, the overall 8-year bRFS rate with prostatectomy was 72%, which is comparable to outcomes observed in large surgical series of patients treated with RP in the same time period (Table 2). In parallel, the 8-year bRFS rate with RT was 70%, comparable to the outcomes observed in the Memorial-Sloan Kettering Cancer Center and Fox Chase Cancer Center series.22-24 In addition, several improvements in the delivery of radiation (conformal technique, high radiation doses) and improvement in surgical techniques (lateral fascial dissections) have contributed to the improvement of observed outcomes. Overall, RP patients treated had significantly higher bRFS rates compared with RT patients (Fig 2). To a large extent, this is secondary to the more unfavorable nature of lesions treated with RT, as evidenced by the distribution of higher stage, higher grade, and higher PSA lesions in Table 1 and the multivariate analysis for all patients (Table 2). However, this is also partially due to the worse outcome associated with standard-dose RT (< 72 Gy) compared with surgery (Figs 5 and 6). We believe these data suggest that radiation doses less than 72 Gy should be considered inadequate to control localized prostate cancers.
The benefit from neoadjuvant AD for
One major criticism of comparisons between radiation and surgery for localized prostate cancer has been the use of two different end point definitions. Using the more stringent definition of reaching and maintaining a PSA level The long natural history of localized prostate cancer makes it difficult to determine from our data which therapy is best in men with life expectancies longer than 8 to 10 years at diagnosis. In pervious reports, we demonstrated that biochemical failure was not an independent predictor of mortality at 10 years after either RP or RT,26,27 and others have demonstrated that the lead time for mortality after biochemical failure to death is longer than 10 years after RP.28,29 Although age was not an independent predictor of outcome in this series, it is theoretically possible that, in men with long life expectancies at diagnosis, the late local recurrence rate might be higher after radiation (or other prostate-sparing treatments such as brachytherapy or cryotherapy) and that this will translate into more local complications or a higher mortality rate. These questions cannot be addressed without additional follow-up. In conclusion, intrinsic tumor characteristics seem to be more important than treatment modality in 8-year biochemical-free relapse rates after radiation or surgery in men with localized prostate cancer. Biochemical failure rates are primarily determined by pretreatment PSA levels, bGS, clinical T stages, and, for RT patients, radiation dose. Longer follow-up is needed to determine whether these observations hold at 10-, 15-, and 20-year posttreatment intervals.
1. Paulson DF, Lin GH, Hinshaw W, et al: Radical surgery versus radiotherapy for adenocarcinoma of the prostate. J Urol 128: 502-504, 1982[Medline] 2. Hanks GE: More on the Uro-Oncology Research Group report of radical surgery vs. radiotherapy for adenocarcinoma of the prostate. Int J Radiat Oncol Biol Phys 14: 1053-1054, 1988[Medline] 3. Stamey T, Ferrari M, Schmid H: The value of serial prostate specific antigen determinations 5 years after radiotherapy: Steeply increasing values characterize 80% of patients. J Urol 150: 1856-1859, 1993[Medline] 4. Brendler CB: Prostate cancer. J Urol 150: 1865-1866, 1993[Medline] 5. Zagars G, von Eschenbach A: Prognostic factors in prostate cancer: Prostate specific antigen—An important marker for prostate cancer treated by external beam radiotherapy. Cancer 72: 538-548, 1993[CrossRef][Medline] 6. Zagars GK: Serum PSA as a tumor marker for patients undergoing definitive radiation therapy. Urol Clin North Am 20: 737-747, 1993[Medline] 7. Zagars GK: Prostate specific antigen as an outcome variable for T1 and T2 prostate cancer treated by radiation therapy. J Urol 152: 1786-1791, 1994[Medline] 8. Zietman AL, Shipley WU: Re: The value of serial prostate specific antigen determinations 5 years after radiotherapy—Steeply increasing values characterize 80% of patients. J Urol 152:1564-1565; discussion 1565-1566, 1994 9. Leibman BD, Dillioglugil O, Scardino PT, et al: Prostate-specific antigen doubling times are similar in patients with recurrence after radical prostatectomy or radiotherapy: A novel analysis. J Clin Oncol 16: 2267-2271, 1998[Abstract]
10. Fowler FJ Jr, McNaughton Collins M, Albertsen PC, et al: Comparison of recommendations by urologists and radiation oncologists for treatment of clinically localized prostate cancer. JAMA 283: 3217-3222, 2000 11. Kupelian PA, Katcher J, Levin H, et al: External beam radiotherapy versus radical prostatectomy for clinical stage T1-2 prostate cancer: Therapeutic implications of stratification by pretreatment PSA levels and biopsy Gleason scores. Cancer J Sci Am 3: 78-87, 1997[Medline] 12. Kupelian PA, Mohan DS, Lyons J, et al: Higher than standard radiation doses (72 Gy or greater) with or without androgen deprivation in the treatment of localized prostate cancer. Int J Radiat Oncol Biol Phys 46: 567-574, 2000[CrossRef][Medline] 13. Lyons JA, Kupelian PA, Mohan DS, et al: Importance of high radiation doses (72 Gy or greater) in the treatment of stage T1-T3 adenocarcinoma of the prostate. Urology 55: 85-90, 2000[CrossRef][Medline] 14. Pollack A, Zagars GK: External beam radiotherapy dose response of prostate cancer. Int J Radiat Oncol Biol Phys 39: 1011-1018, 1997[CrossRef][Medline] 15. Pound CR, Partin AW, Epstein JI, et al: Prostate-specific antigen after anatomic radical retropubic prostatectomy: Patterns of recurrence and cancer control. Urol Clin North Am 24: 395-406, 1997[CrossRef][Medline] 16. Catalona WJ, Ramos CG, Carvalhal GF: Contemporary results of anatomic radical prostatectomy. CA Cancer J Clin 49: 282-296, 1999[Abstract] 17. Eastham J, Scardino P: Radical prostatectomy for clinical stage T1 and T2 prostate cancer, in Vogelzang N, Scardino P, Shipley W, et al (eds): Comprehensive Textbook of Genitourinary Oncology. Philadelphia, PA, Lippincott Williams & Wilkins, 1999, pp 722-738
18. D’Amico AV, Whittington R, Malkowicz SB, et al: Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. JAMA 280: 969-974, 1998 19. Martinez AA, Gonzalez JA, Chung AK, et al: A comparison of external beam radiation therapy versus radical prostatectomy for patients with low risk prostate carcinoma diagnosed, staged, and treated at a single institution. Cancer 88: 425-432, 2000[CrossRef][Medline]
20. Jhaveri FM, Klein EA, Kupelian PA, et al: Declining rates of extracapsular extension after radical prostatectomy: Evidence for continued stage migration. J Clin Oncol 17: 3167-3172, 1999 21. Klein EA, Kupelian PA, Tuason L, et al: Initial dissection of the lateral fascia reduces the positive margin rate in radical prostatectomy. Urology 51: 766-773, 1998[CrossRef][Medline]
22. Kattan M, Zelefsky M, Kupelian PA, et al: A pretreatment nomogram for disease progression following conformal radiotherapy for prostate cancer. J Clin Oncol 18: 3352-3359, 2000 23. Hanks GE, Hanlon AL, Pinover WH, et al: Dose selection for prostate cancer patients based on dose comparison and dose response studies. Int J Radiat Oncol Biol Phys 46: 823-832, 2000[CrossRef][Medline] 24. Hanks GE, Hanlon AL, Schultheiss TE, et al: Dose escalation with 3D conformal treatment: Five year outcomes, treatment optimization, and future directions. Int J Radiat Oncol Biol Phys 41: 501-510, 1998[CrossRef][Medline] 25. Menon M, Tewari A, Divine G, et al: Comparison of long-term survival in men with clinically localized prostate cancer managed conservatively, with definitive radiation or radical prostatectomy. J Urol 165: 151, 2001 26. Kupelian P, Buchsbaum J, Patel C, et al: Impact of biochemical failure as a predictor of overall survival in men with localized prostate cancer treated with radiotherapy. Int J Radiat Oncol Biol Phys 52: 704-711, 2002[CrossRef][Medline] 27. Jhaveri FM, Zippe CD, Klein EA, et al: Biochemical failure does not predict overall survival after radical prostatectomy for localized prostate cancer: 10-year results. Urology 54: 884-890, 1999[CrossRef][Medline] 28. Iselin CE, Robertson JE, Paulson DF: Radical perineal prostatectomy: Oncological outcome during a 20-year period. J Urol 161: 163-168, 1999[CrossRef][Medline]
29. Pound CR, Partin AW, Eisenberger MA, et al: Natural history of progression after PSA elevation following radical prostatectomy. JAMA 281: 1591-1597, 1999 Submitted January 30, 2002; accepted May 7, 2002.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
