- Email: [email protected]
25-Year Prostate Cancer Control and Survival Outcomes: A 40-Year Radical Prostatectomy Single Institution Series
25-Year Prostate Cancer Control and Survival Outcomes: A 40-Year Radical Prostatectomy Single Institution Series Christopher R. Porter,* Koichi Kodama, Robert P. Gibbons, Roy Correa, Jr., Felix K.-H. Chun, Paul Perrotte and Pierre I. Karakiewicz† From the Department of Urology, Virginia Mason Medical Center, Seattle, Washington (CRP, KK, RPG, RC), and Cancer Prognostics and Health Outcomes Unit, University of Montreal, Montreal, Quebec, Canada (FKHC, PP, PIK)
Purpose: We report on 25-year cancer control and survival outcomes after radical prostatectomy in a single center series of patients treated during a 40-year period. Materials and Methods: Between 1954 and 1994, 787 consecutive patients underwent radical prostatectomy at Virginia Mason Medical Center in Seattle, Washington. Kaplan-Meier 25-year probabilities of prostate cancer specific, overall, prostate specific antigen progression-free, local and distant progression-free survival were determined. Multivariate Cox regression models addressed prostate cancer specific mortality. Results: Prostate cancer specific survival, overall survival, prostate specific antigen progression-free survival, local and distant progression-free survival ranged from 99.0% to 81.5%, 93.5% to 19.3%, 84.8% to 54.5%, 95.3% to 87.8% and 95.9% to 78.2%, respectively. In univariate analyses pathological stage, surgical margin status, pathological Gleason sum, delivery of hormonal therapy and radiotherapy represented statistically significant predictors of prostate cancer specific mortality (all p ⱕ0.001). In multivariate analyses only Gleason sum (p ⫽ 0.03) and delivery of hormonal therapy (p ⬍0.001) remained significant. Conclusions: This is one of the most mature radical prostatectomy series. It demonstrates that long-term biochemical cancer control outcomes after radical prostatectomy might be suboptimal. However, local and distant control outcomes are excellent, and cancer specific mortality is minimal even 25 years after surgery. Key Words: prostatic neoplasms, prostatectomy, outcome assessment, survival analysis
neoadjuvant therapy. Clinical and pathological data were logged into a prospective database by one of the authors (RPG) from 1969 to 1999. In later years the records were maintained electronically with institutional review board approval. For the purposes of analyses 35 of 787 men were excluded due to missing age (9) and formal pathology (26), which left 752 evaluable records. Preoperative comorbidity was defined according to the Charlson weighted index of comorbidity.6 For purposes of analyses, the single patient with a weighted index of 4 was grouped with those with an index of 2. RPP was performed in 428 (56.9%) patients while 324 (43.1%) underwent RRP.7,8 PLND at surgery was performed in 328 patients (45.5%) who were selected according to individual surgeon preference and surgical approach. PLND limits consisted of the obturator nerve, external iliac vein, hypogastric vein and the inguinal ligament. PLND pathologies were not recorded in the database and could not be included in the present analyses. RP specimens were evaluated by VMCC pathologists. Specimens were processed as half mount specimens and serially sectioned at 5 mm intervals. Alternate 5 mm sections were fixed in formalin and were paraffin embedded. From 1954 to 1992 pathological tumor stage was recorded according to the Whitmore-Jewett classification and later converted to the 1992 AJCC staging guidelines.9,10 From 1992 to 1994 the 1992 AJCC staging system was used for clinical and pathological staging.
adical prostatectomy represents a treatment modality of choice for men with localized prostate cancer.1– 4 However, few series provide RP cancer control outcomes beyond 20 years.1–3 Conversely, a recent series of men treated conservatively was updated and provides 20-year cancer specific survival outcomes.5 Unfortunately the lack of long-term (median followup greater than 10 years) cancer specific survival data makes comparisons to long-term conservative series difficult. To address this limitation and to allow prognostication beyond the currently available 15-year cutoff, we updated the followup of our large RP series and report on 25-year outcomes in 787 consecutive men treated with RP at our institution between 1954 and 1994.
R
MATERIALS AND METHODS Between December 1954 and May 1994, 787 patients underwent RP at VMMC by 13 surgeons. No patient received
Submitted for publication September 2, 2005. * Correspondence: Department of Urologic Oncology, Virginia Mason Medical Center, Section of Urology and Renal Transplantation, Seattle, Washington 98101 (telephone: 206-223-6162; FAX: 206223-7650; e-mail: [email protected]). † Supported by the Fonds de la Recherche en Santé du Québec, the CHUM Foundation, the Department of Surgery and Les Urologues Associés du CHUM.
See Editorial on page 437.
0022-5347/06/1762-0569/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
569
Vol. 176, 569-574, August 2006 Printed in U.S.A. DOI:10.1016/j.juro.2006.03.094
570
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
From 1992 to 1994 tumors were routinely classified according to the Gleason grading system.11 Before 1992 tumor grade was recorded as well differentiated (I), moderately differentiated (II) and poorly differentiated (III). To recode these data we followed the paradigm outlined by Roehl et al, in which well differentiated tumors are classified as Gleason sum 3, moderately differentiated tumors are assigned Gleason sum 6 and poorly differentiated tumors are assigned Gleason sum 9.1 Positive surgical margins were recorded as presence of cancer cells against the inked resection margin. Serum PSA testing was initiated at VMMC in 1988. Since that time patients were followed at least quarterly for 2 years, then at least biannually for 2 years and then at least annually. Biochemical recurrence was defined as PSA greater than 0.1 ng/ml. Local pelvic disease or disease consistent with regional lymph node metastases was defined by tissue diagnosis or by overt radiographic signs of disease. Distant metastases were diagnosed based on radiographic and/or scintigraphic studies. Adjuvant or salvage hormonal and/or radiotherapy were delivered according to individual surgeon preference. Cause of death was ascertained according to detailed chart review or was obtained from the VMMC cancer registry. The cancer registry uses links with the Washington State Death Certificate Office. PCA must be the first listed cause of death on the certificate for a patient to be classified as having died of PCA. Actuarial analyses addressed the outcomes PSA recurrence, local recurrence, distant recurrence, PCSS and overall survival. In analyses of PCSS, patients without evidence of progression were censored at the time of last followup. Patients who died of other causes were censored at time of death. PSA recurrence-free data were censored if radiotherapy and/or hormonal therapy were delivered before PSA recurrence. PCA specific mortality was also analyzed with univariate and multivariate Cox regression models. Predictors included patient age at surgery, year of surgery, 1992 AJCC pathological stage, surgical margin status, pathological Gleason sum, type of surgery (RPP vs RRP), lymph node dissection status (performed vs not), delivery of hormonal therapy and delivery of radiation therapy. All statistical tests were performed with S-Plus® Professional and statistical significance was set at 0.05. RESULTS Descriptive statistics are shown in table 1. Mean patient age was 63.6 years (median 64.4, range 42.9 to 77.9). In 341 (45%) patients PCA was nonorgan confined. One or multiple positive surgical margins were recorded in 283 specimens (37.6%). Gleason 7 or higher was recorded in 233 (31%) specimens. Radiotherapy was delivered to 118 (15.7%) and hormonal therapy was given to 114 (15.2%) patients. Followup ranged from 0.1 to 40.5 years (mean 11.7, median 11.4). The number of men at risk according to followup is shown in table 2. PSA recurrence was noted in 189 (31.4% of 601 evaluable patients), local recurrence was noted in 63 (8.4%) and distant recurrence was noted in 85 (11.3%). Overall 298 men died (39.6%) and of these 50 died of PCA (6.6%). Part A of the figure illustrates PCA specific survival. Parts B to H of the figure show PCA specific survival stratified according to 7 main classifying variables that were included in univariate and multivariate analyses.
TABLE 1. Characteristics of 752 patients undergoing radical prostatectomy at the Virginia Mason Medical Center between 1954 and 1994 No. Pts (%) Clinical stage (1992): T1 T2a T2b or T2c T3 Pathological T stage (1992): T2a T2b or T2c T3 Pos surgical margins Pathological Gleason sum: 2 3 4 5 6 7 8 9 10 Surgical procedure: Perineal Retropubic Lymph node dissection Weighted Charlson comorbidity index:* 0 1 2 4 Hormonal therapy Radiotherapy: Adjuvant Salvage† Overall recurrence: PSA‡ Local Distant Overall mortality: PCa specific Other cause
53 (7.1) 515 (68.5) 180 (23.9) 4 (0.5) 177 (23.5) 234 (31.1) 341 (45.3) 283 (37.6) 3 (0.4) 20 (2.7) 50 (6.6) 147 (19.5) 299 (39.8) 170 (22.6) 38 (5.1) 22 (2.9) 3 (0.4) 428 (56.9) 324 (43.1) 342 (45.5) 685 (91.1) 43 (5.7) 23 (3.1) 1 (0.1) 114 (15.2) 192 (25.5) 118 (15.7) 75 (10.0) 189 (31.4) 63 (8.4) 85 (11.3) 298 (39.6) 50 (6.6) 248 (33.0)
* Charlson score of 0 to 1 indicates no or minor comorbidities and score of at least 2 indicates significant comorbidities. † One case has adjuvant and salvage, and is set as adjuvant. ‡ Data set of 601 patients.
Table 3 shows actuarial survival estimates according to 5 definitions— overall survival, PCA specific survival, distant recurrence-free survival, local recurrence-free survival and PSA recurrence-free survival. PCA specific survival is further stratified according to pathological Gleason sum. Overall survival ranged from 93.5% at 5 years to 19.3% at 25 years. PCA specific survival ranged from 99.0% at 5 years to 81.5% at 25 years. Distant recurrence-free survival ranged from 95.9% to 78.2% at 5 and 25 years, respectively. Local progression-free survival ranged from 95.3% to 87.8% at 5 and 25 years, respectively. Finally PSA progression-free survival ranged from 84.8% to 54.5% at 5 and 25 years, respectively. Table 4 shows univariate and multivariate Cox regression analyses predicting PCA specific mortality. In univariate analyses pathological stage (p ⬍0.001), surgical margin status (p ⫽ 0.001), Gleason sum (p ⬍0.001), delivery of radiotherapy (p ⬍0.001) and delivery of hormonal therapy (p ⬍0.001) represented statistically significant predictors. In multivariate analyses only Gleason sum (p ⫽ 0.033) and delivery of hormonal therapy (p ⬍0.001) remained statistically significant.
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY TABLE 2. Number of 752 patients at risk after radical prostatectomy according to year of followup Yrs Followup
No. Pts Current Series (%)
No. Pts Han et al Series2 (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
752 (100.0) 719 (95.6) 701 (93.2) 679 (90.3) 660 (87.8) 630 (83.8) 603 (80.2) 570 (75.8) 526 (69.9) 489 (65.0) 444 (59.0) 404 (53.7) 342 (45.5) 294 (39.1) 246 (32.7) 203 (27.0) 167 (22.2) 143 (19.0) 123 (16.4) 108 (14.4) 84 (11.2) 63 (8.4) 49 (6.5) 36 (4.8) 25 (3.3) 16 (2.1) 12 (1.6) 11 (1.5) 9 (1.2) 6 (0.8) 5 (0.7) 4 (0.5) 3 (0.4) 3 (0.4) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1)
2,404 (100.0) 2,093 (87.0) 1,846 (77.0) 1,612 (67.0) 1,396 (58.0) 1,186 (49.0) 1,038 (43.0) 888 (37.0) 746 (31.0) 621 (26.0) 479 (20.0) 343 (14.0) 227 (9.0) 145 (6.0) 78 (3.0) 29 (1.0) 7 (0.3)
DISCUSSION Men with localized PCA have many treatment options ranging from radical surgery to watchful waiting.1–5 Randomized trials comparing long-term outcomes (median outcome more than 10 years) according to treatment type are unavailable. Therefore, long-term outcome data are sorely needed to evaluate treatment results. Long-term outcomes from conservative approaches for localized PCA were the subject of 2 recent publications.5,12 Both reports indicated that men with well differentiated disease have excellent PCA specific survival while men with poorly differentiated disease are likely to die of the disease within 10 years of diagnosis.5,12 Unfortunately 20-year and longer PCA specific RP outcomes are not available.1–3 To address this limitation we report long-term outcomes in 752 men who underwent RP between 1954 and 1994 at a single institution. Our analysis indicates that long-term cancer control outcomes are excellent when PCA specific survival, distant recurrence-free survival and local recurrence-free survival are assessed up to 25 years after RP (tables 3 and 5). For these outcomes eventfree rates at 25 years are 81.5%, 78.2% and 87.8%, respectively. Conversely, assessment of PSA recurrence-free rates demonstrates a steady decrease in event-free rates which change from 84.8% at 5 years, to 61.1% at 15 years, to 54.5% at 25 years. Taken together these data indicate that RP
571
offers excellent cancer control rates which are remarkably stable over time. The exception is PSA recurrence-free survival, which decreases steadily over time. However, PSA recurrence rates in these series have limited impact on longterm prognosis measured by PCA specific survival, distant recurrence-free survival and local recurrence-free survival. Our PSA observations are in agreement with previous findings.13 Based on median followup of 10 years, Ward et al reported that only 8.3% of men with PSA recurrence after RP died of PCA, which represented a mortality figure of 2.7% in a cohort of 3,897 evaluable RP cases.13 Therefore, it may be concluded that PSA recurrence is associated with PCA specific mortality. However, only a small proportion of men with PSA recurrence actually die of PCA. At 25 years after RP, PCA specific survival was 82.2%. When PCA specific survival data were stratified according to pathological Gleason sum, 20-year survival rates were noted as 64.5% for Gleason 8 to 10, 86.3% for Gleason 7, 85.0% for Gleason 6 and 95.5% for Gleason 2 to 5. Our multivariate analysis substantiated that pathological Gleason represents a statistically significant multivariate predictor of PCA specific survival as shown in previous reports.1–3 It is of interest that other variables such as pathological stage and surgical margin status did not reach statistical significance in multivariate PCA specific survival analyses. The multivariate effect of these variables on PCA specific survival was not assessed in other long-term outcome series.1–3 Nonetheless, these variables demonstrated statistically significant multivariate predictor status in analyses addressing PSA recurrence after RP.4 Presence of statistical significance in analyses of short-term end points such as PSA recurrence, and absence of significance in analyses of long-term outcomes such as PCA specific survival, suggest that these variables may not be important with regard to long-term cancer control. This hypothesis will require further analyses in other long-term followup surgical series. It is tempting to compare our data with watchful waiting series. However, direct comparisons need to be made with caution because biases and confounding variables may undermine the validity of such attempts. For example, decisions that contributed to treatment selection (prostatectomy vs observation) cannot be accounted for in such comparisons. Moreover, important differences may stem from pathological specimen evaluation. Watchful waiting series rely on biopsy samples whereas surgical series rely on radical prostatectomy specimens. Up to 48% of cases managed surgically may be upgraded at final pathology, which may confound the relation between treatment type and outcome.14 Moreover, the type of pathological assessment (local vs central review) and the chronological effect of Gleason sum assignment may contribute to further difficulties with valid comparisons.15 With these and other limitations in mind it is still noteworthy that 20-year RP actuarial cancer specific mortality rates are remarkably different from those for patients treated conservatively with respect to well differentiated and poorly differentiated disease. Albertson et al reported 20-year PCA specific mortality rates of 7% (Gleason 2 to 4), 14% (Gleason 5) and 66% (Gleason 8 to 10), while RP actuarial cancer specific mortality rates in our series were 4.5% (Gleason 2 to 5) and 35.5% (Gleason 8 to 10) (table 3).5 The remarkable survival difference for men with poorly differentiated disease suggests a benefit of surgery vs watchful waiting. These results are consistent with a prospective
572 CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
Kaplan-Meier estimates of overall prostate cancer specific survival (A), and stratified according to 7 main classifying variables (B–H) that were included in univariate and multivariate analyses for 752 patients who underwent radical prostatectomy at Virginia Mason Medical Center (1954 to 1994). Mean and median followup was 11.7 and 11.4 years, respectively (range 0.1 to 40.5). X-axis shows time since radical prostatectomy and y-axis shows recurrence-free probability. Solid line represents overall PCSS and PCSS stratified according to 7 main classifying variables. Broken lines represent 95% confidence intervals.
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
573
TABLE 3. Actuarial survival probabilities
Overall survival PCSS: Gleason 2–5 Gleason 6 Gleason 7 Gleason 8–10 Distant recurrence-free survival Local recurrence-free survival PSA recurrence-free survival
5-Yr Probability (CI)/ No. Pts at Risk
10-Yr Probability (CI)/ No. Pts at Risk
15-Yr Probability (CI)/ No. Pts at Risk
20-Yr Probability (CI)/ No. Pts at Risk
25-Yr Probability (CI)/ No. Pts at Risk
0.935 (0.914–0.951)/630 0.990 (0.979–0.995)/630 1.000 (1.000–1.000)/187 0.993 (0.972–0.998)/264 0.993 (0.949–0.999)/131 0.931 (0.826–0.973)/48 0.959 (0.942–0.972)/611
0.803 (0.769–0.832)/444 0.955 (0.934–0.969)/444 1.000 (1.000–1.000)/144 0.951 (0.915–0.972)/197 0.945 (0.881–0.975)/77 0.826 (0.677–0.911)/26 0.905 (0.879–0.926)/419
0.611 (0.565–0.653)/203 0.909 (0.875–0.934)/203 0.955 (0.880–0.984)/58 0.924 (0.876–0.954)/109 0.863 (0.744–0.929)/25 0.726 (0.523–0.854)/11 0.848 (0.811–0.879)/185
0.440 (0.386–0.493)/84 0.857 (0.803–0.898)/84 0.955 (0.880–0.984)/20 0.850 (0.768–0.905)/55 0.863 (0.744–0.929)/7 0.645 (0.400–0.811)/2 0.805 (0.754–0.846)/73
0.193 (0.133–0.261)/16 0.815 (0.729–0.876)/16 0.955 (0.880–0.984)/2 0.790 (0.664–0.873)/14 Not available Not available 0.782 (0.711–0.838)/15
0.953 (0.934–0.966)/601
0.912 (0.887–0.932)/409
0.896 (0.868–0.919)/178
0.878 (0.841–0.907)/74
0.878 (0.841–0.907)/12
0.848 (0.815–0.875)/445
0.712 (0.670–0.749)/288
0.611 (0.562–0.656)/115
0.586 (0.533–0.635)/55
0.545 (0.466–0.617)/9
randomized trial between RP and watchful waiting in which RP reduced cancer specific mortality by 50%.16 In our series men with Gleason 7 disease showed similar improvement in survival when treated with RP, demonstrated by prostate cancer specific mortality of 13.7% vs 45% with conservative therapy.5 Comparison of our data with previous surgical series is less complicated, and indicates that overall cancer specific survival rates at 5, 10 and 15 years are in close agreement with previous estimates.1–3 Han et al reported 15-year PCA specific survival rates of 90% which parallel our rate of 90.2%.2 It is interesting to note that these outcomes are virtually equivalent despite the fact that 64.5% of our patients were treated in the pre-PSA era, whereas the Han et al series is more recent and starts in 1982.2 We performed a detailed assessment of the effect of time on PCA specific survival. Univariate log rank analyses of the effect of treatment era on PCA specific survival (part H of figure) demonstrated a statistically significant difference (p ⫽ 0.048) in only 1 of 6 possible comparisons, when the earliest quartile (1954 to 1981) was compared to a more recent quartile (1992 to 1994). However, more importantly, lack of statistically significant chronological differences in PCA specific survival was found in univariate and multivariate analyses, where continuously coded year of surgery was examined (table 4). Taken together these findings suggest that within our cohort
TABLE 4. Prostate cancer specific survival after radical prostatectomy Univariate Analyses RR (p value) Age at surgery Surgery yr Pathological T stage (1992): T2b/T2c vs T2a T3 Vs T2a Surgical margin status: pos vs neg Pathological Gleason sum: 6 Vs 2–5 7 Vs 2–5 8–10 Vs 2–5 Type of surgical procedure: retropubic vs perineal Lymph node dissection: done vs not done Charlson comorbidity index: 1 Vs 0 2 or More vs 0 Hormonal therapy after surgery yes vs no Radiotherapy after surgery yes vs no
0.979 (0.352) 0.990 (0.594) — (⬍0.001) 1.022 (0.970) 4.619 (0.001) 2.594 (0.001) — (⬍0.001) 3.657 (0.016) 4.558 (0.010) 13.920 (⬍0.001) 0.712 (0.335) 0.695
(0.270)
— (0.843) 0.774 (0.723) 0.006 (0.641) 40.390 (⬍0.001) 5.346 (⬍0.001)
Multivariate Analyses RR (p value) 0.972 1.022 — 1.169 2.881 0.495 — 1.076 1.783 3.417 0.578
(0.283) (0.358) (0.093) (0.796) (0.055) (0.076) (0.033) (0.899) (0.343) (0.047) (0.396)
1.048
(0.935)
— (0.791) 0.655 (0.574) 0.002 (0.694) 30.623 (⬍0.001) 1.561
(0.180)
PCA specific survival has not changed over time. Only 1 of 3 long-term followup series assessed the effect of treatment era on PCA specific survival and found a significant univariate association.1 Three important methodological differences may account for the observed difference between the current study and the other report.1 First, the definition of era divided time into 2 broad categories of before and after PSA introduction, as opposed to a continuous variable in this analysis. Second, only 2% of the other cohort died of PCA vs 6.6% of ours. Finally, mean followup was 5.4 vs 11.7 years in our series. Nonetheless, further analyses are needed to assess temporal differences in PCA specific survival. Several weaknesses are associated with our study. The data represent outcomes from a single institution during 40 years involving 13 surgeons who performed 752 consecutive surgeries. Thus, patient selection and surgical technique may have differed and possibly may have introduced variation in cancer control rates.17 Furthermore, we combined 2 surgical techniques, namely RPP and RRP, which may also add variability. However, recent literature shows that there is no difference between RPP or RRP relative to oncological outcomes, which corroborates our findings.18 Another potential weakness of our analysis is that it encompasses the pre-PSA and early PSA eras. Therefore, it may be argued that men in our series had more adverse tumor characteristics than those undergoing RP in the PSA screening era.19 If so, our PCA specific survival rates underestimate what may be expected in men presenting with localized disease in the present. Moreover, our analysis relies upon the original pathology report from VMMC to record the histological differentiation. Patients before 1988 did not routinely have Gleason scores recorded and, therefore, we relied upon the paradigm used by Roehl et al to assign Gleason scores to
TABLE 5. Comparison of reported outcomes by institution after radical prostatectomy and watchful waiting No. Pts 996 Baylor College4 2,404 Johns Hopkins2 3 3,170 Mayo Clinic Northwestern University1 3,478 University of Connecticut 767 Health Center† Current series 752 * Recurrence-free survival. † Watchful waiting series.
Outcome Reported Study Yrs PCSS PSA* Local* Distant* 10 15 15 10 20
No Yes Yes Yes Yes
Yes Yes Yes Yes No
No Yes Yes No No
No Yes Yes No No
25
Yes
Yes
Yes
Yes
574
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
patients undergoing surgery before 1988.1 Interestingly the proportion of patients with Gleason scores 2 to 5, 6, 7 and 8 to 10 are remarkably similar to those reported by other large surgical series.1–3 Our report spans 4 decades of surgical treatment for localized PCA. In our analysis we have included, as have others, patients with adverse clinical and pathological features.1–3 In fact 15.2% of patients received adjuvant hormonal therapy and 15.7% received adjuvant radiation therapy. Careful review of the patient records have allowed us to categorize the patients with respect to comorbidity status at surgery using the weighted Charlson score. We found that 91% of our patients had no or only minor comorbidities. These factors should be taken in account when evaluating our long-term cancer specific survival rates. CONCLUSIONS This is the most mature of previously published large RP series.1–3 It demonstrates that long-term PSA recurrence outcomes after RP might be suboptimal because nearly 50% of men are at risk for PSA recurrence at 25 years after surgery. However, our data demonstrate excellent local and distant control as well as excellent cancer specific mortalityfree outcomes. On average 80% of patients who undergo RP will remain free of these adverse events 25 years after surgery.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Abbreviations and Acronyms AJCC PCA PCSS PLND PSA RP RPP RRP VMMC
⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽
American Joint Committee on Cancer prostate cancer prostate cancer specific survival pelvic lymph node dissection prostate specific antigen radical prostatectomy radical perineal prostatectomy radical retropubic prostatectomy Virginia Mason Medical Center
REFERENCES Roehl, K. A., Han, M., Ramos, C. G., Antenor, J. A. and Catolona, W. J.: Cancer progression and survival rates following anatomic radical retropubic prostatectomy in 3,478 consecutive patients: long-term results. J Urol, 172: 910, 2004 2. Han, M., Partin, A. W., Pound, C. R., Epstein, J. I. and Walsh, P. C.: Long-term biochemical disease-free and cancer specific survival following anatomic radical retropubic prostatectomy. The 15-year Johns Hopkins experience. Urol Clin North Am, 28: 555, 2001 3. Zincke, H., Oesterling, J. E., Blute, M. L., Bergstralh, E. J., Myers, R. P. and Barrett, D. M.: Longterm (15 years) results after radical prostatectomy for clinically localized (stage T2c or lower) prostate cancer. J Urol, 152: 1850, 1994
13.
14. 15.
16.
1.
17.
18.
19.
Hull, G. W., Rabbani, F., Abbas, F., Wheeler, T. M., Kattan, M. W. and Scardino, P. T.: Cancer control with radical prostatectomy alone in 1000 consecutive patients. J Urol, 167: 528, 2002 Albertson, P. C., Hanley, J. A. and Fine, J.: 20-Year outcomes following conservative management of clinically localized prostate cancer. JAMA, 293: 2095, 2005 Charlson, M. E., Pompei, P., Ales, K. L. and MacKenzie, C. R.: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40: 373, 1987 Gibbons, R. P.: Radical perineal prostatectomy. In: Campbell’s Urology, 7th ed. Edited by P. C. Walsh, A. B. Retik, E. D. Vaughan, Jr. and A. J. Wein. Philadelphia: W. B. Saunders Co., pp. 2589 –2588, 1998 Catalona, W. J.: Radical retropubic prostatectomy. In: Urologic Surgery, 4th ed. Edited by J. F. Glenn. Philadelphia: J. B. Lippincott Co., pp. 616 – 629, 1989 Carter, B. H. and Partin, A. W.: Prostate cancer staging systems. In: Campbell’s Urology, 7th ed. Edited by P. C. Walsh, A. B. Retik, E. D. Vaughan, Jr. and A. J. Wein. Philadelphia: W. B. Saunders Co., p. 2526, 1998 Beahrs, O. H., Henson, D. E., Hutter, R. V. P. and Kennedy, B. J.: American Joint Committee on Cancer: Manual for Staging of Cancer, 4th ed. Philadelphia: J. B. Lippincott Co., pp. 181–186, 1992 Gleason, D. F., Mellinger, G. T. and : the, Veterans Administration Cooperative Urological Research GroupPrediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol, 111: 58, 1974 Johansson, J. E., Andren, O., Andersson, S. O., Dickman, P. W., Holmberg, L., Magnuson, A. et al: Natural History of early, localized prostate cancer. JAMA, 291: 2713, 2004 Ward, J. F., Blute, M. L., Slezak, J., Bergstralh, E. J. and Zincke, H.: The long-term clinical impact of biochemical recurrence of prostate cancer 5 or more years after radical prostatectomy. J Urol, 170: 1872, 2003 King, C. R.: Patterns of prostate cancer biopsy grading: trends and clinical implications. Int J Cancer, 90: 305, 2000 Kondylis, F. I., Moriarty, R. P., Bostwick, D. and Schellhammer, P. F.: Prostate cancer grade assignment: the effect of chronological, interpretive and translational bias. J Urol, 170: 1189, 2003 Holmberg, L., Bill-Axelson, A., Helgesen, F., Salo, J. O., Folmerz, P., Haggman, M. et al: A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. N Engl J Med, 347: 781, 2002 Eastham, J. A., Kattan, M. W., Riedel, E., Begg, C. B., Wheeler, T. M., Gerigk, C. et al: Variations among individual surgeons in the rate of positive surgical margins in radical prostatectomy specimens. J Urol, 170: 2292, 2003 Salomon, L., Levrel, O., de la Taille, A., Anastasiadis, A. G., Saint, F., Zaki, S. et al: Radical prostatectomy by the retropubic, perineal and laparoscopic approach: 12 years of experience in one center. Eur Urol, 42: 104, 2002 Stamey, T. A., Caldwell, M., McNeal, J. E., Nolley, R., Hemenez, M. and Downs, J.: The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years? J Urol, 172: 1297, 2004
Purpose: We report on 25-year cancer control and survival outcomes after radical prostatectomy in a single center series of patients treated during a 40-year period. Materials and Methods: Between 1954 and 1994, 787 consecutive patients underwent radical prostatectomy at Virginia Mason Medical Center in Seattle, Washington. Kaplan-Meier 25-year probabilities of prostate cancer specific, overall, prostate specific antigen progression-free, local and distant progression-free survival were determined. Multivariate Cox regression models addressed prostate cancer specific mortality. Results: Prostate cancer specific survival, overall survival, prostate specific antigen progression-free survival, local and distant progression-free survival ranged from 99.0% to 81.5%, 93.5% to 19.3%, 84.8% to 54.5%, 95.3% to 87.8% and 95.9% to 78.2%, respectively. In univariate analyses pathological stage, surgical margin status, pathological Gleason sum, delivery of hormonal therapy and radiotherapy represented statistically significant predictors of prostate cancer specific mortality (all p ⱕ0.001). In multivariate analyses only Gleason sum (p ⫽ 0.03) and delivery of hormonal therapy (p ⬍0.001) remained significant. Conclusions: This is one of the most mature radical prostatectomy series. It demonstrates that long-term biochemical cancer control outcomes after radical prostatectomy might be suboptimal. However, local and distant control outcomes are excellent, and cancer specific mortality is minimal even 25 years after surgery. Key Words: prostatic neoplasms, prostatectomy, outcome assessment, survival analysis
neoadjuvant therapy. Clinical and pathological data were logged into a prospective database by one of the authors (RPG) from 1969 to 1999. In later years the records were maintained electronically with institutional review board approval. For the purposes of analyses 35 of 787 men were excluded due to missing age (9) and formal pathology (26), which left 752 evaluable records. Preoperative comorbidity was defined according to the Charlson weighted index of comorbidity.6 For purposes of analyses, the single patient with a weighted index of 4 was grouped with those with an index of 2. RPP was performed in 428 (56.9%) patients while 324 (43.1%) underwent RRP.7,8 PLND at surgery was performed in 328 patients (45.5%) who were selected according to individual surgeon preference and surgical approach. PLND limits consisted of the obturator nerve, external iliac vein, hypogastric vein and the inguinal ligament. PLND pathologies were not recorded in the database and could not be included in the present analyses. RP specimens were evaluated by VMCC pathologists. Specimens were processed as half mount specimens and serially sectioned at 5 mm intervals. Alternate 5 mm sections were fixed in formalin and were paraffin embedded. From 1954 to 1992 pathological tumor stage was recorded according to the Whitmore-Jewett classification and later converted to the 1992 AJCC staging guidelines.9,10 From 1992 to 1994 the 1992 AJCC staging system was used for clinical and pathological staging.
adical prostatectomy represents a treatment modality of choice for men with localized prostate cancer.1– 4 However, few series provide RP cancer control outcomes beyond 20 years.1–3 Conversely, a recent series of men treated conservatively was updated and provides 20-year cancer specific survival outcomes.5 Unfortunately the lack of long-term (median followup greater than 10 years) cancer specific survival data makes comparisons to long-term conservative series difficult. To address this limitation and to allow prognostication beyond the currently available 15-year cutoff, we updated the followup of our large RP series and report on 25-year outcomes in 787 consecutive men treated with RP at our institution between 1954 and 1994.
R
MATERIALS AND METHODS Between December 1954 and May 1994, 787 patients underwent RP at VMMC by 13 surgeons. No patient received
Submitted for publication September 2, 2005. * Correspondence: Department of Urologic Oncology, Virginia Mason Medical Center, Section of Urology and Renal Transplantation, Seattle, Washington 98101 (telephone: 206-223-6162; FAX: 206223-7650; e-mail: [email protected]). † Supported by the Fonds de la Recherche en Santé du Québec, the CHUM Foundation, the Department of Surgery and Les Urologues Associés du CHUM.
See Editorial on page 437.
0022-5347/06/1762-0569/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
569
Vol. 176, 569-574, August 2006 Printed in U.S.A. DOI:10.1016/j.juro.2006.03.094
570
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
From 1992 to 1994 tumors were routinely classified according to the Gleason grading system.11 Before 1992 tumor grade was recorded as well differentiated (I), moderately differentiated (II) and poorly differentiated (III). To recode these data we followed the paradigm outlined by Roehl et al, in which well differentiated tumors are classified as Gleason sum 3, moderately differentiated tumors are assigned Gleason sum 6 and poorly differentiated tumors are assigned Gleason sum 9.1 Positive surgical margins were recorded as presence of cancer cells against the inked resection margin. Serum PSA testing was initiated at VMMC in 1988. Since that time patients were followed at least quarterly for 2 years, then at least biannually for 2 years and then at least annually. Biochemical recurrence was defined as PSA greater than 0.1 ng/ml. Local pelvic disease or disease consistent with regional lymph node metastases was defined by tissue diagnosis or by overt radiographic signs of disease. Distant metastases were diagnosed based on radiographic and/or scintigraphic studies. Adjuvant or salvage hormonal and/or radiotherapy were delivered according to individual surgeon preference. Cause of death was ascertained according to detailed chart review or was obtained from the VMMC cancer registry. The cancer registry uses links with the Washington State Death Certificate Office. PCA must be the first listed cause of death on the certificate for a patient to be classified as having died of PCA. Actuarial analyses addressed the outcomes PSA recurrence, local recurrence, distant recurrence, PCSS and overall survival. In analyses of PCSS, patients without evidence of progression were censored at the time of last followup. Patients who died of other causes were censored at time of death. PSA recurrence-free data were censored if radiotherapy and/or hormonal therapy were delivered before PSA recurrence. PCA specific mortality was also analyzed with univariate and multivariate Cox regression models. Predictors included patient age at surgery, year of surgery, 1992 AJCC pathological stage, surgical margin status, pathological Gleason sum, type of surgery (RPP vs RRP), lymph node dissection status (performed vs not), delivery of hormonal therapy and delivery of radiation therapy. All statistical tests were performed with S-Plus® Professional and statistical significance was set at 0.05. RESULTS Descriptive statistics are shown in table 1. Mean patient age was 63.6 years (median 64.4, range 42.9 to 77.9). In 341 (45%) patients PCA was nonorgan confined. One or multiple positive surgical margins were recorded in 283 specimens (37.6%). Gleason 7 or higher was recorded in 233 (31%) specimens. Radiotherapy was delivered to 118 (15.7%) and hormonal therapy was given to 114 (15.2%) patients. Followup ranged from 0.1 to 40.5 years (mean 11.7, median 11.4). The number of men at risk according to followup is shown in table 2. PSA recurrence was noted in 189 (31.4% of 601 evaluable patients), local recurrence was noted in 63 (8.4%) and distant recurrence was noted in 85 (11.3%). Overall 298 men died (39.6%) and of these 50 died of PCA (6.6%). Part A of the figure illustrates PCA specific survival. Parts B to H of the figure show PCA specific survival stratified according to 7 main classifying variables that were included in univariate and multivariate analyses.
TABLE 1. Characteristics of 752 patients undergoing radical prostatectomy at the Virginia Mason Medical Center between 1954 and 1994 No. Pts (%) Clinical stage (1992): T1 T2a T2b or T2c T3 Pathological T stage (1992): T2a T2b or T2c T3 Pos surgical margins Pathological Gleason sum: 2 3 4 5 6 7 8 9 10 Surgical procedure: Perineal Retropubic Lymph node dissection Weighted Charlson comorbidity index:* 0 1 2 4 Hormonal therapy Radiotherapy: Adjuvant Salvage† Overall recurrence: PSA‡ Local Distant Overall mortality: PCa specific Other cause
53 (7.1) 515 (68.5) 180 (23.9) 4 (0.5) 177 (23.5) 234 (31.1) 341 (45.3) 283 (37.6) 3 (0.4) 20 (2.7) 50 (6.6) 147 (19.5) 299 (39.8) 170 (22.6) 38 (5.1) 22 (2.9) 3 (0.4) 428 (56.9) 324 (43.1) 342 (45.5) 685 (91.1) 43 (5.7) 23 (3.1) 1 (0.1) 114 (15.2) 192 (25.5) 118 (15.7) 75 (10.0) 189 (31.4) 63 (8.4) 85 (11.3) 298 (39.6) 50 (6.6) 248 (33.0)
* Charlson score of 0 to 1 indicates no or minor comorbidities and score of at least 2 indicates significant comorbidities. † One case has adjuvant and salvage, and is set as adjuvant. ‡ Data set of 601 patients.
Table 3 shows actuarial survival estimates according to 5 definitions— overall survival, PCA specific survival, distant recurrence-free survival, local recurrence-free survival and PSA recurrence-free survival. PCA specific survival is further stratified according to pathological Gleason sum. Overall survival ranged from 93.5% at 5 years to 19.3% at 25 years. PCA specific survival ranged from 99.0% at 5 years to 81.5% at 25 years. Distant recurrence-free survival ranged from 95.9% to 78.2% at 5 and 25 years, respectively. Local progression-free survival ranged from 95.3% to 87.8% at 5 and 25 years, respectively. Finally PSA progression-free survival ranged from 84.8% to 54.5% at 5 and 25 years, respectively. Table 4 shows univariate and multivariate Cox regression analyses predicting PCA specific mortality. In univariate analyses pathological stage (p ⬍0.001), surgical margin status (p ⫽ 0.001), Gleason sum (p ⬍0.001), delivery of radiotherapy (p ⬍0.001) and delivery of hormonal therapy (p ⬍0.001) represented statistically significant predictors. In multivariate analyses only Gleason sum (p ⫽ 0.033) and delivery of hormonal therapy (p ⬍0.001) remained statistically significant.
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY TABLE 2. Number of 752 patients at risk after radical prostatectomy according to year of followup Yrs Followup
No. Pts Current Series (%)
No. Pts Han et al Series2 (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
752 (100.0) 719 (95.6) 701 (93.2) 679 (90.3) 660 (87.8) 630 (83.8) 603 (80.2) 570 (75.8) 526 (69.9) 489 (65.0) 444 (59.0) 404 (53.7) 342 (45.5) 294 (39.1) 246 (32.7) 203 (27.0) 167 (22.2) 143 (19.0) 123 (16.4) 108 (14.4) 84 (11.2) 63 (8.4) 49 (6.5) 36 (4.8) 25 (3.3) 16 (2.1) 12 (1.6) 11 (1.5) 9 (1.2) 6 (0.8) 5 (0.7) 4 (0.5) 3 (0.4) 3 (0.4) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1) 1 (0.1)
2,404 (100.0) 2,093 (87.0) 1,846 (77.0) 1,612 (67.0) 1,396 (58.0) 1,186 (49.0) 1,038 (43.0) 888 (37.0) 746 (31.0) 621 (26.0) 479 (20.0) 343 (14.0) 227 (9.0) 145 (6.0) 78 (3.0) 29 (1.0) 7 (0.3)
DISCUSSION Men with localized PCA have many treatment options ranging from radical surgery to watchful waiting.1–5 Randomized trials comparing long-term outcomes (median outcome more than 10 years) according to treatment type are unavailable. Therefore, long-term outcome data are sorely needed to evaluate treatment results. Long-term outcomes from conservative approaches for localized PCA were the subject of 2 recent publications.5,12 Both reports indicated that men with well differentiated disease have excellent PCA specific survival while men with poorly differentiated disease are likely to die of the disease within 10 years of diagnosis.5,12 Unfortunately 20-year and longer PCA specific RP outcomes are not available.1–3 To address this limitation we report long-term outcomes in 752 men who underwent RP between 1954 and 1994 at a single institution. Our analysis indicates that long-term cancer control outcomes are excellent when PCA specific survival, distant recurrence-free survival and local recurrence-free survival are assessed up to 25 years after RP (tables 3 and 5). For these outcomes eventfree rates at 25 years are 81.5%, 78.2% and 87.8%, respectively. Conversely, assessment of PSA recurrence-free rates demonstrates a steady decrease in event-free rates which change from 84.8% at 5 years, to 61.1% at 15 years, to 54.5% at 25 years. Taken together these data indicate that RP
571
offers excellent cancer control rates which are remarkably stable over time. The exception is PSA recurrence-free survival, which decreases steadily over time. However, PSA recurrence rates in these series have limited impact on longterm prognosis measured by PCA specific survival, distant recurrence-free survival and local recurrence-free survival. Our PSA observations are in agreement with previous findings.13 Based on median followup of 10 years, Ward et al reported that only 8.3% of men with PSA recurrence after RP died of PCA, which represented a mortality figure of 2.7% in a cohort of 3,897 evaluable RP cases.13 Therefore, it may be concluded that PSA recurrence is associated with PCA specific mortality. However, only a small proportion of men with PSA recurrence actually die of PCA. At 25 years after RP, PCA specific survival was 82.2%. When PCA specific survival data were stratified according to pathological Gleason sum, 20-year survival rates were noted as 64.5% for Gleason 8 to 10, 86.3% for Gleason 7, 85.0% for Gleason 6 and 95.5% for Gleason 2 to 5. Our multivariate analysis substantiated that pathological Gleason represents a statistically significant multivariate predictor of PCA specific survival as shown in previous reports.1–3 It is of interest that other variables such as pathological stage and surgical margin status did not reach statistical significance in multivariate PCA specific survival analyses. The multivariate effect of these variables on PCA specific survival was not assessed in other long-term outcome series.1–3 Nonetheless, these variables demonstrated statistically significant multivariate predictor status in analyses addressing PSA recurrence after RP.4 Presence of statistical significance in analyses of short-term end points such as PSA recurrence, and absence of significance in analyses of long-term outcomes such as PCA specific survival, suggest that these variables may not be important with regard to long-term cancer control. This hypothesis will require further analyses in other long-term followup surgical series. It is tempting to compare our data with watchful waiting series. However, direct comparisons need to be made with caution because biases and confounding variables may undermine the validity of such attempts. For example, decisions that contributed to treatment selection (prostatectomy vs observation) cannot be accounted for in such comparisons. Moreover, important differences may stem from pathological specimen evaluation. Watchful waiting series rely on biopsy samples whereas surgical series rely on radical prostatectomy specimens. Up to 48% of cases managed surgically may be upgraded at final pathology, which may confound the relation between treatment type and outcome.14 Moreover, the type of pathological assessment (local vs central review) and the chronological effect of Gleason sum assignment may contribute to further difficulties with valid comparisons.15 With these and other limitations in mind it is still noteworthy that 20-year RP actuarial cancer specific mortality rates are remarkably different from those for patients treated conservatively with respect to well differentiated and poorly differentiated disease. Albertson et al reported 20-year PCA specific mortality rates of 7% (Gleason 2 to 4), 14% (Gleason 5) and 66% (Gleason 8 to 10), while RP actuarial cancer specific mortality rates in our series were 4.5% (Gleason 2 to 5) and 35.5% (Gleason 8 to 10) (table 3).5 The remarkable survival difference for men with poorly differentiated disease suggests a benefit of surgery vs watchful waiting. These results are consistent with a prospective
572 CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
Kaplan-Meier estimates of overall prostate cancer specific survival (A), and stratified according to 7 main classifying variables (B–H) that were included in univariate and multivariate analyses for 752 patients who underwent radical prostatectomy at Virginia Mason Medical Center (1954 to 1994). Mean and median followup was 11.7 and 11.4 years, respectively (range 0.1 to 40.5). X-axis shows time since radical prostatectomy and y-axis shows recurrence-free probability. Solid line represents overall PCSS and PCSS stratified according to 7 main classifying variables. Broken lines represent 95% confidence intervals.
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
573
TABLE 3. Actuarial survival probabilities
Overall survival PCSS: Gleason 2–5 Gleason 6 Gleason 7 Gleason 8–10 Distant recurrence-free survival Local recurrence-free survival PSA recurrence-free survival
5-Yr Probability (CI)/ No. Pts at Risk
10-Yr Probability (CI)/ No. Pts at Risk
15-Yr Probability (CI)/ No. Pts at Risk
20-Yr Probability (CI)/ No. Pts at Risk
25-Yr Probability (CI)/ No. Pts at Risk
0.935 (0.914–0.951)/630 0.990 (0.979–0.995)/630 1.000 (1.000–1.000)/187 0.993 (0.972–0.998)/264 0.993 (0.949–0.999)/131 0.931 (0.826–0.973)/48 0.959 (0.942–0.972)/611
0.803 (0.769–0.832)/444 0.955 (0.934–0.969)/444 1.000 (1.000–1.000)/144 0.951 (0.915–0.972)/197 0.945 (0.881–0.975)/77 0.826 (0.677–0.911)/26 0.905 (0.879–0.926)/419
0.611 (0.565–0.653)/203 0.909 (0.875–0.934)/203 0.955 (0.880–0.984)/58 0.924 (0.876–0.954)/109 0.863 (0.744–0.929)/25 0.726 (0.523–0.854)/11 0.848 (0.811–0.879)/185
0.440 (0.386–0.493)/84 0.857 (0.803–0.898)/84 0.955 (0.880–0.984)/20 0.850 (0.768–0.905)/55 0.863 (0.744–0.929)/7 0.645 (0.400–0.811)/2 0.805 (0.754–0.846)/73
0.193 (0.133–0.261)/16 0.815 (0.729–0.876)/16 0.955 (0.880–0.984)/2 0.790 (0.664–0.873)/14 Not available Not available 0.782 (0.711–0.838)/15
0.953 (0.934–0.966)/601
0.912 (0.887–0.932)/409
0.896 (0.868–0.919)/178
0.878 (0.841–0.907)/74
0.878 (0.841–0.907)/12
0.848 (0.815–0.875)/445
0.712 (0.670–0.749)/288
0.611 (0.562–0.656)/115
0.586 (0.533–0.635)/55
0.545 (0.466–0.617)/9
randomized trial between RP and watchful waiting in which RP reduced cancer specific mortality by 50%.16 In our series men with Gleason 7 disease showed similar improvement in survival when treated with RP, demonstrated by prostate cancer specific mortality of 13.7% vs 45% with conservative therapy.5 Comparison of our data with previous surgical series is less complicated, and indicates that overall cancer specific survival rates at 5, 10 and 15 years are in close agreement with previous estimates.1–3 Han et al reported 15-year PCA specific survival rates of 90% which parallel our rate of 90.2%.2 It is interesting to note that these outcomes are virtually equivalent despite the fact that 64.5% of our patients were treated in the pre-PSA era, whereas the Han et al series is more recent and starts in 1982.2 We performed a detailed assessment of the effect of time on PCA specific survival. Univariate log rank analyses of the effect of treatment era on PCA specific survival (part H of figure) demonstrated a statistically significant difference (p ⫽ 0.048) in only 1 of 6 possible comparisons, when the earliest quartile (1954 to 1981) was compared to a more recent quartile (1992 to 1994). However, more importantly, lack of statistically significant chronological differences in PCA specific survival was found in univariate and multivariate analyses, where continuously coded year of surgery was examined (table 4). Taken together these findings suggest that within our cohort
TABLE 4. Prostate cancer specific survival after radical prostatectomy Univariate Analyses RR (p value) Age at surgery Surgery yr Pathological T stage (1992): T2b/T2c vs T2a T3 Vs T2a Surgical margin status: pos vs neg Pathological Gleason sum: 6 Vs 2–5 7 Vs 2–5 8–10 Vs 2–5 Type of surgical procedure: retropubic vs perineal Lymph node dissection: done vs not done Charlson comorbidity index: 1 Vs 0 2 or More vs 0 Hormonal therapy after surgery yes vs no Radiotherapy after surgery yes vs no
0.979 (0.352) 0.990 (0.594) — (⬍0.001) 1.022 (0.970) 4.619 (0.001) 2.594 (0.001) — (⬍0.001) 3.657 (0.016) 4.558 (0.010) 13.920 (⬍0.001) 0.712 (0.335) 0.695
(0.270)
— (0.843) 0.774 (0.723) 0.006 (0.641) 40.390 (⬍0.001) 5.346 (⬍0.001)
Multivariate Analyses RR (p value) 0.972 1.022 — 1.169 2.881 0.495 — 1.076 1.783 3.417 0.578
(0.283) (0.358) (0.093) (0.796) (0.055) (0.076) (0.033) (0.899) (0.343) (0.047) (0.396)
1.048
(0.935)
— (0.791) 0.655 (0.574) 0.002 (0.694) 30.623 (⬍0.001) 1.561
(0.180)
PCA specific survival has not changed over time. Only 1 of 3 long-term followup series assessed the effect of treatment era on PCA specific survival and found a significant univariate association.1 Three important methodological differences may account for the observed difference between the current study and the other report.1 First, the definition of era divided time into 2 broad categories of before and after PSA introduction, as opposed to a continuous variable in this analysis. Second, only 2% of the other cohort died of PCA vs 6.6% of ours. Finally, mean followup was 5.4 vs 11.7 years in our series. Nonetheless, further analyses are needed to assess temporal differences in PCA specific survival. Several weaknesses are associated with our study. The data represent outcomes from a single institution during 40 years involving 13 surgeons who performed 752 consecutive surgeries. Thus, patient selection and surgical technique may have differed and possibly may have introduced variation in cancer control rates.17 Furthermore, we combined 2 surgical techniques, namely RPP and RRP, which may also add variability. However, recent literature shows that there is no difference between RPP or RRP relative to oncological outcomes, which corroborates our findings.18 Another potential weakness of our analysis is that it encompasses the pre-PSA and early PSA eras. Therefore, it may be argued that men in our series had more adverse tumor characteristics than those undergoing RP in the PSA screening era.19 If so, our PCA specific survival rates underestimate what may be expected in men presenting with localized disease in the present. Moreover, our analysis relies upon the original pathology report from VMMC to record the histological differentiation. Patients before 1988 did not routinely have Gleason scores recorded and, therefore, we relied upon the paradigm used by Roehl et al to assign Gleason scores to
TABLE 5. Comparison of reported outcomes by institution after radical prostatectomy and watchful waiting No. Pts 996 Baylor College4 2,404 Johns Hopkins2 3 3,170 Mayo Clinic Northwestern University1 3,478 University of Connecticut 767 Health Center† Current series 752 * Recurrence-free survival. † Watchful waiting series.
Outcome Reported Study Yrs PCSS PSA* Local* Distant* 10 15 15 10 20
No Yes Yes Yes Yes
Yes Yes Yes Yes No
No Yes Yes No No
No Yes Yes No No
25
Yes
Yes
Yes
Yes
574
CANCER SPECIFIC SURVIVAL AFTER RADICAL PROSTATECTOMY
patients undergoing surgery before 1988.1 Interestingly the proportion of patients with Gleason scores 2 to 5, 6, 7 and 8 to 10 are remarkably similar to those reported by other large surgical series.1–3 Our report spans 4 decades of surgical treatment for localized PCA. In our analysis we have included, as have others, patients with adverse clinical and pathological features.1–3 In fact 15.2% of patients received adjuvant hormonal therapy and 15.7% received adjuvant radiation therapy. Careful review of the patient records have allowed us to categorize the patients with respect to comorbidity status at surgery using the weighted Charlson score. We found that 91% of our patients had no or only minor comorbidities. These factors should be taken in account when evaluating our long-term cancer specific survival rates. CONCLUSIONS This is the most mature of previously published large RP series.1–3 It demonstrates that long-term PSA recurrence outcomes after RP might be suboptimal because nearly 50% of men are at risk for PSA recurrence at 25 years after surgery. However, our data demonstrate excellent local and distant control as well as excellent cancer specific mortalityfree outcomes. On average 80% of patients who undergo RP will remain free of these adverse events 25 years after surgery.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Abbreviations and Acronyms AJCC PCA PCSS PLND PSA RP RPP RRP VMMC
⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽
American Joint Committee on Cancer prostate cancer prostate cancer specific survival pelvic lymph node dissection prostate specific antigen radical prostatectomy radical perineal prostatectomy radical retropubic prostatectomy Virginia Mason Medical Center
REFERENCES Roehl, K. A., Han, M., Ramos, C. G., Antenor, J. A. and Catolona, W. J.: Cancer progression and survival rates following anatomic radical retropubic prostatectomy in 3,478 consecutive patients: long-term results. J Urol, 172: 910, 2004 2. Han, M., Partin, A. W., Pound, C. R., Epstein, J. I. and Walsh, P. C.: Long-term biochemical disease-free and cancer specific survival following anatomic radical retropubic prostatectomy. The 15-year Johns Hopkins experience. Urol Clin North Am, 28: 555, 2001 3. Zincke, H., Oesterling, J. E., Blute, M. L., Bergstralh, E. J., Myers, R. P. and Barrett, D. M.: Longterm (15 years) results after radical prostatectomy for clinically localized (stage T2c or lower) prostate cancer. J Urol, 152: 1850, 1994
13.
14. 15.
16.
1.
17.
18.
19.
Hull, G. W., Rabbani, F., Abbas, F., Wheeler, T. M., Kattan, M. W. and Scardino, P. T.: Cancer control with radical prostatectomy alone in 1000 consecutive patients. J Urol, 167: 528, 2002 Albertson, P. C., Hanley, J. A. and Fine, J.: 20-Year outcomes following conservative management of clinically localized prostate cancer. JAMA, 293: 2095, 2005 Charlson, M. E., Pompei, P., Ales, K. L. and MacKenzie, C. R.: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40: 373, 1987 Gibbons, R. P.: Radical perineal prostatectomy. In: Campbell’s Urology, 7th ed. Edited by P. C. Walsh, A. B. Retik, E. D. Vaughan, Jr. and A. J. Wein. Philadelphia: W. B. Saunders Co., pp. 2589 –2588, 1998 Catalona, W. J.: Radical retropubic prostatectomy. In: Urologic Surgery, 4th ed. Edited by J. F. Glenn. Philadelphia: J. B. Lippincott Co., pp. 616 – 629, 1989 Carter, B. H. and Partin, A. W.: Prostate cancer staging systems. In: Campbell’s Urology, 7th ed. Edited by P. C. Walsh, A. B. Retik, E. D. Vaughan, Jr. and A. J. Wein. Philadelphia: W. B. Saunders Co., p. 2526, 1998 Beahrs, O. H., Henson, D. E., Hutter, R. V. P. and Kennedy, B. J.: American Joint Committee on Cancer: Manual for Staging of Cancer, 4th ed. Philadelphia: J. B. Lippincott Co., pp. 181–186, 1992 Gleason, D. F., Mellinger, G. T. and : the, Veterans Administration Cooperative Urological Research GroupPrediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol, 111: 58, 1974 Johansson, J. E., Andren, O., Andersson, S. O., Dickman, P. W., Holmberg, L., Magnuson, A. et al: Natural History of early, localized prostate cancer. JAMA, 291: 2713, 2004 Ward, J. F., Blute, M. L., Slezak, J., Bergstralh, E. J. and Zincke, H.: The long-term clinical impact of biochemical recurrence of prostate cancer 5 or more years after radical prostatectomy. J Urol, 170: 1872, 2003 King, C. R.: Patterns of prostate cancer biopsy grading: trends and clinical implications. Int J Cancer, 90: 305, 2000 Kondylis, F. I., Moriarty, R. P., Bostwick, D. and Schellhammer, P. F.: Prostate cancer grade assignment: the effect of chronological, interpretive and translational bias. J Urol, 170: 1189, 2003 Holmberg, L., Bill-Axelson, A., Helgesen, F., Salo, J. O., Folmerz, P., Haggman, M. et al: A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. N Engl J Med, 347: 781, 2002 Eastham, J. A., Kattan, M. W., Riedel, E., Begg, C. B., Wheeler, T. M., Gerigk, C. et al: Variations among individual surgeons in the rate of positive surgical margins in radical prostatectomy specimens. J Urol, 170: 2292, 2003 Salomon, L., Levrel, O., de la Taille, A., Anastasiadis, A. G., Saint, F., Zaki, S. et al: Radical prostatectomy by the retropubic, perineal and laparoscopic approach: 12 years of experience in one center. Eur Urol, 42: 104, 2002 Stamey, T. A., Caldwell, M., McNeal, J. E., Nolley, R., Hemenez, M. and Downs, J.: The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years? J Urol, 172: 1297, 2004