ADVERSE PROGNOSTIC SIGNIFICANCE OF CAPSULAR INCISION WITHRADICAL RETROPUBIC PROSTATECTOMY

MATTHEW D. SHUFORD, MICHAEL S. COOKSON,*,† SAM S. CHANG,‡ AYUMI K. SHINTANI,ATHANASIOS TSIATIS, JOSEPH A. SMITH, JR.§ AND SCOTT B. SHAPPELL

From the Departments of Urologic Surgery and Pathology, Vanderbilt University Medical Center, Nashville, Tennessee

ABSTRACT

Purpose: The prognostic significance of capsular incision (CPI) at radical retropubic prostatec-tomy remains to be defined. To evaluate this we compared prostate specific antigen recurrence forwith CPI to that with established pathological groups.

Materials and Methods: From January 1998 to December 2000, 409 men underwent radicalretropubic prostatectomy at our medical center. CPI was defined as a positive posterior, lateralor posterolateral surgical margin without documented extraprostatic extension (EPE). Excludingpatients with preoperative androgen ablation, positive lymph nodes or seminal vesicle involve-ment there were 129 with organ confined disease and negative surgical margins (pT2/�M), 18with CPI, 29 with EPE and negative surgical margins (pT3a/�M), and 24 with EPE and positivesurgical margins (pT3a/�M). We compared time to biochemical recurrence among these 4 groupsusing Kaplan-Meier estimates. Cox proportional hazard regression was performed to determinethe HR of CPI vs the other groups, while controlling for age, prostate specific antigen, tumorvolume and Gleason score.

Results: The 3-year likelihood of freedom from biochemical recurrence in the CPI group was65%, for pT2/�M it was 96%, for pT3a/�M it was 91% and for pT3a/�M it was 58%. The adjustedHR with the 95% CI showed that the risk of biochemical recurrence with CPI was 8.4 timeshigher than that with pT2/�M (p � 0.002), 5.9 times higher than that with pT3a/�M (p � 0.046)and the same as that with pT3a/�M (p � 0.840).

Conclusions: Isolated posterior, lateral and posterolateral CPI by our definition occurs notuncommonly and it may represent true incision of the capsule and/or difficulty in diagnosing EPEdue to a lack of extraprostatic tissue in the surgical specimen. However, the prognostic signifi-cance of CPI as defined appears similar to that of pT3a with positive margins.

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

Pathological classification of prostate specimens followingradical retropubic prostatectomy (RRP) provides importantprognostic information. The outcomes of organ confined andextraprostatic disease have been well documented and thestatus of surgical margins at sites of extraprostatic extension(EPE) have prognostic significance.1–6 However, the outcomein patients with capsular incision (CPI) has yet to be fullydefined.

CPI has been defined as tumor extending to the inkedmargin with the absence of histologically documented EPE.4Some cases may represent true incision into the prostaticcapsule in the posterior, lateral and posterolateral aspects ofthe gland, where the capsule is best defined histologicallyand where incision is most likely to occur. CPI thus definedmay expose tumor or benign prostatic glands at the inkedsurgical margin.5 Other cases may represent tumor with apositive margin that was not classifiable as a pT3a tumorbecause of the lack of periprostatic tissue.

Defining the prognostic significance of CPI is particularly

relevant with nerve sparing prostatectomy, where dissectionof the neurovascular bundles may increase the risk of CPI.However, there is uncertainty regarding the significance ofthis finding. Since CPI was originally described in the early1990s, several studies have specifically addressed patientoutcomes.4–7 Most have suggested that CPI has no adverseeffect but it is essential to define more clearly its significance.We focused on tumor related outcomes in this group of pa-tients.

MATERIALS AND METHODS

From January 1998 through December 2000, 409 consec-utive men underwent RRP for clinically localized prostateadenocarcinoma, as previously described.8 We excluded 50men with positive lymph nodes or seminal vesicle involve-ment, 40 with preoperative antiandrogen therapy, 5 with noprostate cancer, 32 with a positive margin at the apex orbase, 30 randomized to immediate antiandrogen therapy aspart of a research study and 53 with followup less than 6months. Thus, 200 patients were available. Postoperativeserum prostate specific antigen (PSA) values were obtainedevery 6 months for 2 years and yearly thereafter. Biochemi-cal recurrence was defined as serum PSA greater than 0.2ng/ml.

Prostatic specimens were prepared as totally submittedwhole mounted specimens. Grade, stage, tumor volume andsurgical margin (SM) status were assessed by a single pa-thologist (SBS), as previously described.9 CPI was defined astumor extending to the inked margin at the posterior, lateral

Accepted for publication February 13, 2004.* Correspondence: Department of Urologic Surgery, Vanderbilt

University Medical Center, A-1302 Medical Center North, Nashville,Tennessee 37232 (telephone: 615-343-2338; FAX: 615-343-7023; e-mail: michael.cookson@vanderbilt.edu).

† Financial interest and/or other relationship with Photocure, Ab-bott, Ethicon, Cytogen, Schering Plough, Shire Pharmaceuticals, IlexOncology and Aventis.

‡ Financial interest and/or other relationship with Abbott, Cyto-gen, Praecis and GlaxoSmithKline.

§ Financial interest and/or other relationship with Pharmacia/Merck.

0022-5347/04/1721-0119/0 Vol. 172, 119–123, July 2004THE JOURNAL OF UROLOGY® Printed in U.S.A.Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION DOI: 10.1097/01.ju.0000132137.02846.ec

119

or posterolateral borders with the absence of histologicallydocumented EPE or other extraprostatic disease. CPI casesincluded only pT2 cases with tumor at the inked margin andnot benign prostate glands since the focus was tumor excisionand as agreed at a consensus conference.10 We limited ouranalysis to CPI patients with positive SMs at the posterior,lateral and posterolateral borders of the prostatic specimento focus on CPI as a consequence of nerve sparing surgeryand since they are the sites most commonly involved whenEPE is present.6 Figure 1 shows pathology specimens withtypical or more subtle differences in pT2/�M, pT3a/�M,pT3a/�M and pT2/�M (CPI).

Some studies have used the perhaps more directly relevantdefinition of CPI as cases with the histological appearance offrank incision into the capsule, which is more subjective toapply.5 To facilitate more direct comparisons among studieswe retrospectively reviewed our CPI cases, defined as pT2with positive posterior, lateral or posterolateral SMs. Whileblinded to progression status, cases were histopathologicallyclassified into those that appeared to be a consequence of trueincision into the prostate vs those in which inked externalcontours focally involved by tumor were smooth or indeter-minate.

The distribution of relevant baseline variables (age, tumorvolume, PSA and RRP Gleason score) within each group arepresented with the median and IQR. Equivalence of themedians across the 4 groups was assessed using the Kruskal-Wallis test. The 3-year cumulative probability of being freefrom biochemical recurrence was determined by Kaplan-Meier estimates. The 3-year cumulative incidence rate ofbiochemical recurrence was obtained along with the Clopper-Pearson 95% exact CI. Multivariate Cox proportional hazardregression was used to obtain the HRs of biochemical recur-

rence comparing capsular incision to the other 3 groups. Themultivariate regression model included tumor volume,Gleason score and preoperative serum PSA to adjust for theeffect of these variables.

RESULTS

Of the 200 patients included in the study 129 (64%) hadpathologically confirmed, organ confined disease and nega-tive surgical margins (pT2/�M), 29 (14.5%) had EPE withnegative margins (pT3a/�M), 24 (12%) had EPE with posi-tive surgical margins (pT3a/�M) and 18 (9%) had CPI. Of the18 CPI cases 13 had positive surgical margins at the postero-lateral aspect, 3 had a positive lateral margin and 3 had apositive posterior margin (positive lateral and posterior mar-gin in 1). Of the 18 cases 11 were believed to be histologicallycompatible with true capsular incision (surgeon incision intoprostate capsule).

Table 1 lists baseline tumor and demographic features ofthe 4 pathological categories. There was no statistically sig-nificant difference in median age among groups (p � 0.17).Median specimen Gleason score was 7 in all groups exceptpT2/�M (Gleason 6, p � 0.0008). Median PSA was actuallylowest in the CPI group at 5.3 ng/ml. It increased to 5.8 ng/mlin the pT2/�M group, 6.7 ng/ml in the pT3a/�M group and8.6 ng/ml in the pT3a/�M group with the differences beingstatistically significant. Median tumor volume was lowestin the pT2/�M group and highest in the pT3a/�M groupwith the differences between groups being significant(p � 0.0001).

Minimum mean followup was 22 months in all groups(mean 22 in all except pT3a/�M, which was 25). Biochemicalrecurrence occurred in only 4 of the 129 pT2/�M cases, (3%)

FIG. 1. Capsule and margin status in whole mounted RRP specimens that defined pathology subgroups analyzed. A, pT2/�M with tumorclearly confined within prostate with intact prostate capsule and inked surgical margin at top. B, pT3a/�M in case of EPE but negative SMat EPE site. Tumor extends through capsule, including site indicated by asterisk, but does not extend to inked margin at this focus (�).Tumor extends to within approximately 1 mm of inked margin but does not contact ink. Arrowheads indicate prostate capsule. C, pT3a/�Min case of EPE and positive SM at EPE site. Although there is little periprostatic soft tissue, tumor extends beyond capsule (asterisk), whereit may elicit desmoplastic response with fibrous tissue similar in appearance to prostate capsule, and extends to inked SM (arrow). D to F,representative pT2/�M or CPI cases. Arrowheads indicate capsule. D, tumor extends to inked SM (arrow), although there is no periprostaticsoft tissue and definable EPE at this focus. Prostate has smooth external contour at this aspect. E, tumor extends to inked SM at 2 distinctfoci (arrows) without evident periprostatic tissue or diagnosable EPE at these sites. Prostate has fairly smooth external contour at positivemargin site at top center but contour is more jagged and suggestive of iatrogenic incision at positive margin site at top right. Arrowheadsindicate prostate capsule. F, cauterized tumor extends to inked SM (arrows) without evident periprostatic tissue or diagnosable EPE at thesesites. Prostate external contour is jagged at positive margin site, suggesting possible iatrogenic incision into capsule at positive SM site.Arrowheads indicate prostate capsule. E and F, note ink dots and marks remaining from tumor delineation at sign out for tumor volumedetermination.

PROGNOSTIC SIGNIFICANCE OF CAPSULAR INCISION WITH PROSTATECTOMY120

at an average of 7.75, in 2 of the 29 pT3a/�M (6.8%) at anaverage of 14 and in 8 of the 24 pT3a/�M (33%) at an averageof 6.75 months. PSA recurrence was observed in 5 of the 18CPI cases (27%) at an average of 8 months (table 2). Figure 2shows the 3-year cumulative probability of freedom frombiochemical recurrence. The 3-year incidence rate per 1,000person-months of biochemical recurrence along with the 95%CI is shown. The 3-year incidence rate of recurrence was 1.4(95% CI 0.4 to 3.7) in the pT2/�M group, 18.5 (95% CI 6 to43.1) in the pT2/�M group, 3 (95% CI 0.4 to 10.9) in thepT3a/�M group and 21.7 (95% CI 9.3 to 42.7) in the pT3a/�Mgroup. Table 2 also shows adjusted HRs with the 95% CIcontrolling for serum PSA, Gleason score and tumor volume.

The CPI group had an 8.4 times higher risk of recurrencethan the pT2/�M group (95% CI 2.2 to 31.8, p � 0.002), a 5.97times higher risk than the pT3a/�M group (95% CI 1.0 to34.6, p � 0.046) and virtually the same risk of recurrence asthe pT3a/�M groups (HR 1.1, 95% CI 0.32 to 4.1, p � 0.84).These HRs demonstrate that CPI has a significantly worseprognosis than pT2/�M and pT3a/�M, and it does not differfrom pT3a/�M.

Of the 18 patients with CPI 12 underwent a bilateral nervesparing procedure and 2 underwent a unilateral nerve spar-ing procedure. However, regarding the possible future reduc-tion of CPI by planned wide excision eg on the side with adetermined increased risk of EPE, these 2 patients weretreated with a left nerve sparing procedure but the positiveSMs were on the right side, including one that histologicallyappeared as true CPI. Four of the 18 patients with CPIunderwent bilateral wide excision. Three of the 5 patientswith CPI who had PSA recurrence underwent wide excision,while 1 each underwent a bilateral and a unilateral nervesparing procedure. In the 1 case of recurrence after a left onlynerve sparing procedure the location of CPI was the rightside. Three CPI recurrences were histologically interpretedas true surgical incision and 2 were indeterminate.

DISCUSSION

CPI has been defined in multiple ways, including histolog-ically recognized surgical incision into the capsule (includingtumor or benign glands at ink) or tumors negative for EPEwith positive surgical margins. It is most typically consideredto indicate a positive SM as a consequence of iatrogenicsurgical incision into the prostate capsule. The prostate cap-sule is best defined histologically in the posterior, lateral andposterolateral aspects of the gland. Similarly because EPEmost typically occurs in the posterolateral aspect of the pros-tate,11 a truly positive SM as a result of surgical incision intothe capsule is most likely to occur at this location.

Ohori et al defined CPI as “a positive surgical margin in anarea without periprostatic tissue. . .in a cancer confined tothe prostate, indicating that the plane of dissection entered

the prostatic capsule.”4 Barocas et al subsequently definedCPI as cases where “the surgeon inadvertently develops theplane of resection within the prostate rather than exterior tothe prostate.”5 They broadened the definition to ensure theinclusion of all cases independent of margin status, CPI intobenign or malignant glands and the presence or absence ofEPE. These 2 studies did not show that CPI is an adverseprognostic factor for biochemical progression following radi-cal prostatectomy.

Our definition is similar to that of Ohori et al.4 Some seriesalso included positive apical margins. However, the prostaticcapsule is not well defined at the true apex and issues ofapical dissection are different than those regarding decisionsfor nerve sparing surgery. Furthermore, involvement of thesemargins is less related to mechanisms of usual EPE11 andpreoperative strategies to identify patients at risk for EPE.12

Our positive apical margin rate is slightly less than ourpositive posterolateral surgical margin rate (eg 35% vs 45%for clinical stage T1c and 37% vs 43% for clinical stage T2cases).9 Our CPI cases did not include those with positiveapical margins. However, the majority of our CPI cases ap-peared to have histologically recognizable incision into thecapsule. Hence, our cases closely match those defined byBarocas et al.5 Although that study included cases of tumoror benign glands at the inked margin, the majority of suchcases had prostatic carcinoma at the margin. Also, as in ourseries, the study of Barocas et al did not include otherwisepT2 tumors with only positive apical margins. However, incontrast to the results of the current study, these previousinvestigations did not describe capsular incision as an ad-verse prognostic factor.4, 5

Our definition of CPI includes incision into the capsule bythe surgeon as well as tumor extending to the edge of thespecimen without adequate periprostatic tissue to be certainof EPE status. Precisely defining the prognosis of CPI notonly provides important prognostic information, but also mayindicate the need for better preoperative selection of patientsfor a nerve sparing approach. In this study 12 of the 18patients with CPI underwent a bilateral nerve sparing pro-cedure, of whom only 1 had biochemical recurrence. Threepatients with recurrence underwent bilateral wide excisionand 2 who underwent a unilateral nerve sparing procedure,including 1 with recurrence, had CPI on the opposite side.While these numbers are small, they suggest that a nervesparing procedure is not the only risk factor leading to bio-chemical recurrence associated with CPI.

Of the prostates of our 409 patients 200 met the criteria forinclusion in our study and 18 (9%) had CPI. The incidence ofCPI was 3.1% to 43%, as reported in the literature.2–7, 13 Thislarge disparity may be due to several reasons, includingdifferences in surgical technique, the frequency with whichnerve sparing surgery is performed and the stringency of

TABLE 1. Clinical features and RRP tumor pathology by groups analyzed

Variable pT2/�M CPI pT3a/�M pT3a/�M p Value

No. pts 129 18 29 24Median age (range) 59 (40–72) 62 (47–71) 61 (40–71) 62.5 (43–76) 0.17Median Gleason score (range) 6 (5–9) 7 (5–8) 7 (5–9) 7 (5–9) 0.0008Median ng/ml serum PSA (range) 5.8 (0.5–24) 5.3 (3.4–20.8) 6.7 (0.27–27) 8.6 (4.2–23) 0.01Median cc tumor vol (range) 0.87 (0.02–8.77) 1.31 (0.25–8.2) 2.90 (0.58–17.5) 3.3 (1.28–11.6) �0.0001

TABLE 2. Biochemical recurrence in patients in different RRP pathological classifications

pT2/�M CPI pT3a/�M pT3a/�M

No. pts 129 18 29 24No. recurrences (%) 4 (3) 5 (27) 2 (6.8) 8 (33)3-Yr incidence rate/1,000 person-mos (95% exact CI) 1.4 (0.4, 3.7) 18.5 (6, 43.1) 3 (0.4, 10.9) 21.7 (9.3, 42.7)Adjusted CPI HR vs other (95% CI)* 8.4 (2.2–31.8) 1.0 5.97 (1.0–34.6) 1.14 (0.32–4.1)p Value 0.002 0.046 0.84

* Recurrence rate in CPI group as reference to each of remaining 3 groups.

PROGNOSTIC SIGNIFICANCE OF CAPSULAR INCISION WITH PROSTATECTOMY 121

pathological analysis. Pathological analysis may not be per-formed using totally submitted or whole mounted prostaticspecimens and, therefore, some instances of CPI may bemissed. For example, in the study of Barocas et al, in whichwhole mounted specimens were not used, there was a lowerincidence of CPI of 3.1%.5 The incidence of CPI in the studyof Cheng et al6 and the current series were higher, whichmay reflect differences in pathological analysis since the 2studies used whole mounted prostate specimens. The inci-dence of CPI in the literature is likely under reported due toambiguity in the definition and the method of pathologicalanalysis.

Our 3-year actuarial likelihood of freedom from recurrencefor CPI was significantly worse than for pT2/�M or pT3a/�Mdisease (fig. 2). In fact, patients with CPI were more than 8times more likely to have biochemical recurrence than thosewith pT2/�M disease and more than 5 times more likely tohave recurrence than those with pT3a/�M disease. Previousinvestigators have analyzed the outcome in CPI cases (table3). Freedom from recurrence is similar in all studies withregard to pT2/�M, pT3a/�M and pT3a/�M disease. How-ever, there is a difference with regard to the prognostic sig-nificance of CPI. Previous studies do not show that CPIconfers a statistically significantly worse prognosis thanpT2/�M disease with the possible exception of the studies ofCheng6 and Boccon-Gibod13 et al. Boccon-Gibod et al specif-ically compared the perineal approach with the retropubicapproach but found that CPI into tumor had a biochemicalrecurrence-free survival rate of 63% vs 100% for organ con-fined disease. Cheng et al found that biochemical recurrence-free survival for CPI into tumor was 78% vs 90% for organconfined disease.6

In contrast, the studies of Ohori4 and Barocas5 et al showedno adverse effect of CPI. Numerous elements could contrib-ute to the differences in reported outcome, including patientselection, surgical technique and the method of pathologicalanalysis. It is also worth noting that in the study of Barocaset al patients with CPI had a lower 3-year likelihood offreedom from disease progression than those with pT2/�Mand pT3a/�M disease, (ie 87.8% vs 96.4% and 91.3%, respec-tively), but not as low as for pT3a/�M (73.9%). The lowerincidences of CPI and the inclusion of positive apical marginsin pT2/�M cases in these studies may have biased againstprogression because positive apical margins do not appear toimpart as adverse a prognosis as positive posterolateral mar-gins.14 In contrast, the number of patients with CPI in thestudy of Ohori et al4 was more similar to that in the currentreport. Another observation that can be made from theseresults is that margin positivity is more important than EPEwith regard to biochemical progression. This is consistentwith the Gleason, PSA, seminal vesicle and margin score ofBlute et al for PSA recurrence, which showed margin posi-tivity to be a significant predictor of biochemical recurrence,while extraprostatic extension was not.15

There are some limitations to our analysis. The first limi-tation is relatively short followup. While continued followupwould likely show more recurrences in all groups, the well-defined pathological categories of pT2, pT3a/�M andpT3a/�M already accurately reflect the biochemical recur-rence rates of previous studies with longer followup.4–6 CPIhas already begun to diverge as an adverse prognostic indi-cator and any further recurrences with time would only in-creases its poor prognostic implications. A second limitationis our relatively small numbers. Under powered analysis can

FIG. 2. Three-year likelihood of freedom from biochemical recurrence in patients in different RRP pathological classifications

TABLE 3. Outcome comparison of CPI in institutional studies

Variable Barocas et al5 Ohori et al4 Cheng et al6 Present Series

No. pts 368 478 298 200No. CPI 92 23 72 18% CPI 3.1* 4.8 19* 9% Recurrence-free likelihood (pathological classification):

pT2 96.4 95 90 96Benign glands

CPI (malignant glands) 88.1/87.8 100 78 64pT3a/�M 91.3 84 90 90pT3a/�M 73.9 59 55 58

* Reported data including men not in outcome analysis.

PROGNOSTIC SIGNIFICANCE OF CAPSULAR INCISION WITH PROSTATECTOMY122

inflate p values. However, our differences in recurrence ratesalready show significant differences between the CPI, andpT3a/�M and pT2/�M groups, and larger numbers wouldnot change these results. The insignificant difference be-tween CPI and pT3a/�M is unlikely to change, given that the3-year cumulative probability of freedom from biochemicalrecurrence curves almost overlap, but these results must beverified in a larger study. Any effect of sample size is re-flected in the Clopper Peason exact 95% CI.

CONCLUSIONS

Isolated posterior, lateral and posterolateral CI after RRPcan be identified histopathologically in some patients. Theprognostic significance of CI appears similar to that of pT3awith positive margins. Future attempts to decrease posteriorand lateral CI include improved preoperative assessment ofrisk factors as well as intraoperative identification of pa-tients better treated with wide excision to minimize the riskof cancer recurrence. Meticulous reporting of surgical tech-nique and rigorous pathological assessment of the surgicalspecimen based on a clear definition of CI will help assessfurther the incidence and significance of CI.

REFERENCES

1. Han, M., Partin, A. W., Pound, C. R., Epstein, J. I. and Walsh,P. C.: Long-term biochemical disease-free and cancer-specificsurvival following anatomic radical retropubic prostatectomy.The 15-year John’s Hopkins experience. Urol Clin North Am,28: 555, 2001

2. van den Ouden, D., Bentvelsen, F. M., Boeve, E. R. and Schroder,F. H.: Positive margins after radical prostatectomy: correla-tion with local recurrence and distant progression. Br J Urol,72: 489, 1993

3. Epstein, J. I., Partin, A. W., Sauvageot, J. and Walsh, P. C.:Prediction of progression following radical prostatectomy. Amultivariate analysis of 721 men with long-term follow-up.Am J Surg Pathol, 20: 286, 1996

4. Ohori, M., Wheeler, T. M., Kattan, M. W., Goto, Y. and Scardino,P. T.: Prognostic significance of positive surgical margins inradical prostatectomy specimens. J Urol, 154: 1818, 1995

5. Barocas, D. A., Han, M., Epstein, J. I., Chan, D. Y., Trock, B. J.,

Walsh, P. W. et al: Does capsular incision at radical retropubicprostatectomy affect disease-free survival in otherwise organ-confined prostate cancer? Urology, 58: 746, 2001

6. Cheng, L., Darson, M. F., Bergstralh, E. J., Slezak, J., Myers,R. P. and Bostwick, D. G.: Correlation of margin status andextraprostatic extension with progression of prostate carci-noma. Cancer, 86: 1775, 1999

7. Stamey, T. A., Villers, A. A., McNeal, J. E., Link, P. C. andFreiha, F. S.: Positive surgical margins at radical prostatec-tomy: importance of the apical dissection. J Urol, 143: 1166,1990

8. Smith, J. A., Jr.: Techniques to decrease morbidity with radicalprostatectomy. AUA Update Series, vol. 19, lesson 35, p. 273,2000

9. Jack, G. S., Cookson, M. S., Coffey, C. S., Vader, V., Roberts,R. L., Chang, S. S. et al: Pathological parameters of radicalprostatectomy for clinical stages T1c versus T2 prostate ade-nocarcinoma: decreased pathological stage and increased de-tection of transition zone tumors. J Urol, 168: 519, 2002

10. Sakr, W. A., Wheeler, T. M., Blute, M., Bodo, M., Calle-Rodrigue,R., Henson, D. E. et al: Staging and reporting of prostatecancer—sampling of the radical prostatectomy specimen. Can-cer, 78: 366, 1996

11. McNeal, J. E.: Cancer volume and site of origin of adenocarci-noma in the prostate: relationship to local and distant spread.Hum Pathol, 23: 258, 1992

12. Bismar, T. A., Lewis, J. S., Jr., Vollmer, R. T. and Humphrey,P. A.: Multiple measures of carcinoma extent versus perineu-ral invasion in prostate needle biopsy tissue in prediction ofpathologic stage in a screening population. Am J Surg Pathol,27: 432, 2003

13. Boccon-Gibod, L., Ravery, V., Vordos, D., Toublanc, M., Delmas,V. and Boccon-Gibod, L.: Radical prostatectomy for prostatecancer: the perineal approach increases the risk of surgicallyinduced positive margins and capsular incisions. J Urol, 160:1383, 1998

14. Epstein, J. I.: Incidence and significance of positive margins inradical prostatectomy specimens. Urol Clin North Am, 23: 651,1996

15. Blute, M. L., Bergstralh, E. J., Iocca, A., Scherer, B. and Zincke,H.: Use of Gleason score, prostate specific antigen, seminalvesicle and margin status to predict biochemical failure afterradical prostatectomy. J Urol, 165: 119, 2001

PROGNOSTIC SIGNIFICANCE OF CAPSULAR INCISION WITH PROSTATECTOMY 123