PROGNOSTIC FACTORS FOR MULTIFOCAL PROSTATE CANCER INRADICAL PROSTATECTOMY SPECIMENS: LACK OF SIGNIFICANCE OF

SECONDARY CANCERS

MASANORI NOGUCHI, THOMAS A. STAMEY, JOHN E. MCNEAL* AND ROSALIE NOLLEYFrom the Department of Urology, Kurume University School of Medicine (MN), Kurume, Japan, and Department of Urology, Stanford

University School of Medicine (TAS, JEM and RN), Stanford, California

ABSTRACT

Purpose: We evaluated secondary cancers in the prostate in relation to predictions of patho-logical stage and prognosis.

Materials and Methods: A total of 222 men with T1c (impalpable) prostate cancer and 6 or moresystematic needle biopsies were matched with radical prostatectomy and classified into 3 groupsaccording to tumor multifocality and secondary cancer volumes, including a single tumor in 54(24%), an index (largest) tumor with secondary cancers less than 0.5 cc in 86 (39%) and an indextumor with secondary cancers greater than 0.5 cc in 82 (37%). Logistic analysis was used topredict adverse histological features. Cox proportional hazards analysis was used to predictprostate specific antigen (PSA) failure after radical prostatectomy.

Results: There were no differences among the 3 groups with respect to preoperative serumPSA, number of positive cores, percent Gleason grade 4/5 cancer in the needle biopsy or histo-logical features in radical prostatectomy specimens. On logistic analysis neither serum PSA norpre-radical biopsy predicted adverse histological features in radical prostatectomy specimens.The Cox regression model showed that primary predictors of PSA failure were percent Gleasongrade 4/5 cancer in the biopsy (HR � 2.6, p � 0.015) and prostatectomy (HR � 2.4, p � 0.04)specimens, and the number of positive cores (HR � 2.5, p � 0.04). When comparing PSA failurerates among the 3 groups, the multifocal group with smaller secondary cancers showed a betterprognosis than the single tumor group (p � 0.019).

Conclusions: Secondary cancers in multifocal prostate tumors did not adversely influence theresults of preoperative clinical parameters, including PSA and needle biopsy findings. PercentGleason grade 4/5 cancer in needle biopsies and prostatectomy specimens is the most powerfulpredictor of biochemical failure in men with stage T1c prostate cancer after radical prostatec-tomy.

KEY WORDS: prostate, prostatic neoplasms, biopsy, prostate-specific antigen, prostatectomy

Early detection strategies using serum prostate specificantigen (PSA) and ultrasound guided systematic sextant bi-opsy have resulted in massive stage migration with greaterthan 50% of men showing nonpalpable cancer (T1c) at initialpresentation.1 The majority of clinical stage T1c cases arethought to be biologically significant based on tumor volume,grade and pathological stage but there is no agreement onthe limits of clinically insignificant cancers.2, 3 Moreover, themajority of radical prostatectomy specimens contain 2 ormore independent tumors.4, 5 It is unclear whether thesesecondary cancers contribute to serum PSA failure after rad-ical prostatectomy.

We examined the relationship between preoperative nee-dle biopsy findings and histopathological features in radicalprostatectomy specimens by comparing multifocal tumorswith unifocal tumors and determined whether the number ofpositive cores, core cancer length and Gleason grade on bi-opsy are related to the prognostic factors of organ confinedstatus, Gleason grade, capsular penetration and positive sur-gical margins in prostatectomy specimens. Importantly weexplored the impact of secondary cancers as multifocal tu-mors on disease progression.

MATERIALS AND METHODS

Patient population. A total of 222 prostate biopsies positivefor cancer and matched radical prostatectomy specimenswere selected from the urological pathology files of 450 con-secutive patients who underwent radical retropubic prosta-tectomy for stage T1c disease at Stanford University MedicalCenter between January 1988 and January 1997. Of the men244 were excluded from analysis, including 130 with lessthan 6 needle cores, 85 treated with hormones, irradiation ortransurethral resection before surgery and an additional 13with incomplete data. Mean and median patient age was 63years (range 43 to 79). The prostate gland was consideredbenign (no palpable evidence of asymmetry, nodularity, ir-regularity or induration) based on digital rectal examination.All PSA values were determined using the automatedmonoclonal-monoclonal AIA-600 assay (Tosoh, Foster City,California).

Systematic sextant biopsy specimens. Systematic sextantbiopsies were obtained from 6 or more separate regions withat least 3 per side. The length of cancer in each biopsy corewas measured in mm and the percent of cancer composed ofGleason grade 4 or 5 cancer was estimated. We calculated thetotal length of cancer in all cores, the number of individualcores with cancer, and the percent of Gleason grade 4 and 5cancer.

Radical prostatectomy specimens. Prostatectomy speci-

Accepted for publication March 21, 2003.* Corresponding author and requests for reprints: Department of

Urology, S-287, Stanford University Medical Center, 300 PasteurDr., Stanford, California 94305-5118.

0022-5347/03/1702-0459/0 Vol. 170, 459–463, August 2003THE JOURNAL OF UROLOGY® Printed in U.S.A.Copyright © 2003 by AMERICAN UROLOGICAL ASSOCIATION DOI: 10.1097/01.ju.0000070928.49986.04

459

mens were evaluated according to the Stanford protocol us-ing serial transverse sections at 3 mm intervals.4, 5 In eachprostate the volume of the largest cancer as well as secondarycancers was determined using a computer assisted imageanalysis system.6 Tumor grade was determined according tothe Gleason system of 5 grades and the percent of each canceroccupied by Gleason grades 4 and 5 was estimated. Theextent of capsule penetration and positive margins weremarked on the slide and quantified in linear cm. Seminalvesicle invasion and positive regional lymph nodes were re-corded. All biopsies and radical prostatectomy specimenswere independently evaluated in the department of urologyby one of us (JEM). Tumor volume less than 0.5 cc is gener-ally accepted as a criterion for defining a clinically insignif-icant tumor in radical prostatectomy specimens. All caseswere classified into 3 subgroups in accordance with tumortype and secondary cancer volume, including unifocal groupwith a solitary tumor in the radical prostatectomy specimen,multifocal group 1 with the largest tumor less than 0.5 cc ofthe secondary tumor and multifocal group 2 with the largesttumor greater than 0.5 cc of the secondary tumor.

Followup at 6 to 12-month intervals included a brief his-tory, rectal examination and serum PSA. Biochemical failurewas defined as a PSA of at least 0.07 ng/ml on the Tosohassay, increasing on subsequent serum samples. Of 182 menwho had no biochemical evidence of disease 65 did not have arecent serum PSA measurement by the Tosoh assay. Thus,the available Hybritech assay (Hybritech Beckman-CoulterCorp., San Diego, California) value of less than 0.2 ng/ml wasaccepted as evidence of undetectable PSA. Time to PSA fail-ure was the time of the first detectable value and time zerowas defined as the day of surgery.

Statistics. Data were analyzed using commercially avail-able computer software. Logistic regression analysis wasused to evaluate whether any preoperative parameters wereable to predict histological features in radical prostatectomyspecimens. The Cox proportional hazard regression modelwas used to calculate the hazard rate of PSA failure afterradical prostatectomy with the 95% CI. Independent predic-

tors of biochemical failure were selected in stepwise multi-variate Cox proportional hazard regression models. Thedisease-free survival rate was calculated by the Kaplan-Meier method. The log rank test was used to analyze tumortypes related to recurrence. Student’s t and the chi-squaretests were used to compare quantitative and categorical vari-ables, respectively. Test results were considered significantat p �0.05.

RESULTS

Preoperative characteristics. Table 1 lists preoperativecharacteristics of the 222 patients undergoing radical retro-pubic prostatectomy for clinical stage T1c prostate cancer. Ofthe 222 patients 54 (24%) had only 1 prostate cancer (unifo-cal) and 168 (76%) had multifocal prostate cancer. Of thelatter patients 86 had a secondary cancer volume of less than0.5 cc (multifocal group 1) and 82 had a secondary cancervolume of greater than 0.5 cc (multifocal group 2). The meannumber of secondary cancers in the multifocal tumor groupswas 3 (range 1 to 7). Table 1 shows preoperative PSA andneedle biopsy findings in the unifocal, multifocal 1 and mul-tifocal 2 groups. There were no differences among the 3tumor types with respect to PSA, the number of positive coresor percent Gleason grade 4 and 5 cancer in needle biopsies.The mean percent Gleason grade 4/5 cancer was evenly dis-tributed across all 3 tumor types (tables 1 and 2). Impor-tantly total cancer length in needle biopsies in multifocalgroups 1 and 2 was significantly longer than cancer length inthe unifocal group (p �0.05), suggesting strongly that sec-ondary cancers definitely contribute to positive biopsy rates.

Histopathological features according to tumor type. Table 2lists histological features of the radical prostatectomy speci-mens. All except 47 lesions were organ confined. Capsularpenetration occurred in 20% of cases, positive surgical mar-gins were present in 9% and seminal vesicle invasion wasevident in 1%, of which all were associated with the largestcancer. In 6 cases metastatic cancer was identified in thepelvic lymph nodes. Of 222 cases 146 (66%) had a Gleason of

TABLE 1. Preoperative clinical characteristics of 222 men with nonpalpable prostate cancer by tumor type

TotalsTumor Type

Unifocal* Multifocal 1* Multifocal 2*

No. pts 222 54 86 82Mean age (range) 62.5 61.5 (48–73) 62.8 (43–79) 62.9 (47–74)No. ng/ml PSA (%):

4 or Less 4 (2) 0 3 (3) 1 (1)4.1–10 160 (72) 44 (81) 65 (76) 51 (62)10.1–20 55 (25) 8 (15) 17 (20) 30 (37)Greater than 20 (%) 3 (1) 2 (4) 1 (1) 0Mean (range) 8.8 (4–50.4) 8.9 (4.1–50.4) 8.3 (4.0–32.7) 9.2 (4.1–19.1)

Total biopsy cores:Mean (range) 6.3 (6–13) 6.7 (6–13) 6.3 (6–10) 6.3 (6–10)Median 6 6 6 6

No. pos biopsy cores (%):1 78 (35) 23 (43) 27 (31) 28 (34)2 79 (36) 22 (41) 32 (38) 25 (31)3 42 (19) 4 (7) 19 (22) 19 (23)Greater than 3 23 (10) 5 (9) 8 (9) 10 (12)Mean (range) 2.1 (1–7) 1.9 (1–5) 2.1 (1–5) 2.2 (1–7)

No. total Ca mm length on biopsy (%):3 or Less 80 (36) 25 (46) 28 (33) 27 (33)3.1–6.0 66 (30) 14 (26) 27 (31) 25 (31)6.1–9.0 29 (13) 10 (19) 8 (9) 11 (13)Greater than 9 47 (21) 5 (9) 23 (27) 19 (23)Mean (range) 6.5 (0.5–34) 4.8 (0.5–17) 7.2 (0.5–34)* 7.0 (0.5–34)*

No. Gleason grade 4/5 (%):0 132 (59) 31 (57) 47 (54) 54 (66)5–20 35 (16) 7 (13) 17 (20) 11 (13)25–40 17 (8) 6 (11) 6 (7) 5 (6)Greater than 40 38 (17) 10 (19) 16 (19) 12 (15)Mean (range) 17 (0–100) 19 (0–100) 18 (0–100) 14 (0–100)

Unifocal—solitary tumor in radical prostatectomy specimen, multifocal 1—largest tumor less than 0.5 cc of secondary tumor and multifocal 2—largest tumorgreater than 0.5 cc of secondary tumor.

* Significantly longer than in unifocal group (p � 0.05).

LACK OF SIGNIFICANCE OF SECONDARY CANCERS460

score 7 (3 � 4 in 49% and 4 � 3 in 17%), 69 (31%) had aGleason score of 6 and only 7 (3%) had other scores. MeanGleason grades 4 and 5 cancer in the largest tumor waspresent in 22% of cases (range 0% to 95%). There was nocorrelation between tumor type (unifocal, multifocal 1 ormultifocal 2) and capsular penetration, seminal vesicle inva-sion, positive surgical margins or lymph node metastasis(table 2). The largest cancer volume in the multifocal 2 groupwas significantly larger than in the unifocal (p �0.05) andmultifocal 1 (p �0.001) groups (table 2). Pearson’s correlationdemonstrated that total cancer volume in the prostatectomyspecimen was significantly related to the largest cancer vol-ume (fig. 1). Relationships among the risks of nonorgan con-fined disease, capsular penetration, positive surgical mar-gins and preoperative variables were examined on univariateanalysis using a logistic model. Analysis results showed nodifference in the risk of adverse histological features due topreoperative PSA or biopsy findings (data not shown).

Progression-free survival. Overall mean followup of the 222patients was 52.4 months (range 6 to 152). Only 39 men(18%) showed increasing PSA. To evaluate the impact ofsecondary cancer on PSA failure-free survival the disease-free survival rate was estimated and stratified by tumor type(fig. 2). PSA failure-free survival rates were consistentlyhigher in the multifocal 1 group than in the unifocal group(p � 0.019). These low rates of progression-free survival inthe unifocal group suggested that PSA failure might be re-

lated to the largest cancer and smaller secondary cancers(less than 0.5 cc) may not have affected the prognosis.

Prognostic factors for progression. Cox proportional hazardregression analysis was performed to determine factors pre-dictive of biochemical failure (table 3). On univariate analy-sis preoperative PSA, percent Gleason grade 4 and 5 cancerin the biopsy and prostatectomy specimens, seminal vesicleinvasion and positive lymph nodes were significant predic-tors of biochemical failure. However, the number of men withseminal vesicle invasion or lymph node metastasis was toosmall to be statistically reliable. Secondary cancer volumewas not significant. Multivariate analysis identified the per-cent Gleason grades 4 and 5 in the biopsy as the strongestindependent predictor of biochemical failure after radicalprostatectomy (HR � 2.6, 95% CI 1.2 to 5.7, p � 0.015),followed by percent Gleason grades 4 and 5 in the prostatec-tomy specimen (HR � 2.4. 95% CI 1.0 to 5.4, p � 0.04) andnumber of positive cores on biopsy (HR � 2.5, 95% CI 1.0 to6.2, p � 0.04, table 3).

DISCUSSION

Prostate cancer has been recognized as a multifocal dis-ease.4, 5 In prostatectomy specimens multifocal foci can beseparated from the index (largest) cancer and it has beenshown that the largest tumor is typically larger than allsmaller lesions combined as well as the most poorly differen-

TABLE 2. Pathological features of radical prostatectomy specimens by tumor type

TotalsTumor Type

Unifocal* Multifocal 1* Multifocal 2*

No. pts 222 54 86 82Mean gm prostate wt (range) 59.6 (24–183) 64.3 (27–176) 60.7 (24–183) 55.3 (29–103)Mean cc Ca vol (range):

Largest 1.9 (0.01–4.0) 1.8 (0.01–4.0) 1.6 (0.01–4.0) 2.2 (0.7–3.7)*Secondary 0.5 (0–3.4) 0 0.2 (0.01–0.49) 1.3 (0.5–3.4)†

Mean % Gleason grade 4/5 (range) 22 (0–95) 21 (0–95) 20 (0–90) 24 (0–95)No. capsular penetration (%) 44 (20) 9 (17) 16 (19) 19 (23)No. seminal vesicle invasion (%) 3 (1) 1 (2) 0 2 (2)No. surgical margins (%) 20 (9) 4 (7) 7 (8) 9 (11)No. lymph node (%) 6 (3) 1 (2) 0 5 (6)

Unifocal—solitary tumor in radical prostatectomy specimen, multifocal 1—largest tumor less than 0.5 cc of secondary tumor and multifocal—largest tumorgreater than 0.5 cc of secondary tumor.

* Significantly higher than unifocal (p �0.05) and multifocal 1 (p �0.001).† Significantly higher than multifocal 1 (p �0.0001).

FIG. 1. Relationship between largest and total cancer volume in222 prostatectomy specimens. About 75% of observations on x axiswere related to observations on y axis.

FIG. 2. Postoperative estimates of PSA failure-free survival inpatients stratified by tumor type, including unifocal with solitarycancer in radical prostatectomy specimen, multifocal 1 with largesttumor with less than 0.5 cc of secondary tumor and multifocal 2 withlargest tumor with greater than 0.5 cc of secondary tumor. PSAfailure-free survival rates were consistently higher in multifocalgroups 1 and 2 than in unifocal group.

LACK OF SIGNIFICANCE OF SECONDARY CANCERS 461

tiated.4 The percent Gleason grade 4/5 and cancer volume inthe largest tumor, positive lymph node findings and intra-prostatic vascular invasion have been independently associ-ated with prostate cancer progression.7 Therefore, to predictthese prognostic factors a number of investigators have an-alyzed the relationship between preoperative biopsy findingsand histopathological findings in prostatectomy specimens.However, these relationships were weak and disappoint-ing.8–11 Multiple independent foci of prostate adenocarci-noma, grade heterogeneity within a single tumor, a differ-ence in the anatomical location of the tumor or a small biopsycore might cause these weak relationships.

In the current study 76% of all patients had multifocaltumors and there were no differences between unifocal andmultifocal tumors with respect to serum PSA and biopsyfindings except total cancer length. However, total cancerlength in needle biopsies in multifocal groups 1 and 2 weresignificantly longer than in the unifocal group (p �0.05). It isprobably a major reason why biopsies do not reflect the indexcancer well. Djavan et al reported that preoperative PSAdensity of the transition zone and free-to-total PSA weresignificantly different in patients with unifocal vs multifocaldisease but neither PSA nor PSA density was able to predictunifocality or multifocality.12 Their results suggest that sec-ondary cancers in multifocal tumors do not significantly in-fluence the results of preoperative PSA and needle biopsyfindings. Moreover, we have been unable to determine anyrelationship between the largest peripheral zone cancer vol-ume or grade and serum PSA.13

Several recent studies indicate that neither the number ofpositive cores nor percent prostate biopsy tissue with canceris predictive of adverse histological findings.14–16 Other stud-ies have focused on the development of an algorithm usingbiopsy findings, such as greater than 3 sextant biopsies con-taining any Gleason grade 4 or 5 cancer,17 to define a subsetof patients who are at an extremely high risk for pelvic lymphnode metastasis. However, these investigators did not con-sider whether multifocality in prostate cancer affected biopsyfindings. In the current study logistic regression analysisshowed no difference in the risk of nonorgan confined dis-ease, capsular penetration or positive surgical margins intheir relationship to preoperative PSA or biopsy findingsbetween unifocal and multifocal tumors.

Our Cox regression model shows that the primary predic-tors of PSA failure are the percent Gleason grade 4/5 inbiopsy and prostatectomy specimens, and the number of pos-itive cores. Secondary cancer volume did not predict biochem-ical PSA failure. Recently we reported a morphological anal-

ysis of secondary cancers in 486 radical prostatectomies,which included the patients in the current study.18 In thatseries we noted a steady decrease in largest cancer size witheach increase in the number of smaller cancers, an observa-tion that indicates the biological importance of index cancers.In fact, the PSA failure rate in patients with multifocaltumors in the current study was consistently better for mul-tifocal than for unifocal tumors, confirming earlier data18

that secondary cancers have a more favorable prognosis thana single solitary cancer (fig. 2). These findings support theconcept that smaller incidental tumors found in the prostateprobably have little prognostic significance for the detectionof the largest cancer.4

CONCLUSIONS

The current findings suggest that secondary cancers in theprostate do not contribute to preoperative PSA. Secondarycancers in the prostate are unlikely to be clinically signifi-cant. Significance rests with the index or largest canceralone. In the 222 men with stage T1c prostate cancer thepercent Gleason grade 4/5 cancer in the biopsy and radicalprostatectomy specimens, and the number of positive biopsycores significantly predicted biochemical failure.

REFERENCES

1. Sun, L., Gancarezyk, K., Paquette, E. L., McLeod, D. G., Kane,C., Kusuda, L. et al: Introduction to Department of DefenseCenter for Prostate Disease Research Multicenter NationalProstate Cancer Database, and analysis of changes in thePSA-era. Urol Oncol, 6: 203, 2001

2. Oesterling, J. E., Suman, V. J., Zinke, H. and Bostwick, D. G.:PSA-detected (clinical stage T1c or B0) prostate cancer. Patho-logically significant tumors. Urol Clin North Am, 20: 687, 1993

3. Epstein, J. I., Walsh, P. C., Carmichael, M. and Brendler, C. B.:Pathologic and clinical findings to predict tumor extent ofnonpalpable (stage T1c) prostate cancer. JAMA, 271: 368,1994

4. McNeal, J. E., Price, H. M., Redwine, E. A., Freiha, F. S. andStamey, T. A.: Stage A versus stage B adenocarcinoma of theprostate: morphological comparison and biological signifi-cance. J Urol, 139: 61, 1988

5. Villers, A., McNeal, J. E., Freiha, F. S. and Stamey, T. A.:Multiple cancers in the prostate. Morphologic features of clin-ically recognized versus incidental tumors. Cancer, 70: 2313,1992

6. Noguchi, M., Stamey, T. A., McNeal, J. E. and Yemoto, C. M.:Assessment of morphometric measurements of prostate cancervolume. Cancer, 89: 1056, 2000

7. Stamey, T. A., McNeal, J. E., Yemoto, C. M., Sigal, B. M. and

TABLE 3. Cox proportional hazards analysis of factors predicting biochemical recurrence after radical prostatectomy

Potential Predictors Cutoffs HR (95% CI) p Value

Univariate analysis:Age* 63 or Greater/less than 63 0.782 (0.412–1.486) 0.453Serum PSA (ng/ml)* 9 or Greater/less than 9 2.194 (1.164–2.194) 0.015No. pos. biopsy cores* 2 or Greater/less than 2 1.931 (0.885–4.214) 0.098Total biopsy Ca length (mm)* 6 or Greater/less than 6 1.123 (0.597–2.112) 0.719% Gleason grade 4/5 in biopsy*,† 17 or Greater/less than 17 3.645 (1.907–6.969) �0.0001Largest Ca vol (cc)* 2 or Greater/less than 2 1.749 (0.921–3.322) 0.087Secondary Ca vol (cc)* 0.5 or Greater/less than 0.5 1.221 (0.644–2.315) 0.539% Prostatectomy specimen Gleason grade 4/5† 22 or Greater/less than 22 4.793 (2.454–9.362) �0.0001Capsular penetration Pos/neg 1.592 (0.816–3.107) 0.172Seminal vesicle invasion Pos/neg 5.914 (1.400–24.987) 0.016Pos surgical margins Pos/neg 0.935 (0.363–2.410) 0.889Pos lymph node Pos/neg 3.808 (1.478–9.809) 0.006

Multivariate analysis:% Biopsy Gleason grade 4/5 17 or Greater/less than 17 2.621 (1.201–5.720) 0.015% Prostatectomy specimen Gleason grade 4/5 22 or Greater/less than 22 2.377 (1.038–5.4440) 0.040No. pos biopsy cores 2 or Greater/less than 2 2.537 (1.038–6.199) 0.041

Of the 222 men 39 (18%) had biochemical failure.* Patient age, PSA, number of positive cores, cancer length, cancer volume and percent Gleason grade 4/5 were based on mean values with remaining predictors

treated as dichotomous variables.† Percent Gleason grade 4/5 cancer in preoperative biopsy and prostatectomy specimens were most significant variables for predicting serum PSA recurrence

after prostatectomy with (hazard ratio 3.6 and 4.8, respectively, p �0.0001).

LACK OF SIGNIFICANCE OF SECONDARY CANCERS462

Johnstone, I. M.: Biological determinants of cancer progres-sion in men with prostate cancer. JAMA, 281: 1395, 1999

8. Cupp, M. R., Bostwick, D. G., Myers, R. P. and Oesterling, J. E.:The volume of prostate cancer in the biopsy specimen cannotreliably predict the quantity of cancer in the radical prosta-tectomy specimen on an individual basis. J Urol, 153: 1543,1995

9. Humphrey, P. A., Baty, J. and Keetch, D.: Relationship betweenserum prostate specific antigen, needle biopsy findings, andhistopathologic features of prostate carcinoma in radical pros-tatectomy tissues. Cancer, 75: 1842, 1995

10. Dietrick, D. D., McNeal, J. E. and Stamey, T. A.: Core cancerlength in ultrasound-guided systematic sextant biopsies: apreoperative evaluation of prostate cancer volume. Urology,45: 987, 1995

11. Noguchi, M., Stamey, T. A., McNeal, J. E. and Yemoto, C. M.:Relationship between systematic biopsies and histological fea-tures of 222 radical prostatectomy specimens: lack of predic-tion of tumor significance for men with nonpalpable prostatecancer. J Urol, 166: 104, 2001

12. Djavan, B., Susani, M., Bursa, B., Basharkhah, A., Simak, R.and Marberger, M.: Predictability and significance of multifo-cal prostate cancer in the radical prostatectomy specimen.Tech Urol, 5: 139, 1999

13. Stamey, T. A., Johnstone, I. M., McNeal, J. E., Lu, A. Y. andYemoto, C. M.: Preoperative serum prostate specific antigenlevels between 2 and 22 ng./ml correlate poorly with post-

radical prostatectomy cancer morphology: prostate specific an-tigen cure rates appear constant between 2–9 ng./ml. J Urol,167: 103, 2002

14. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D.,Fondurulia, J., Chen, M. H. et al: Clinical utility of the per-centage of positive prostate biopsies in defining biochemicaloutcome after radical prostatectomy for patients with clini-cally localized prostate cancer. J Clin Oncol, 18: 1164, 2000

15. Presti, J. C., Jr., Shinohara, K., Bacchetti, P., Tigrani, V. andBhargava, V.: Positive fraction of systematic biopsies predictsrisk of relapse after radical prostatectomy. Urology, 52: 1079,1998

16. Freedland, S. J., Csathy, G. S., Dorey, F. and Aronson, W. J.:Percent prostate needle biopsy tissue with cancer is morepredictive of biochemical failure or adverse pathology afterradical prostatectomy than prostate specific antigen or Glea-son score. J Urol, 167: 516, 2002

17. Conrad, S., Graefen, M., Pichlmeier, U., Henke, R.-P.,Erbersdobler, A., Hammerer, P. G. et al: Prospective valida-tion of an algorithm with systematic sextant biopsy to predictpelvic lymph node metastasis in patients with clinically local-ized prostatic carcinoma. J Urol, 167: 521, 2002

18. Wise, A. M., Stamey, T. A., McNeal, J. E. and Clayton, J. L.:Morphologic and clinical significance of multifocal prostatecancers in radical prostatectomy specimens. Urology, 60: 264,2002

LACK OF SIGNIFICANCE OF SECONDARY CANCERS 463