Pathological Stage T2 Subgroups to Predict Biochemical Recurrence After Prostatectomy

Pathological Stage T2 Subgroups to Predict Biochemical Recurrence After Prostatectomy Yakup Kordan, Sam S. Chang,* Shady Salem, Michael S. Cookson, Peter E. Clark,† Rodney Davis, S. Duke Herrell, Roxelyn Baumgartner, Sharon Phillips, Joseph A. Smith, Jr. and Daniel A. Barocas‡ From the Departments of Urologic Surgery and Biostatistics (SP), Vanderbilt University Medical Center, Nashville, Tennessee

Purpose: We evaluated whether the 2002 TNM substages of pathological T2 prostate cancer predict intermediate term biochemical recurrence-free survival. Materials and Methods: The cohort consisted of men who underwent radical prostatectomy between January 2000 and June 2008, and had pT2 disease at final pathological evaluation. We excluded patients with prior treatment, less than 6 months of followup or missing data, leaving 1,370 available for analysis, including 340 with pT2a, 35 with pT2b and 995 with pT2c disease. Clinical and pathological characteristics were compared between groups using univariate analysis. Biochemical recurrence-free survival was compared between substages using Kaplan-Meier analysis. A Cox proportional hazards model was used to evaluate tumor substage as a biochemical recurrence-free survival predictor. Results: Median followup was 21 months. No differences were seen in the likelihood of biochemical recurrence-free survival between T2 subclasses (p ⫽ 0.174). No patient with T2b disease had recurrence. The 3 and 5-year likelihood of freedom from biochemical recurrence was 95.5% (95% CI 90.9 – 97.8) and 93.8% (95% CI 87.3–97.0) for pT2a, and 94.3% (95% CI 91.8 –96.0) and 87.5% (95% CI 82.7–91.1) for pT2c, respectively. Multivariate analysis showed that significant predictors of biochemical recurrence-free survival were margin status (HR 2.7, 95% CI 1.3–5.5, p ⫽ 0.006), preoperative prostate specific antigen (HR 1.0, 95% CI 1.0 –1.1, p ⫽ 0.029), pathological Gleason score 7 (HR 2.5, 95% CI 1.1–5.7, p ⫽ 0.024) and pathological Gleason score 8 –10 (HR 6.2, 95% CI 2.2–17.4, p ⬍0.001). Compared to pathological stage T2a neither pT2b nor pT2c predicted biochemical recurrence-free survival (p ⫽ 0.99 and 0.42, respectively). Conclusions: Current pT2 prostate cancer substages may not have prognostic significance for intermediate term outcomes. If borne out during longer followup, future staging systems may collapse the substages into a single category.

Abbreviations and Acronyms AJCC ⫽ American Joint Committee for Cancer Staging and End Result Reporting BCR ⫽ biochemical recurrence PCa ⫽ prostate carcinoma PSA ⫽ prostate specific antigen RP ⫽ radical prostatectomy RRP ⫽ radical retropubic prostatectomy Submitted for publication February 25, 2009. Study received institutional review board approval. * Financial interest and/or other relationship with Sanofi-Aventis and Amgen † Financial interest and/or other relationship with Tension and Galil. ‡ Correspondence: Department of Urologic Surgery, Vanderbilt University Medical Center, A 1302 Medical Center North, Nashville, Tennessee 37232-2765 (telephone: 615-322-2101; FAX: 615322-8990; e-mail: [email protected]).

Key Words: prostate, prostatic neoplasms, neoplasm staging, prostate-specific antigen, prostatectomy EARLY PCa detection has led to significant stage migration and, thus, to an increased proportion of men diagnosed with clinically localized disease.1,2 However, 4% to 32% of men with pathologically organ confined disease have recurrence for

reasons that remain unclear. 3– 6 Several researchers have identified factors that predict recurrence in localized PCa cases, such as preoperative PSA, clinical stage, Gleason score and surgical margin status.7,8 While pathological staging is consid-

0022-5347/09/1825-2291/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 182, 2291-2295, November 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.07.020

www.jurology.com

2291

2292

PATHOLOGICAL STAGE T2 SUBGROUPS AND BIOCHEMICAL RECURRENCE

ered an important prognostic factor in many cancers, it is not clear whether pathologically organ confined PCa substages add value to these predictive models. Accurate pathological staging is essential to determine the likelihood of recurrence, guide management and standardize reporting. The TNM staging system was first proposed for PCa staging in 1975, and in 1992 AJCC and UICC adopted a new system.9,10 Since then, TNM classification has gained general acceptance for clinical and pathological staging.11 According to the 1992 TNM staging system T2 disease is defined as a palpable, prostate confined tumor and subdivided into 3 categories, including T2a— unilateral tumor involving less than half of a lobe, T2b— unilateral tumor involving greater than half of a lobe and T2c— bilateral tumor. In 1997 a revision was done in which unilateral tumors (formerly T2a and T2b) were merged into 1 category (T2a) and T2c was changed to T2b.12 Subsequent research showed significantly different BCR-free survival in 1992 clinical T2a vs cT2b cases but similar BCR-free survival in cT2b and cT2c cases after prostatectomy.13–15 As a result, in 2002 AJCC revised the TNM staging system and reverted to the previous 1992 classification of T2 disease.16 While clinical staging is based on digital rectal examination findings, pathological staging of PCa is determined by histological identification of the tumor extent in the prostate and surrounding tissues. As anticipated, there is a significant discrepancy between clinical and pathological staging, and only 36% to 41% of staging correlates.17,18 Moreover, all mentioned studies mainly focused on PCa clinical staging to predict the future outcome after RRP. 8 –10 Nevertheless, there are sparse published data on whether the T2 PCa pathological subclassification carries prognostic value similar to that of its clinical counterpart. 3,8,11,19 –22 Recent reports questioned the prognostic significance of pathological substaging, suggesting that current 2002 TNM pathological T2 substages may not confer any prognostic value for predicting BCR after RP. Thus, we evaluated whether pathological T2 subcategories (TNM 2002) have prognostic significance after RP with regard to BCR-free survival.

MATERIALS AND METHODS

and pathological data were collected prospectively and stored in the institutional review board approved Vanderbilt Prostatectomy Database. Medical charts were examined for clinical followup data so that only patients with followup at our institution are represented. Patients who received prior treatment, eg radiation therapy, chemotherapy or hormonal therapy (165), or had less than 6 months of followup (1,037) or missing data (124) were excluded from analysis. Of the remaining 1,764 patients 1,371 (77.7%) had pathological stage T2 disease. To isolate the effect of stage the single patient with pathological T2 disease and positive lymph nodes was excluded from analysis. Thus, the final sample included 1,370 patients who underwent open (630) or robot assisted (740) RP. All surgeries were done at our institution by 1 of 7 urological oncologists. There was no predetermined followup regimen in all patients but generally they were followed by routine evaluation and PSA monitoring every 6 months in the first 2 years and annually thereafter. BCR was defined as postoperative PSA greater than 0.2 ng/ml confirmed on 1 or more subsequent assays, or a PSA increase and adjuvant therapy.

Pathological Evaluation Prostates were measured and weighed before fixation in formalin. External surfaces were inked. Before May 2003 RP specimens were evaluated according to whole mount protocol described by Jack et al.23 Thereafter all RP specimens were step sectioned, and systematically sampled and evaluated according to a standard protocol, as described by Srigley.24 In whole mount processed prostates tumor volume was determined by planimetry and in systematic sampling tumor volume was estimated by multiplying the estimated percent of tumor involvement of the submitted prostate portions times prostate size. In all cases Gleason grade was assigned and cases were subclassified by pT2 subgroup using 2002 TNM staging criteria,16 including pT2a in 340, pT2b in 35 and pT2c in 995.

Statistical Analysis Clinical and pathological characteristics were compared between groups using the Kruskal-Wallis and Fisher exact tests as appropriate. The detection rate of pathological T2 substages was compared across pathological processing methods. BCR was compared between substages using Kaplan-Meier analysis. A multivariate Cox proportional hazards model was used to evaluate tumor substage as a possible predictor of BCR. Included in this model were preoperative serum PSA, pathological Gleason score, surgical margin status and tumor volume since each is associated with outcome. Also, we controlled for pathological sampling technique since this is a potential confounder. All statistical analysis was done using Stata® 10.0.

Patient Population The population consisted of 3,090 men with biopsy proven, clinically localized PCa who underwent RP between January 2000 and June 2008 at our institution. Staging lymphadenectomy was done according to surgeon preference and 98.5% of patients underwent lymphadenectomy. Clinical, demographic, perioperative

RESULTS Table 1 lists patient clinicopathological characteristics. There were no statistically significant differences between the groups in age at surgery, PSA, clinical stage or prostate size. There was a sig-

Table 1. Clinicopathological features in 1,370 men with RP for pT2 PCa Characteristic

pT2A

pT2B

pT2C

Mean ⫾ SD age Median PSA ng/mo (IQR) No. clinical stage (%): Nonpalpable Palpable No. biopsy Gleason score (%): 6 or Less 7 8–10 Median gm prostate size (IQR) Median gm estimated tumor vol (IQR) No. pos surgical margin (%) No. pathological Gleason score (%): 6 or Less 7 8–10

60.7 (7.3) 5.2 (2.5)

60.0 (7.1) 5.6 (3.6)

60.1 (7.5) 5.3 (2.9)

261 (77.0) 78 (23.0)

27 (77.1) 8 (22.9)

768 (78.5) 210 (21.5)

p Value 0.480 0.410 0.876

0.003 265 (79.1) 61 (18.2) 9 (2.7) 45.5 (28.2)

17 (48.6) 15 (42.9) 3 (8.6) 45.4 (20.4)

753 (76.2) 199 (20.1) 36 (3.6) 43.2 (21.2)

0.443

1.6 (2.3)

4.6 (4.8)

3.5 (4.4)

⬍0.001

5 (14.3)

193 (19.4)

⬍0.001

32

(9.4)

0.006 230 (68.5) 94 (28.0) 12 (3.6)

16 (45.7) 17 (48.6) 2 (5.7)

580 (58.5) 367 (37.0) 44 (4.4)

nificant difference in the rate of positive surgical margins across groups, including 9.4% for pT2a, 14.3% for pT2b and 19.4% for pT2c (p ⬍0.001). There were also significant differences in biopsy Gleason score, estimated tumor volume and pathological Gleason score (table 1). There were significant differences in pathological substage distributions according to pathological sampling method, suggesting differences in the detection of substages by sampling method or differences in the underlying population with time. When whole mount processing was used, pT2b disease was unlikely to be detected (1 of 360 cases or 0.3% vs 34 of 1,010 or 3.4%, p ⬍0.001). However, while using the whole mount and systemic sampling techniques, 122 (33.9%) and 218 (21.6%) of pT2a, and 237 (65.8%) and 758 (75.0%) of pT2c cases were detected, respectively. Patients with pT2b disease had higher biopsy Gleason scores than those with pT2a or pT2c (Gleason 7 or greater in 42.9% vs 18.2% and 20.1%, respectively). They also had higher pathological Gleason scores (Gleason 7 or greater in 48.6% vs 28% and 37%, respectively). The largest median tumor volume was in the pT2b group, that is 1.6 cc for pT2a, 4.6 cc for pT2b and 3.5 cc for pT2c (p ⬍0.001). Median followup was 21.2 months (IQR 11.6–38.7). No patient with T2b disease had recurrence but in the pT2a subgroup 12 (3.5%) and in the pT2c group 45 (4.5%) had recurrence. No BCR likelihood differences were seen between T2 subclasses (log rank test p ⫽ 0.174, see figure). The 3 and 5-year likelihood of freedom from BCR was 95.5% (95% CI 90.9–97.8) and 93.8% (95% CI 87.3–

Proportion free of biochemical recurrence

PATHOLOGICAL STAGE T2 SUBGROUPS AND BIOCHEMICAL RECURRENCE

2293

BCR of Patients with PT2 Disease, by Sub-Stage 1.00 0.75 0.50 Log-rank p = 0.174 0.25 0.00 0

Number at risk T2A 340 T2B 35 T2C 995

12

24

274 26 715

177 14 427

36 48 60 72 Time since surgery (months) 106 5 267

75 2 170 T2A T2C

56 1 93

27 1 37

84

96

8 1 11

3 1 4

T2B

Estimated Kaplan-Meier BCR in 1,370 patients with RRP by pT2 substage.

97.0) for pT2a, and 94.3% (95% CI 91.8–96.0) and 87.5% (95% CI 82.7–91.1) for pT2c, respectively. On multivariate analysis controlled for pathological processing method significant BCR predictors were margin status (HR 2.7, 95% CI 1.3–5.5, p ⫽ 0.006), preoperative PSA (HR 1.0, 95% CI 1.0 –1.1, p ⫽ 0.029), pathological Gleason score 7 (HR 2.5, 95% CI 1.1–5.7, p ⫽ 0.024) and pathological Gleason score 8 –10 (HR 6.2, 95% CI 2.2–17.4, p ⬍0.001, table 2). With pathological stage T2a as the comparator neither stage pT2b (p ⫽ 1) nor pT2c (p ⫽ 0.42) predicted BCR.

DISCUSSION The studies that helped shape 2002 AJCC staging with respect to the prognostic value of substaging T2 disease focused on clinical stage. For example, Stamey et al reported that according to the 1992 AJCC TNM staging system cT2a tumors were significantly differTable 2. BCR Cox multivariate regression analysis Characteristics Preop PSA Pathological Gleason grade: 6 or Less 7 8–10 Tumor vol Pos surgical margin Pathological T stage: 2a 2b 2c Pathological sampling technique: Whole mount Systematic sampling

OR (95% CI)

p Value

1.03 (1.00–1.06)

0.029

1 (referent) 2.53 (1.13–5.66) 6.18 (2.19–17.4) 1.02 (0.95–1.09) 2.70 (1.33–5.47)

0.024 0.001 0.56 0.006

1 (referent) ⬍0.1 (0–infinity) 1.46 (0.58–3.65)

0.99 0.42

1 (referent) 0.84 (0.40–1.77)

0.65

2294

PATHOLOGICAL STAGE T2 SUBGROUPS AND BIOCHEMICAL RECURRENCE

ent from cT2b and cT2c tumors in pathological and BCR features.13 Han14 and Cagiannos15 et al reported that patients with 1992 cT2a and cT2b tumors had significantly different BCR rates after RRP, whereas those with 1992 cT2b and cT2c tumors had similar rates. They concluded that the 3-tiered 1992 subclassification was preferable to the 2-tiered 1997 T2 subclassification. Furthermore, Han et al recommended combining the T2b and T2c groups into 1 subgroup.14 As a result, in 2002 AJCC revised the TNM staging system and reverted to the previous 1992 classification of T2 disease.16 Surprisingly few studies have focused on whether the same prognostic use holds true for pathological T2 subgroups.3,8,11,19 –22 May et al analyzed 152 cases of pT2 disease and found no significant difference among 1992 T2 substages with respect to BCR at 5-year followup.19 However, sample size was quite small. Freedland et al examined BCR after RRP in 1,606 cases stratified into pT2a and pT2b substages according to 1997 TNM staging criteria.3 They reported that pT2 substages were not significant predictors of BCR on univariate or multivariate analysis. Chun et al evaluated 1,726 European cases and staged them according to 1992/2002 and 1997 TNM staging criteria.11 On univariate and multivariate analysis total PSA, positive surgical margin status, and primary and secondary pathological Gleason scores had prognostic significance but 1992/2002 and 1997 TNM pathological T2 substaging failed to add significant prognostic information. Similarly DeCastro et al studied 906 pT2 cases to assess the prognostic value of substaging.20 On univariate analysis they observed a slightly decreased risk of recurrence in the pT2a subgroup but the difference did not attain statistical significance and on multivariate analysis the association was even weaker. In 372 Asian men Hong et al21 and in 360 men van Oort et al22 evaluated the value of pathological T2 subclassification according to 2002 TNM staging criteria. Their results agreed with the mentioned studies and they found no difference in BCR among pT2 subgroups. We found that patients with pT2c disease had slightly higher risk disease characteristics but as in previous studies we found no significant difference in the likelihood of BCR among 1992/2002 TNM pathological T2 subclasses (p ⫽ 0.174). Our study confirms other commonly identified independent predictors of BCR, including margin status (p ⬍0.001), preoperative PSA (p ⫽ 0.029), and pathological Gleason scores 7 (p ⫽ 0.024) and 8-10 (p ⬍0.001), as prognostic factors for BCR after RP. Unlike the findings of others on the value of clinical T2 substaging pathological substaging in our study seemed to add no additional prognostic information for predicting intermediate term BCR. It remains uncertain whether any pathological subclassification is needed for T2 prostate carcinoma.

Only 2.6% of our pathological T2 cases had the pT2b classification. Furthermore, in RP specimens evaluated by the whole mount technique pT2b stage was identified in only 1 patient. Several groups have cast doubt on the relevance of T2b since it is so rarely observed. For example, attempts were made to identify pT2b disease in RRP specimens in 369 men by Eichelberger and Cheng,25 in 224 by Quintal et al26 and in 360 by van Oort et al.22 Each group used whole mount processing and stratified patients according to 1992/2002 TNM staging criteria. No group identified any pT2b substage in their study. They also concluded that, considering the nature of PCa multifocality, it is unlikely for a tumor that occupies more than half of 1 lobe to remain confined in that lobe without extension to the opposite lobe. Hong et al agreed with these studies and reported that they observed an accompanying smaller satellite tumor focus in the contralateral lobe even with a small index tumor.21 Since not all of our specimens were processed by whole mount sectioning and we did not perform a formal pathology review for multifocality, we cannot conclude that pT2b does not exist. However, although some studies show a higher rate of pT2b disease,11,19,20 our data suggest that pT2b PCa is rare. In our series tumor volume was greater in patients with pT2b disease than in those with pT2a or pT2c. However, pT2b appeared to carry a lower recurrence rate than the other subgroups. There are several possible explanations for this paradox. Large tumors that do not extend across the midline may be biologically less aggressive than others. However, a more likely explanation is that higher tumor volume in pT2b cases is an artifact of the pathological processing method. The estimation technique used with systematic sampling tends to overestimate tumor volume compared to planimetry, for example a median of 3.84 vs 1.71 cc in all 1,370 patients (p ⬍0.01). Since 97% of pT2b cases were evaluated by systematic sampling but the other substages were more evenly divided, tumor volume in pT2b cases appeared larger than in the other groups. We found a different distribution of pathological T2 substages by pathological processing method. Specifically pT2b disease was rarely detected using the whole mount processing method (0.3% vs 3.4%). Again, this may be because pT2b is rare and contralateral tumor foci may more likely be detected in a completely submitted prostate. The remaining pT2 cancers were more likely to be pT2a in the whole mount group than in the systematic sampling group (33.9% vs 21.6%). The explanation for this finding is not as clear since, although the different pathological methods were used sequentially, secular trends such as stage migration during the study period may explain the difference in substage detection. Others have assessed the differences between whole mount and systematic sampling methods but neither method is clearly superior.27,28

PATHOLOGICAL STAGE T2 SUBGROUPS AND BIOCHEMICAL RECURRENCE

For example, Hollenbeck et al reported no significant difference between whole mount and traditional sampling techniques, and concluded that whole mount sampling did not improve the detection of adverse pathological features.27 Thus, the optimal processing method remains to be determined. We included prostates evaluated by the 2 methods. To be a confounder a variable must be associated with the exposure and the outcome. While the processing method is clearly associated with exposure (pathological stage), it is not associated with outcome (BCR) and did not influence the results of our multivariate model. Thus, it did not act as a confounder. Our study has all of the recognized potential limitations and biases of retrospective studies. An

2295

important limitation is the relatively short followup considering the long natural history of PCa. However, our findings add to a group of studies with varying followups that fairly consistently show the absence of a significant difference in BCR by pT2 subgroup.3,20

CONCLUSIONS Results show that the current pT2 PCa substages may not have prognostic significance for intermediate term outcomes. If borne out during longer followup, future staging systems may collapse the substages into 1 or 2 categories rather than 3.

REFERENCES 1. Amling CL, Blute ML, Lerner SE et al: Influence of prostate-specific antigen testing on the spectrum of patients with prostate cancer undergoing radical prostatectomy at a large referral practice. Mayo Clin Proc 1998; 73: 401. 2. Berger AP, Spranger R, Kofler K et al: Early detection of prostate cancer with low PSA cut-off values leads to significant stage migration in radical prostatectomy specimens. Prostate 2003; 57: 93. 3. Freedland SJ, Partin AW, Epstein JI et al: Biochemical failure after radical prostatectomy in men with pathologic organ-confined disease: pT2a versus pT2b. Cancer 2004; 100: 1646. 4. Khan MA, Partin AW, Mangold LA et al: Probability of biochemical recurrence by analysis of pathologic stage, Gleason score, and margin status for localized prostate cancer. Urology 2003; 62: 866. 5. Bianco FJ Jr, Wood DP Jr, Cher ML et al: Tenyear survival after radical prostatectomy: specimen Gleason score is the predictor in organconfined prostate cancer. Clin Prostate Cancer 2003; 1: 242. 6. Chun FK, Graefen M, Zacharias M et al: Anatomic radical retropubic prostatectomy-long-term recurrence-free survival rates for localized prostate cancer. World J Urol 2006; 24: 273. 7. Kupelian P, Katcher J, Levin H et al: Correlation of clinical and pathologic factors with rising prostate-specific antigen profiles after radical prostatectomy alone for clinically localized prostate cancer. Urology 1996; 48: 249. 8. Hernandez DJ, Nielsen ME, Han M et al: Natural history of pathologically organ-confined (pT2), Gleason score 6 or less, prostate cancer after radical prostatectomy. Urology 2008; 72: 172. 9. Wallace DM, Chisolm GD and Hendry WF: T.N.M. classification for urological tumors (U.I.C.C)— 1974. Br J Urol 1975; 47: 1.

10. Schroder FH, Hermanek P, Denis L et al: The TNM classification of prostate cancer. Prostate, suppl., 1992; 4: 129. 11. Chun FK, Briganti A, Lebeau T et al: The 2002 AJCC pT2 substages confer no prognostic information on the rate of biochemical recurrence after radical prostatectomy. Eur Urol 2006; 49: 273. 12. Fleming ID, Cooper JS, Henson DE et al: AJCC Cancer Staging Manual, 5th ed. Philadelphia: Lippincott-Raven 1997; pp 231–232. 13. Stamey TA, Sozen TS, Yemoto CM et al: Classification of localized untreated prostate cancer based on 791 men treated only with radical prostatectomy: common ground for therapeutic trials and TNM subgroups. J Urol 1998; 159: 2009. 14. Han M, Walsh PC, Partin AW et al: Ability of the 1992 and 1997 American Joint Committee on Cancer staging systems for prostate cancer to predict progression-free survival after radical prostatectomy for stage T2 disease. J Urol 2000; 164: 89.

system to predict progression-free survival after radical prostatectomy. BJU Int 2001; 88: 702. 20. DeCastro GJ, McCann T, Benson MC et al: Pathologic T2 stage subgroups and recurrence-free survival after radical prostatectomy. Urology 2008; 72: 1214. 21. Hong SK, Han BK, Chung JS et al: Evaluation of pT2 subdivisions in the TNM staging system for prostate cancer. BJU Int 2008; 102: 1092. 22. van Oort IM, Witjes JA, Kok DE et al: The prognostic role of the pathological T2 subclassification for prostate cancer in the 2002 TumourNodes-Metastasis staging system. BJU Int 2008; 102: 438. 23. Jack GS, Cookson MS, Coffey CS et al: Pathological parameters of radical prostatectomy for clinical stages T1c versus T2 prostate adenocarcinoma: decreased pathological stage and increased detection of transition zone tumors. J Urol 2002; 168: 519. 24. Srigley JR: Key issues in handling and reporting radical prostatectomy specimens. Arch Pathol Lab Med 2006; 130: 303.

15. Cagiannos I, Graefen M, Karakiewicz PI et al: Analysis of clinical stage T2 prostate cancer: do current subclassifications represent an improvement? J Clin Oncol 2002; 20: 2025.

25. Eichelberger LE and Cheng L: Does pT2b prostate carcinoma exist? Critical appraisal of the 2002 TNM classification of prostate carcinoma. Cancer 2004; 100: 2573.

16. Banerjee M, Biswas D, Sakr W et al: Recursive partitioning for prognostic grouping of patients with clinically localized prostate carcinoma. Cancer 2000; 89: 404.

26. Quintal MM, Magna LA, Guimaraes MS et al: Prostate cancer pathologic stage pT2b (2002 TNM staging system): does it exist? Int Braz J Urol 2006; 32: 43.

17. Bostwick DG, Myers RP and Oesterling JE: Staging of prostate cancer. Semin Surg Oncol 1994; 10: 60.

27. Hollenbeck BK, Bassily N, Wei JT et al: Whole mounted radical prostatectomy specimens do not increase detection of adverse pathological features. J Urol 2000; 164: 1583.

18. Hoedemaeker RF, Vis AN and Van Der Kwast TH: Staging prostate cancer. Microsc Res Tech 2000; 51: 423. 19. May F, Hartung R and Breul J: The ability of the American Joint Committee on Cancer Staging

28. Grossfeld GD, Chang JJ, Broering JM et al: Does the completeness of prostate sampling predict outcome for patients undergoing radical prostatectomy? Data from the CAPSURE database. Urology 2000; 56: 430.