Gustavo F. Carvalhal, M.D.1 Peter A. Humphrey, M.D., Ph.D.2 Phataraporn Thorson, M.D.2 Van Van, M.D., Ph.D.1 Christian G. Ramos, M.D.1 William J. Catalona, M.D.1
1 Division of Urologic Surgery. Washington University School of Medicine. St. Louis, Missouri.
2 Division of Surgical Pathology, Washington University School of Medicine, St. Louis, Missouri.
Supported by a grant from the Coordination of Higher Education and Graduate Training of the Brazilian Government (CAPES), Brasilia, Brazil (G.F.C.) and an award from CaP CURE (P.A.H.). W.J.C. was supported by Department of Defense grants, Cap CURE awards, Monsanto, Uromed, the Urologic Research Foundation, and Hybritech Beckman.
Gustavo F. Carvalhal's current address: Rua Luciana de Abreu 210/201, Porto Allegre, RS, Brazil CEP 90570-100.
Christian G. Ramos' current address. Hospital del Trabajador ACHS, Ramon Carnicer 185, Prvidencia, Santiago, Chile.
Peter A. Humphrey's current address: Department of Pathology, Washington University Medical Center, 660 S. Euclid Avenue. Campus Box 8118, St. Louis, MO 63110.
Phataraporn Thorson's current address: Department of Pathology, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Box 085, Houston, TX 77030-4095.
Address for reprints: Peter A. Humphrey, M.D., Ph.D., Department of Pathology, Washington University Medical Center, Campus Box 8118, 660 South Euclid Avenue, St. Louis, MO 63110; Fax: 314747-2040; E-mail: humphrey@path.wustl.edu
Received January 17, 2000; revision received May 22, 2000; accepted May 22, 2000.
© 2000 American Cancer Society
BACKGROUND. The importance of tumor size measurements as predictors of disease recurrence after radical prostatectomy in patients with prostate carcinoma remains somewhat controversial, and many pathologists do not report these measurements routinely. The authors studied the correlation between the visual estimate of the percentage of carcinoma in prostatic tissue from radical prostatectomy specimens and prostate carcinoma recurrence rates in a series of 595 patients who underwent radical prostatectomy.
METHODS. A total of 595 men with clinically localized prostate carcinoma were treated by the same surgeon (W.J.C.) from 1993 through 1997. The percentage of carcinoma in radical prostatectomy specimens from these patients was assessed microscopically through visual estimate. The authors used Kaplan-Meier product limit estimates, log-rank statistics, and the Cox proportional hazards model to evaluate the percentage of carcinoma in the pathologic specimens as predictors of recurrence free survival.
RESULTS. Of the 595 patients, 46 (8%) had evidence of tumor recurrence. The mean percentage of carcinoma in the prostatectomy specimen was 11.3% in the group of patients who did not have disease recurrence and 23.8% in the group of patients who did experience disease recurrence. The percentage of carcinoma, preoperative prostate specific antigen levels, tumor differentiation (histologic Gleason grade), and pathologic stage all were significant predictors of disease recurrence according to the Kaplan-Meier method (all log-rank p values < 0.0001). Using the Cox proportional hazards model, controlling for all of these variables, only pathologic tumor stage, Gleason score, and percentage of carcinoma proved to be independent predictors of disease recurrence. In the final model, which included pathologic stage, Gleason score, and percentage of carcinoma, for each 5% increase in the percentage of carcinoma in the surgical specimen, there was an 11% incremental increase in the chance of tumor recurrence.
CONCLUSIONS. The visual estimate of the percentage of carcinoma in prostatic tissue specimens from patients who undergo radical prostatectomy is a practical, simple, and inexpensive method that provides important prognostic information after radical prostatectomy. Cancer 2000;89:1308-14.
© 2000 American Cancer Society.
KEYWORDS: prostate, carcinoma, prognosis, prostatectomy, prostate specific antigen.
Increasing tumor size has been correlated with many adverse prognostic findings in patients with prostate carcinoma. such as elevated pretreatment serum prostate specific antigen (PSA) levels, a lower percentage of free PSA to total PSA levels, higher histologic tumor grade, DNA aneuploidy, extraprostatic tumor extension, positive surgical margins, advanced pathologic stage, and tumor progression after treatment.1-13 In addition, tumor size contributes directly to the current definitions of clinically important versus possibly unimportant prostate carcinoma.14-16 Therefore, it has been recommended that the amount of carcinoma present in prostatic specimens should be recorded routinely by the pathologist.17,l8 More specifically, for radical pros. tatectomy specimens, it has been recommended the percentage of the prostate that is involved by carcinoma should be noted.17,l8
Although most authors agree that tumor size (percentage of carcinoma or tumor volume) in patients with prostate carcinoma should be reported because of its prognostic importance, in some analyses, tumor size has not been reported to be an independent predictor of tumor recurrence,19,20 and there is still insufficient or inconclusive data linking tumor volume to overall and disease specific survival rates in patients with prostate carcinoma2l Furthermore, histologic grade and tumor volume are not always related, and some tumors may extend outside the prostate before achieving large volumes22,23 Moreover, there are important technical difficulties in performing tumor size measurements in prostate carcinoma specimens. First, there is no agreement about what should be measured: the sum of the volumes of all tumor foci, the volume of the largest lesion, the volume of the most biologically significant lesion, or some combination of these measurements!4 Second, the macro- scopic morphologic features of prostate carcinoma tumors often are subtle, and macroscopic identification of tumors often is impossible, especially with smaller lesions that are detected in the modern era of PSA-based screening. Therefore, unlike the case with many other malignancies, it is often not possible to measure macroscopic tumor size in patients with prostate carcinoma. Currently, the most rigorous mea. surement of tumor size in prostatic carcinoma, that of tumor volume, typically requires complete embedding of the prostatectomy specimen and microscopic examination. This is not practical and usually is done only in research settings. Third, computer-assisted morphometric measurement performed with imaging analyzers, which considered the gold standard in the determination of tumor volume in patients with prostate carcinoma, is time consuming and labor intensive, and image analyzers are not available at all hospitals.
We previously reported that the percentage of carcinoma estimated by using a grid technique was highly correlated with computerassisted morphometric measurements of tumor volume and was correlated with disease progression and patient survival in patients with prostate carcinoma.4,5,8 In this report, we evaluate whether the percentage of carcinoma determined by simple visual inspection is an independent predictor of disease recurrence after radical prostqtectomy in patients with prostate carcinoma.
MATERIALS AND METHODS
Patients
The clinical and pathologic data were analyzed from 595 men with clinically localized prostate carcinoma who underwent radical retropubic prostatectomy between January 1993 and December 1997 at our institution. All operations were performed by the same surgeon (W.J.C.) with a previously reported standardized technique.25 Preoperative evaluation consisted of history and physical examination, including digital rectal examination, the determination of serum acid phosphatase and serum PSA levels, and radioisotope bone scanning with confirmatory imaging studies or bone biopsies when necessary. Many men also underwent abdominal and pelvic computerized tomography. Tumors were considered to be clinical Stage TI tumors if they were impalpable and were detected through either transurethral resection or needle biopsy of the prostate performed because of elevated serum PSA levels. Palpable but clinically localized tumors were considered to be clinical Stage T2 tumors. All patients in this series had clinically localized disease. No patient received neoadjuvant hormonal therapy. Patients were followed with serum PSA measurements at 6-month intervals and digital rectal examinations at yearly intervals. Patients who did not comply with this schedule were contacted by a research assistant. We considered disease recurrence to be the presence of either postoperative detectable serum PSA levels (~ 0.3 ng/mL), local tumor recurrence, or distant metastases.
Pathologic Staging and Estimation of the Percentage of Carcinoma
The pathologic examination of radical prostatectomy specimens and the visual estimate of the percentage of carcinoma were performed by a group of pathologists from the Laboratory of Surgical Pathology at Barnes-Jewish Hospital and Washington University Medical Center. In all cases, the entire pathologic specimens were coated with India ink, and thin, 2-mm margins were obtained from the bladder and the distal urethra. The specimens were fixed in 10% formalin, and the seminal vesicles were amputated and examined separately. The prostate gland was bisected and sectioned at 4-mm intervals from the apex to the base in each case. Each block of the prostatic parenchyma was designated by quadrant (right apex, right base, left apex, and left base). On average, eight to ten blocks of prostatic parenchyma were obtained from each case. Pathologic stage was described originally following a modification of the 1992 American Joint Committee for Cancer (AJCC) classification system;26 however, we translated the pathologic stage (pT) into a modification of the recommendations of the 1997 AJCC Cancer Staging Manual27 for the purposes of the current analysis. Pathologically organ-confined tumors were categorized as pT2 tumors. Tumors with microscopic periprostatic tumor extensions or cancerous surgical margins unilaterally or bilaterally were categorized as pT3a tumors, tumors with seminal vesicle invasion were classified as pT3b tumors, and tumors with lymph node metastases were categorized as N1 tumors. On the basis of Gleason scores, tumors were classified as well-differentiated (Gleason score, 2-4), moderately differentiated (Gleason score, 5-6), moderately to poorly differentiated (Gleason score, 7), and poorly differentiated (Gleason score, 8-10).
To determine the percentage of carcinoma, we used a modification of the technique of visual estimation of the percentage of tumor volume first described by Cantrell et al.l Glass slides with histologic sections from each block were examined under the microscope, and the percentage of carcinoma was estimated by visual inspection without circling of foci or morphometric measurements. The method of Cantrell et al.l reported volume estimates in four broad categories (0-5%, 6-25%, 26-50%, and 51-100%); however, we generally reported our results at 5% or lower increments.
Statistical Analysis
For univariate analyses, we compared Kaplan-Meier product-limit estimates using log-rank statistics to determine whether variables, such as patient age, pre-operative PSA levels, clinical stage, pathologic stage, tumor grade, and percentage of carcinoma, were significant predictors of disease recurrence in patients with prostate carcinoma. Serum prostatic acid phosphatase was not tested as a variable. To establish appropriate cut-off points for the continuous variables (patient age, preoperative PSA levels, and percentage of carcinoma), we produced separate survival curves for each quartile of the study subjects and lumped together groups whose curves yielded similar results. With respect to the categorical variables (clinical stage, pathologic stage, and tumor differentiation), survival curves were produced for each predefined category. We considered log-rank values of p < 0.05 to be statistically significant. For multivariate analysis, we used the Cox proportional hazards method to determine whether the variables that were significant predictors of disease recurrence in univariate analyses remained independent predictors of recurrence after controlling for all other variables entered in the model.
RESULTS
Distribution of Clinical and Pathologic Characteristics
The distributions of the clinical and pathologic characteristics of the 595 patients are summarized in Table 1. The mean patient age was 58 years (range, 38-68 years). Preoperative serum PSA levels were in the range of 4-10 ng/mL in 63% of patients. Fifty-eight percent of the patients had impalpable tumors. At the time of pathologic examination after radical prostatectomy, the vast majority of the tumors were either moderately differentiated (64% with Gleason scores of 5-6) or moderately to poorly differentiated (30% with a Gleason score of 7). Most tumors were pathologically organ confined (66%). Unilateral or bilateral extracapsular extension was present in 26% of patients, seminal vesicle invasion was detected in 7% of patients, and lymph node metastases were present in fewer than 1% of patients. The percentage of carcinoma in the surgical specimen varied widely (mean ± standard deviation [SD], 11% ± 12%; range, 1-80%). The follow-up interval ranged from 1 month to 71 months (mean ± SD, 19 ± 13 months). The actuarial failure rates (Kaplan-Meier method) were 4.7% at 1 year, 9.3% at 2 years, and 15.6% at 3 years.
Univariate Analysis of Predictors of Disease Recurrence
In univariate analyses, preoperative PSA levels, Gleason grade, pathologic stage, and percentage of carcinoma in the surgical specimen all were significant predictors of disease recurrence. Figure 1 shows the probability of disease free survival after radical prostatectomy stratified by preoperative PSA levels. The distribution of PSA levels in quartiles has generated the values 0-4.6, 4.7-6.1, 6.2-9.1, and > 9.1; the first 2 quartiles were lumped for analysis due to the similarities of their survival curves. Lower preoperative PSA levels were associated significantly with increased recurrence free survival rates (log rank, p > 0.0001).
Figure 2 demonstrates the probability of disease free survival after radical prostatectomy stratified by pathologic Gleason scores. Due to the small numbers of patients with well-differentiated (Gleason score, 2-4) tumors (n = 5 patients; 1%), these were analyzed together with the patients with moderately differenti- ated (Gleason score, 5-6) tumors. Lower Gleason scores were associated with a higher probability of recurrence free survival in our series (log rank, p < 0.0001).
Figure 3 shows that the pathologic tumor stage also correlated with recurrence free survival (log rank, p < 0.0001). Organ-confined (pT2) tumors offered the best chance of recurrence free survival, whereas the presence of seminal vesicle invasion or lymph node metastases was associated with the highest recurrence rates.
Although the visual estimates of the percentage of carcinoma in the surgical specimens varied widely, the mean percentage of carcinoma in specimens from the group of patients who did not have a disease recurrence was significantly lower than in the group of patients who had disease recurrence (11.3% vs. 23.8%). The quartile distribution of the visual estimate of the percentage of carcinoma in the pathologic specimens was 0-5%, 5.1-7.5%, 7.6-15%, and > 15%; due to the similarities of their survival curves, the intermediate quartiles (5.1-7.5% and 7.6-15%) were lumped together in the final analysis. Figure 4 shows that the visual estimate of the percentage of carcinoma was correlated inversely with recurrence free survival rates (log rank, p < 0.0001).
Clinical tumor stage had a marginal correlation with recurrence free survival, with impalpable (T1) tumors demonstrating a trend toward lower disease recurrence rates compared with palpable (T2) tumors (log rank, p < 0.005; data not shown) .Patient age at the time of surgery did not correlate with recurrence free survival rates (log rank, p = 0.8; data not shown).
Multivariate Analysis of Predictors of Disease Recurrence
The final results of the Cox regression analysis evaluating the ability of clinical and pathologic variables to predict recurrence free survival are shown in Table 2. We initially entered into the model all variables that were associated significantly with recurrence free survival in the univariate analyses. However, the final statistical model included as independent predictors of recurrence free survival only histologic grade (Gleason score), pathologic stage, and visual estimate of the percentage of carcinoma. Patients with Gleason scores of 7 and 8-10, respectively, had a 2.3 and a 3.1 times higher risk of disease recurrence compared with patients in the reference level of Gleason scores (2-6). Patients with extraprostatic tumor extension or positive margins had a 2.4 times higher risk of tumor recurrence than patients with organ-confined disease. Patients with seminal vesicle invasion by tumor or lymph node metastases had a 4.8 times higher risk of tumor recurrence than patients with organ-confined tumors. The visual estimate of the percentage of carcinoma was predictive of tumor recurrence after controlling for all other predictors. Here, the risk ratio of 1.021 (Table 2) indicates that, for every 1% unit in- crease in tumor size, there is a 2.1% increased risk of failure (1.021- 1.0). Thus, there is an increase of 11% in the risk of tumor recurrence for each 5% increase in the percentage of carcinoma.
DISCUSSION
We report that a simple, light microscopic, visual estimate of the percentage of carcinoma in radical prostatectomy tissue specimens from patients with prostatic carcinoma is an important and independent predictor of disease recurrence after radical prostatectomy. This estimate can be performed in all surgical pathology laboratories without additional time or expense, and it does not require complete embedding of the prostate gland, time-consuming morphometric analysis, or an image analyzer. The ability to predict disease recurrence after radical prostatectomy is important, because this will allow for the identification of patients who may be candidates for adjuvant therapy.
Although increasing tumor size has been associated with adverse pathologic findings in prostate carcinoma, there is controversy surrounding its usefulness as a predictor of disease recurrence in patients with prostate carcinoma. Before the advent of wide-spread PSA screening for prostate carcinoma, Cantrell et al.1 studied the correlation of the percentage of carcinoma (determined both by the pathologist's visual estimation of the percentage of carcinoma after circling foci and by a semiquantitative grid method) and clinical disease recurrence in 117 patients with Stage A prostate carcinoma. All patients underwent either transurethral resection or enucleation of the prostate and were followed for 2-15 years. Those authors found that the estimated percentage of carcinoma was the most important predictor of clinical disease recurrence, even more useful than the Gleason score. In the PSA screening era, Stamey et al.2 studied 68 patients who underwent radical prostatectomy and noted that development of bone metastases and conversion from undetectable to increasing serum PSA levels postoperatively were more common in patients with tumors > 3.0 cc in volume, although there was a strong correlation of tumor volume with tumor grade. In an expanded series of 379 men, this group found that tumor volume in radical prostatectomy tissues from patients with prostate carcinoma was an independent predictor of clinical disease progression.9 Blute and coworkers19 did not find a positive association between tumor volume and disease progression in either univariate analysis or multivariate analysis of a case-control population of 76 age-matched patients with pathologically localized disease, although the method of volume measurement was not clearly specified. Subsequently, Epstein et al20 analyzed the impact of tumor volume and the percentage of carcinoma on the recurrence rates of 185 men who underwent radical prostatectomy. Although both tumor volume and the percentage of carcinoma were increased significantly in the group of patients who had evidence of tumor recurrence, none of these variables was capable of independently predicting tumor recurrence when controlling for Gleason score and surgical margin status. More recently, Renshaw et al28 investigated whether simple tumor measurements, such as the greatest tumor dimension and area, could be of independent prognostic value. Although both measurements were associated with tumor recurrence in univariate analyses, they were only of marginal significance in a multivariate analysis that included pathologic stage and preoperative PSA levels. Moreover, prostatic carcinoma tumors, particularly those detected through PSA screening, often are multifocal, without a dominant nodule, and this creates difficulty in using the greatest tumor dimension as a measure of prostate carcinoma tumor size. Eastham and Scardino29 reported that, in Baylor's series, tumor volume was a significant predictor of disease recurrence both in univariate analysis and multivariate analysis but did not suggest the routine measurement of tumor volume in view of the small additional contribution of tumor volume measurement to their final predictive model. We previously reported that the circling and grid techniques of visually estimating the percentage of carcinoma correlate well with morphometric calculations of tumor volume performed through image analysis.30 Furthermore, we have shown the independent prognostic value of the grid technique in determination of the percentage of carcinoma in a smaller series of 73 patients with clinically palpable tumors.5
The current results confirm the usefulness of the visual estimate of the percentage of carcinoma as an independent predictor of disease recurrence after radical prostatectomy in a larger, more recent series of patients with tumors that were diagnosed mainly through PSA screening. One of the most appealing advantages of the method of visually estimating the percentage of carcinoma is that it is obtained from routine sections, obviating the need for complete embedding of all of the prostatic tissue. Indeed, complete embedding of prostatic tissue is not performed routinely by most pathology laboratories31 and is not likely to become standard practice, especially in this era of cost constraints. Moreover, the technique of a simple visual estimate of the percentage of carcinoma does not require expensive and sophisticated computerized morphometric analyzers.
Several potential limitations of this investigation should be noted. First, we did not compare other measures of tumor size, such as the morphometrically determined percentage of carcinoma or tumor volume, for their ability to predict disease recurrence. In addition, the percentage of carcinoma was deter- mined by different pathologists in our series, and we did not assess for possible interexaminer variability in this report. Furthermore, the follow-up was relatively short. Because the median follow-up was only 19 months, additional follow-up will be necessary to confirm these findings. Finally, because our population contained many patients with prostate carcinoma that was detected during screening, our results may not be generalizable to all clinical populations.
In summary, the visual estimate of the percentage of carcinoma in prostatic carcinoma tissue specimens from patients who have undergone radical prostatec- tomy is a practical, simple, and inexpensive method that generates independent prognostic information after radical prostatectomy. We recommend routine reporting of the visual estimate of the percentage of carcinoma in clinical practice.
REFERENCES
1. Cantrell BB, DeKlerk DP, Eggleston JC, Boitnott JK, Walsh PC. Pathological factors that influence prognosis in Stage A prostatic cancer: the influence of extent versus grade. J Urol 1981;125:516-20.
2. Stamey TA, McNeal JE, Freiha FS, Redwine E. Morphometric and clinical studies on 68 consecutive radical prostatecto- mies. J Urol 1988;139:1235-41.
3. Partin AW, Epstein JE, Cho KR, Gittelsohn AM, Walsh PC. Morphometric measurement of tumor volume and percent of gland involvement as predictors of pathological stage in clinical Stage B prostate cancer. J Urol 1989;141:341-5.
4. Humphrey PA, Vollmer RT. Intraglandular tumor extent and prognosis in prostatic carcinoma: application of a grid method to prostatectomy specimens. Hum Pathol 1990;21: 799-804.
5. Humphrey PA, Walther PJ, Currin SM, Vollmer RT. Histologic grade, DNA ploidy, and intraglandular tumor extent as indicators of tumor progression of clinical Stage B prostatic carcinoma. Am J Surg Pathol 1991;15:1165-70.
6. McNeal JE. Cancer volume and site of origin of adenocarcinoma in prostate: relationship to local and distant spread. Hum Pathol 1992;23:258-66.
7. Bostwick DG, Graham SD Jr., Napalkov P, Abrahamsson PA, Sant'agnese PA, Algaba F, et al. Staging of early prostate cancer: a proposed tumor volume-based progno~tic index. Urology 1993;41:403-11.
8. Humphrey PA, Frazier HA, Vollmer RT, Paulson DF. Stratification of pathologic features in radical prostatectomy specimens that are predictive of elevated initial postoperative serum prostate-specific antigen levels. Cancer 1993;71: 1821-7.
9. Stamey TA, McNeal JE, Yemoto CM, Sigal CM, Johnston IM. Biological determinants of cancer progression in men with prostate cancer. JAMA 1999;281:1395-400.
10. Elackwell KL, Eostwick DG, Myers RP, Zinke H, Oesterling JE. Combining prostate specific antigen with cancer and gland volume to predict more reliably pathological stage: the influence of prostate specific antigen cancer density. J Urol 1994;151:1565-70.
11. Eabaian RJ, Troncoso P, Steelhammer LC, Lloreta-Trull J, Ramirez El. Tumor volume and prostate specific antigen: implications for early detection and defining a window of curability. J UrolI995;154:1808-12.
12. Arcangeli CG, Humphrey PA, Smith DS, Harmon TJ, shepherd DL, Keetch DW, et al. Percentage of free serum prostate specific antigen as a predictor of pathologic features of prostate cancer in a screening population. Urology 1998;51: 558-65.
13. PettawayCA, Troncoso P, Ramirez El, Johnston DA, Steelhammer L, Eabaian RJ. Prostate specific antigen and pathological features of prostate cancer in black and white patients: a comparative study based on radical prostatectomy specimens. J Urol 1998;160:437-42.
14. Stamey TA, Freiha FS, McNeal JE, Redwine EA, Whittemore AS, Schmid HP. Localized prostate cancer: relationship of tumor volume to clinical significance for treatment of pros- tate cancer. Cancer 1993;71:993-8.
15. Ohori M, Wheeler TM, Dunn JK, Stamey TA, Scardino PT. The pathological features and prognosis of prostate cancer detectable with current diagnostic tests. J Urol 1994;152: 1714-20.
16. Epstein II, Walsh PC, Erendler CE. Radical prostatectomy for impalpable prostate cancer: the Johns Hopkins experience with tumors found on transurethral resection (Stages Tla and Tlb) and on needle biopsy (Stage TIc). J Urol 1994;152: 1721-9.
17. Henson DE, Hutter RVP, Farrow G, for members of the Cancer Committee, College of American Pathologists, and the Task Force on the Examination of Specimens Removed from Patients with Prostate Cancer. Practice protocol for the examination of specimens removed from patients with carcinoma of the prostate gland. Cancer Committee, College of American Pathologists. Arch Pathol Lab Med 1994;118 (I): 779-83.
18. Amin ME, Grignon D, Eostwick D, Reuter V, Troncoso P, Ayala AG for the Association of Directors of Anatomic and Surgical Pathology. Recommendations for the reporting of resected prostate carcinomas. Am J Clin Pathol 1996;105: 667-70.
19. Elute ML, Nativ 0, Zincke H, Farrow GM, Therneau T, Lieber MM. Pattern of failure after radical retropubic prostatectomy for clinically and pathologically localized adenocarcinoma of the prostate: influence of tumor deoxyribonucleic acid ploidy. J Urol 1989;142:1262-5.
20. Epstein JI, Carmichael M, Partin AW, Walsh PC. Is tumor volume an independent predictor of progression following radical prostatectomy? A multivariate analysis of 185 clinical Stage B adenocarcinomas of the prostate with 5 years of follow up. J Urol 1993;149:1478-81.
21. Hutter RVP, Montie JE, Busch C, Grignon DJ, Lieber M, Logothotis C. Current prognostic factors and their relevance to staging. Workgroup 3. International consultation on pros- tatic intraepithelial neoplasia and pathologic staging for prostatic carcinoma. Cancer 1996;78:369-71.
22. Lange PH. Prognostic factors and clinical significance of cancers detected by screening: growth rate and progression of prostate cancer [editorial]. J Urol 1994;152:1707-8.
23. Miller GJ, Cygan JM. Morphology of prostate cancer: the effects of multifocality on histological grade, tumor volume and capsule penetration. J Urol 1994;152:1709-13.
24. Grignon DJ, Sakr WA. Pathologic staging of prostate carcinoma: what are the issues? International consultation on prostatic intraepithelial neoplasia and pathologic staging of prostatic carcinoma. Cancer 1996;78:337-40.
25. McCarthy JF, Catalona WJ. Nerve-sparing radical retropubic prostatectomy. In: Marshall FF, editor. Textbook of operative urology. Philadelphia: W.B. Saunders, 1996:537-44.
26. Beahrs OH, Henson DE, Hutter RVP. American Joint Com- mittee on Cancer manual for staging cancer, 4th edition. Philadelphia: J.B. Lippincott, 1992.
27. Fleming ID, Cooper JS, Henson DE, et al., eds. American Joint Committee on Cancer. Cancer Staging manual, 5th edition. Philadelphia: J.B. Uppincott, 1997.
28. Renshaw M, Richie JP, Loughlin KR, Jiroutek M, Chung A, D'Amico AV. The greatest dimension of prostate carcinoma is a simple, inexpensive predictor of prostate specific antigen failure in radical prostatectomy specimens. Cancer 1998;83:748-52.
29. Eastham JA, Scardino PT. Radical prostatectomy. In: Walsh PC, Retik AB, Vaughan ED Jr., Wein AJ, editors. Campbell's Urology. 7th edition. Philadelphia: W.B. Saunders, 1998: 2547-64.
30. Humphrey PA, Vollmer RT. Percentage carcinoma as a measure of prostatic tumor size in radical prostatectomy tissues. Mod Pathol 1997;10:326-33.
31. True LD. Surgical pathology examination of the prostate gland. Practice survey by the American Society of Clinical Pathologists. Am J Clin Pathol 1994;102:572-9.