- Email: [email protected]
Influence of Capsular Penetration on Progression Following Radical Prostatectomy: A Study of 196 Cases with Long-Term Followup
0022-534 7/93/lSOl-0135$03,00/0 Vol. 150, 135-141,
THE JOURNAL OF UROLOGY
Printed in
Copyright© 1993 by AMERICAN UROLOGICAL ASSOCIATION, INC.
INFLUENCE OF CAPSULAR PENETRATION ON PROGRESSION FOLLOWING RADICAL PROSTATECTOMY: A STUDY OF 196 CASES WITH LONG-TERM FOLLOWUP JONATHAN I. EPSTEIN,* MARNE J. CARMICHAEL, GALINA PIZOVt
AND
PATRICK C. WALSH
From the Departments of Pathology and Urology, The Johns Hopkins University School of Medicine and the James Buchanen Brady Urological Institute, The Johns Hopkins Hospital, Baltimore, Maryland
ABSTRACT
We studied 196 radical prostatectomy cases performed for clinical stage B prostate cancer with capsular penetration; in all cases seminal vesicles and lymph nodes were free of tumor. The mean followup in patients who showed no evidence of progression was 5 years. Focal capsular penetration was seen in 93 cases. There was no difference in progression in this group, irrespective of whether margins were negative or positive. High grade tumors (Gleason score 7 or more) had a significantly higher risk of progression compared to lower grade tumors (p = 0.0002). Established capsular penetration was seen in 103 tumors. Cancers with established capsular penetration had a higher risk of progression than those with focal capsular penetration. Established capsular penetration tumors were stratified into 3 groups with increasing risks of progression: 1) margins were negative and grade was low, 2) margins were positive or grade was high yet both adverse features were not present or 3) margins were positive and grade was high. The differences in progression among these 3 groups were statistically significant. Because of the negligible influence of positive margins in patients with focal capsular penetration the status of capsular margins should not influence the decision on whether to administer immediate postoperative adjuvant therapy. To evaluate the efficacy of adjuvant therapy following radical prostatectomy, tumors with capsular penetration should be stratified into groups having similar risks of progression according to the extent of capsular penetration, surgical margins of resection and grade. KEY WORDS:
prostatic neoplasms, neoplasm invasiveness, prostatectomy
An often repeated truism in urology is that once prostate cancer has spread outside the gland the window of opportunity to cure the patient is lost. Studies have demonstrated that capsular penetration is related to other variables associated with aggressive tumor behavior, such as increased tumor volume, seminal vesicle invasion, lymph node metastases and poor histological grade. 1 However, only a few studies have assessed the relationship between capsular penetration and recurrence following radical prostatectomy. 2- 5 Prior studies evaluating the prognostic significance of capsular penetration have suffered from the following deficiencies: 1) prostates were not serially sectioned and embedded in their entirety, 2) radical prostatectomy specimens were not stratified by capsular margins of resection, 3) Gleason score 7 tumor was considered as intermediate grade tumor rather than high grade tumor, 4) cases when tumor extended into but not through the capsule were considered as showing capsular penetration, 5) extent of capsular penetration was not quantified, 6) pathological information was gleaned from pathology reports in which ambiguous terminology may have led to erroneous interpretation and 7) followup information antedated the use of serum prostate specific antigen (PSA) levels to detect recurrence. The current study evaluated 196 radical prostatectomy specimens with pathological stage C disease in which the seminal vesicles and pelvic lymph nodes were free of tumor. Cases were stratified according to the extent of capsular penetration, status of capsular margins and Gleason score. These parameters were correlated with progression. MATERIALS AND METHODS
Between April 1982 and the end of 1988, 600 consecutive men underwent retropubic prostatectomy by 1 surgeon (P. C. Accepted for publication December 18, 1992.
* Requests for reprints: Department of Pathology, The Johns Hopkins Hospital, Baltimore, Maryland 21205. t Current address: Hadassah University Hospital, Jerusalem, Israel.
W.) at our hospital. Preoperatively, all patients were evaluated by careful palpation of the primary lesion, serum enzymatic prostatic acid phosphatase (PAP), bone scan, pathological review of the biopsy material and, since June 1986, PSA levels. From the 600 men 196 cases satisfied the following requirements and were included in this study: 1) the patients had clinical stage B disease, 2) all patients whose disease did not progress had a minimum of 2 years of follovvup, 3) all men had at least 1 serum PSA determination postoperatively and had a PSA value at the last visit, 4) all prostates were serially sectioned, embedded in their entirety and assessable as to margins of resection, 5) all tumors demonstrated penetration through the capsule into periprostatic soft tissue and 6) in all cases seminal vesicles and lymph nodes were free of tumor. A total of 15 cases in which tumor invaded the periseminal vesicle soft tissue was also dropped from the analysis, since a prior study showed these tumors to have an intermediate prognosis between those showing capsular penetration alone and those showing seminal vesicle invasion. 6 The surgical procedure on the first 100 patients was performed as described in 1983. 7 Thereafter, once the anatomy of the pelvic fascia and neurovascular bundles was better characterized it was possible to excise the neurovascular bundle(s) widely when indicated. All subsequent procedures were performed via the technique presently in use. 8 Only 2 patients received immediate adjuvant radiotherapy to the prostatic bed and both of these men eventually experienced progression. Otherwise, no patient received adjuvant radiotherapy or hormonal therapy. All men were evaluated at 3 and 6 months postoperatively. Patients with tumor involving the surgical margins were followed semiannually, while all others were evaluated yearly. Followup examination included a complete history, rectal examination, serum acid phosphatase and, since June 1986, serum PSA. Because the patients came from 47 of
135
136
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
the 50 states in the United States it was not possible to evaluate all men at our hospital yearly. However, in those men followup information from their referring urologist was obtained by mail yearly. PSA was considered elevated if it was greater than the lower limit of detection for the reference laboratory; at our hospital using the Tandem-R PSA* solid phase 2-site immunoradiometric assay the lower limit of detection is 0.2 ng./ml. Patients with symptoms of recurrent disease or elevated serum markers were evaluated with a complete history, a repeat careful rectal examination, chest x-ray, bone scan and computerized tomography of the pelvis. Local recurrence was defined as palpable induration at the operative site; in many cases if this induration was associated with an elevated PSA level the men were classified as having local recurrence without biopsy confirmation. Distant metastases were confirmed by bone scan. Following resection the intact prostate was weighed, coated over the entire surface with india ink and fixed in 10% buffered formalin for 18 to 24 hours. Following fixation the distal and proximal thin shave urethral margins were removed for histological examination. These shave margins were taken as cross sections, with the urethra running through the center. The gland was then serially sectioned at 2 to 3 mm. intervals perpendicular to the long (apical-basal) axis of the gland. Each cross section of the prostate was further divided (that is, halves, quarters and so forth) to conform to the size of a cassette. The entire prostate was embedded and designated in a manner permitting the localization of each section within the prostate. The tumor was graded using the Gleason grading system. Capsular penetration was defined as tumor extending out of the prostate into periprostatic soft tissue. Capsular penetration was subdivided into those in which only a few neoplastic glands were present exterior to the prostate (focal capsular penetration, fig. 1, A) and cases with more extensive extraprostatic spread of tumor (established capsular penetration, fig. 1, B). Cases with focal capsular penetration revealed on only 1 or 2 slides a few cancer glands outside of the prostate, which may not have been identified had we not totally submitted the prostate for histological analysis. Capsular margins of resection were designated as negative or positive, as previously described.9 Statistics were performed using the statistics graphic data measurement software program. t End points measured as interval to progression were calculated by the Kaplan-Meier method for estimation. Differences between Kaplan-Meier curves were tested for statistical significance using the Wilcoxon-Gehan test. RESULTS
Focal capsular penetration. There were 93 cases showing focal capsular penetration: 77 men (82%) showed no evidence of progression with a mean followup of 5 years and 16 men experienced progression. The overall progression-free probability curve for tumors with focal capsular penetration is illustrated in figure 2. Margins were free of tumor in 55 of these specimens and in 38 cases the margins were focally positive. There was no difference in progression, irrespective of whether margins were negative or positive (fig. 3, A). The mean Gleason grade for tumors with focal capsular penetration was 6.1. High grade tumors had a significantly increased risk of progression compared to tumors with a Gleason score of less than 7 (p = 0.0002, fig. 3, B). The interval to failure stratified by type of recurrence is shown in table 1. There were 5 patients with focal capsular penetration who did not have a serum PSA level drawn until 2 to 3 years following radical prostatectomy; progression was noted at that time. Given that the curves for tumors with positive and negative margins are superimposed, and the marked difference in progression when tumors were stratified
* Hybritech Inc., San Diego, California.
t STATA Computing Resource, Los Angeles, California.
FIG. 1. A, several neoplastic glands immediately exterior to prostate show perineural invasion in periprostatic tissue (right side), consistent with focal capsular penetration. Reduced from X25. B, established capsular penetration resulting in irregular protuberance to capsular surface because of tumor-associated desmoplasia. Reduced from Xl8.
by grade, our results would not have been altered significantly whether these few patients had progression 1 or 2 years earlier. Established capsular penetration. A total of 103 tumors demonstrated established capsular penetration. The mean followup for patients showing no evidence of progression was 5 years, while 36 men (35%) experienced progression. The actuarial risk of progression in tumors with established capsular penetration is illustrated in figure 2. The difference in progression in cancers with established capsular penetration compared to those with focal capsular penetration just missed statistical significance (p = 0.06). Of the cases 29 had negative margins, 70 tumors had focally positive margins and 4 showed extensively positive margins. Tumors removed with negative margins had a lower likelihood of progressing (p = 0.006, fig. 4, A). The mean Gleason score for tumor with established capsular penetration was 6. 7. The difference in progression when tumors with established capsular penetration were stratified by Gleason score barely missed statistical significance (p = 0.07, fig. 4, B). Because low grade tumors (Gleason score less than 7) with positive margins had similar progression rates as high grade (Gleason score 7 or more) tumors with negative margins, these 2 groups were combined. The resulting 3 groups had significantly different progression probabilities: low risk-negative margins and low grade, intermediate risk-negative margins and high grade tumor or positive margins and low grade tumor, and high risk-positive margins and high grade (fig. 5). The
137
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
Correlation of Focal and Established Capsular Penetration with Progression '.l (7) >, _µ
(18)
l:'.:(:==9l=c::-:--u__
..... ..... .....
_j__~c6)
__j__
(4)
.75 -
D
ro
focal cap pen.
D 0
L CL
"'"'
.5 -
....L
(3)
I
C
established
0
. .-,
in
cap. pen.
.25 -
I!>
L
Ol
D
L
a.. 0 I
I
I
.4
0
8
first year of recurrence FIG. 2. Correlation of extent of capsular penetration (cap. pen.) with progression. Numbers in parentheses represent number of censored patients.
A
Focal Capsular Penetration: Correlation of Surgical ~arg1ns with Progression 1 -
>
-
(12)
+-'
C1 ll
pas. margin
(5)
rl
(4)
rl
D
ro
D
(7)
(4)
(9)
B
13)
(2)
11 1 _ ,----i1c_~ ( 7 )
(2)
(3)
.75 -
~
ro
D 0
75 -
(6)
Cll
C1l
L
CL
.5 -
Gleason ~7
0)
I
0)
£:
s:
5 -
I
If) If)
L
(31 Gleason
D
C
QJ
(5) (5)
rl
0) 0)
L
(12)
rl
margin
Q_
(10)
~
neg.
L
(11)
>-
(2)
0
Focal Capsular Penetration: Correlation of Gleason Sum with Progression
C 0 rl If)
.25 -
01 0 L
Ul Q)
25 -
L
01
Q_
0 (_
0 -
Q_
I 0
I
I
I
2
3
4
I 5
I
I
I
I
5
7
8
9
0 I
I
I
0
3
5
first year of recurrence FIG,
I
8
I
g
first year of recurrence
3. Focal capsular penetration. A, correlation of surgical margins with progression. B, correlation of Gleason sum with progression
curve for low grade tumors with negative margins decreases to a progression-free probability of zero at year 8. However, this decrease represents only 1 patient followed to 8 years who had progression. The difference in progression between the low and intermediate risk tumors was p = 0.05. High risk tumors had a higher likelihood of progression as compared to the intermediate risk tumors (p = 0.04). The difference in progression between patients with low and high risk tumors was highly significant (p = 0.004). The interval to progression in these different groups, classified by the type of recurrence, is depicted in table 2. In 14 patients the first postoperative serum PSA levels were drawn 2 or more years following radical prostatectomy, with progression noted at that time. All 14 of these cases had positive margins and 12 of these tumors were high grade. Since these men may have had progression at even an earlier date had serum PSA levels been available, the differences in tumors stratified by margins and grade may be more significant than what was calculated using the information available. Of the 14 men with delayed measurements of serum PSA 12 had high grade tumor with positive margins. Furthermore, all 3 patients
with PSA levels drawn 4 to 5 years following radical prostatecTABLE 1.
Type of Failure
Focal capsular penetration: interval to progression No. Cases
Yrs. to Failure
Gleason score less than 7 (59 pts.)
PSA Local Distant
4 0
4 (mean), 2-8 (range)*
0
Gleason score 7 or more (34 pts.)
PSA Local Local Local Distant
8 1 1 1 1
3 (mean), 2-5 (range)t PSA at 2, local at 3+ PSA at 2, local at 4 Local at 3 Distant at 2
* Delayed measurement of PSA in 2 men (2 and 3 years postoperatively) with abnormal PSA noted at first measurement. t Delayed measurement of PSA in 2 men (2 and 2 years postoperatively) with abnormal PSA noted at first measurement. Delayed measurement of PSA with abnormal PSA noted at first measurement.
+
138
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
A
B
Established Capsular Penetration: Correlation of Surgical Margins with Progression (1)
1-
1 -----=====ic_
_c_~_'19)
(3)
?;'
>,
~
(6)
(4)
rl rl
rl
:.8 ro
D
negative
.75 -
0
L
(_
Cl..
Cl..
.;:
UJ
Q) (_
.5 -
I C
positive
.s
ClJ L 01 0 L
-s
.5 -
I C
Cl) Cl)
.75 -
"'
D 0
margin
D
Q) Q)
Established Capsular Penetration: Correlation of Gleason Sum with Progression
(2)
margin
en
Cl) Q)
.25 -
Gleason ~7 .25 -
L 01 0 L
D..
Cl..
0 -
0 I
0
I
2
I
3
I
I
4
first year of recurrence
6
I
I
7
B
I
I
4
0
I
5
I
6
I
B
first year of recurrence
FIG. 4. Established capsular penetration. A, correlation of surgical margins with progression. B, correlation of Gleason sum with progression
Established Capsular Penetration: Correlation of Surgical Margins and Gleason Sum wi tn Progression (3)
1-
(4)
(2)
(2)
neg. margins, Gleason
>, .µ
..... •rl •rl
.0
ro
.75 -
.0 D L CL
w w L
pos. margins. Gleason
C3)( 2) .5 -
't-
i C 0
..... "'w (/)
L
pos. margins, Gleason~?
.25 -
CJ)
0
L
CL
0 I
0
I
2
I
3
I
4
I
5
I
6
first year of recurrence
I
7
I
8
FIG. 5. Established capsular penetration. Correlation of surgical margins and Gleason sum with progression. neg., negative. pos., positive
tomy had high grade tumors with positive margins. It is likely that at least some of these men would have been shown to have had progression at an earlier date had serum PSA levels been available immediately following radical prostatectomy. Therefore, tumors in the high risk group with positive margins and high grade tumors are probably associated with even a worse risk of progression as compared to the intermediate and lower risk groups than we were able to calculate using the available data. Also, because measurements of PSA were more often delayed in tumors with established penetration compared to cases with focal penetration, tumors with established penetration probably have a relatively worse progression rate than illustrated in figure 2. DISCUSSION
Despite the use of terminology such as capsular penetration, the prostate lacks a well defined capsule. 10 • 11 In some areas there appears to be a separation between the most peripheral nonneoplastic acini and the edge of the prostate, giving the appearance of a fibromuscular capsule. However, elsewhere nonneoplastic glands extend right up to the edge of the prostate. Towards the base of the prostate the edge of the gland is even more ill-defined with nonneoplastic glands appearing to extend
out of the prostate. Anteriorly, the edge of the gland is also poorly delimited where the anterior fibromuscular stroma of the prostate merges in with smooth muscle that extends to the pubic bone. Because one cannot distinguish where the prostatic capsule begins and the normal prostatic stroma ends, it makes no sense to classify tumors as going into but not through the prostatic capsule. Furthermore, the majority of studies have demonstrated that prognosis is adversely affected only when tumor penetrates full thickness through the prostatic capsule and extends into adjacent periprostatic soft tissue. 12 Consequently, tumors should be categorized as either being confined to the gland or extending out of the prostate. Since some physicians differentiate between capsular penetration and capsular perforation, pathologists should define their use of these terms until nomenclature is standardized. Recognizing that the prostate lacks a true discrete fibrous or fibromuscular capsule, the term prostatic capsule, nevertheless, is still convenient to denote the edge of the gland. Only a few studies have measured the. extent of capsular penetration. Stamey et al quantified capsular penetration based on the length of capsular penetration summed along different slides. 12 However, because prostate cancer often spreads along nerves parallel to the prostatic capsule one can see tumor barely
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY TABLE 2.
Established capsular penetration: interval to progression
Type of Failure
No. Cases
Yrs. to Failure
Gleason score 7 or more and neg. margins (13 pts.) PSA 1 8 Local 0 Distant 0 Gleason score less than 7 and pas. margins or Gleason score 7 or more and neg. margins (41 pts.)
PSA Local Local Local Distant
8 1 1 1
4 (mean), 2-7 (range)* PSA at 1, local at 2 Local at 4 Local at 6
0
Gleason score 7 or more and pas. margins (49 pts.)
PSA Local Local Local Local Local Distant Distant Distant Distant
12 1 2 1 2 1 2 1 1 1
4 (mean), 2-6 (range)t PSA at 3, local at 4+ Local at 4+ PSA at 3, local at 5+ Local at 5 PSA at 7, local at 8 PSA at 2, distant at 3 PSA at 2, distant at 3+ Distant at 3+ PSA at 2, local, distant at 4+
* Delayed measurement of PSA in 2 men (2 and 3 years postoperatively) with abnormal PSA noted at time of first measurement. t Delayed measurement of PSA in 5 men (2, 2, 3, 4 and 5 years postoperatively) with abnormal PSA noted at time of first measurement. Delayed measurement of PSA with abnormal PSA noted at time of first measurement.
+
out of the gland skimming along for significant distances. Also, cases with extensive spread of tumor perpendicularly to the prostatic capsule into the neurovascular bundle might be misdiagnosed as showing minimal capsular penetration if there is not significant horizontal spread of tumor as well. At our institution we have classified cases of capsular penetration as showing focal capsular penetration when only a few neoplastic glands are present in periprostatic soft tissue. 9 These glands are barely exterior to the prostate and tend to grow horizontally parallel to the prostate rather than extending away from the prostatic gland. Tumors with a greater degree of extracapsular spread are designated as having established capsular penetration. Using this definition, our study demonstrates that tumors with established capsular penetration had a higher risk of progression than those showing only focal capsular penetration. The only parameter that helped stratify tumors with focal capsular penetration was the Gleason score. A somewhat surprising finding was the virtually identical progression rate in cancers with focal capsular penetration regardless of whether capsular margins were positive or negative. There are 2 potential explanations for this disparity. First, many of the margins designated as positive in patients with focal capsular penetration represent an artifactually positive margin. In general, there exists in prostate cancer a great potential for over-diagnosing positive margins as a consequence of the limited, easily disruptable soft tissue that surrounds radical prostatectomy specimens.13 In prostate cancers when there is only minimal tumor spread out of the prostate and there is a positive margin, our study suggests that there is even a greater likelihood that these positive margins are artifactual. The other explanation is that tumors displaying only focal capsular penetration are inherently less aggressive, and even if tumor is cut across it has little potential for regrowth and progression. Progression in cases of established penetration was adversely influenced by positive margins and high grade tumor. We were able to stratify tumors into 3 groups associated with increasing risks of progression based on whether 1) margins were negative and grade was low, 2) margins were either positive or grade was
139
high and yet both adverse features were not present, or 3) margins were positive and grade was high. It is difficult to compare the current findings with those of prior studies. Paulson et al compared 74 tumors that had established capsular penetration and negative margins (specimen confined) to 123 tumors that had established capsular penetration and positive margins (margin positive). 3At 10 years postoperatively the disease-free probabilities in these 2 groups were 70% and 40%, respectively. In their study progression was determined by an elevated PAP level. Other differences between our study and theirs were: in their study the inclusion of cases with positive seminal vesicles and stage A disease, a significant number of patients received postoperative radiotherapy although this was not found to influence failure and cases were subdivided into Gleason scores of 7 or less versus a Gleason score of more than 7. We and others have demonstrated that Gleason score 7 tumor behaves more closely to high grade tumor than intermediate grade tumor. 9• 14 Schellhammer analyzed a group of patients with pathological stage C disease, including cases with positive seminal vesicles, in whom recurrence was determined by biopsy. 2 There was no evidence of disease postoperatively in 77% and 54% of the men at 5 and 10 years, respectively. Data from his study were obtained by review of pathology reports rather than by direct interaction with pathologists. As he quotes, "Terminology used to describe involvement of the capsule is also imprecise. The terms involvement, invasion, permeation, perforation, or penetration are often used interchangeably and do not identify specifically whether the tumor involved the inner most layer of the fibrous capsule or extends through the capsule to the pericapsular tissue." This statement reflects that the terminology used to describe the relationship of tumor to the capsule is often ambiguous, and attempts to glean data from pathology reports regarding the status of the capsule are fraught with inaccuracies. The importance of having direct involvement by a study pathologist was recently demonstrated at our institution; when urologists initially reviewed pathology reports generated by different pathologists the status of the margins of resection was misinterpreted in a number of cases. Just as the terminology used to describe the prostatic capsule is often open to interpretation, margins are also described using a variety of terms. Pathologists often use the terms extending to, involving and reaching the margin synonymously to denote a positive margin. In contrast, terms describing a tumor approaching or extending close to a margin implies that tumor goes near the margin but tumor is not cut across. Difficulties in assessing the distal margin of the prostate, as discussed in detail elsewhere, may also lead to ambiguities regarding the margins of resection. Articles in which much of the data concern the pathology of the radical prostatectomy specimen should include among the authors a pathologist who has reviewed the material and helped verify the accuracy of the data. One of the few studies to assess the relationship of capsular penetration to progression using serum PSA levels as a means of detecting recurrences is that by Stein et al from UCLA. 4 The 5 and 10-year disease-free progression rates for tumors with capsular penetration were 79% and 62%, respectively. In this study tumors that "penetrated into or through" the capsule were considered as showing pathological stage C disease. Adjuvant radiotherapy was administered to some of these men, although the authors claim that this did not affect their results. Gleason score 7 tumors were considered to be moderately differentiated and cases were not stratified by margins. In addition, the data were obtained by review of pathology reports without the direct input of a pathologist. Despite differing methodologies, all of these studies in conjunction with the current work suggest that not all prostate cancers with capsular penetration experience progression 8 to 10 years postoperatively. Another recent study to address the significance of capsular
140
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
penetration is that by Partin et al from our institution analyzing a different group of patients (table 3). 5 In that study, of which the lead author of the current paper is also a coauthor, particular attention was focused on tumors that invaded the region of the neurovascular bundle. 5 The lower incidence of positive margins found by Partin et al_ is attributable to_ their requiring established capsular penetration to be present m the region of the neurovascular bundle. This is the site where additional soft tissue may be removed to increase the likelihood of negative surgical margins. Tumors with positive margins progressed at a more rapid rate and with greater frequ_ency than in the current study, probably as a result of a higher proportion of extensive positive margins_. Only 5% ?~ the patients with established capsular penetration and positive margins had extensively positive margins in the current_study, as compared to 19% in the group studied by Partin et al. Another explanation is that some patients with positive margins in our study did not have serum PSA levels drawn for se~eral years following prostatectomy; these men may have experienced progression earlier that went undetected. A m?re surpri~ing result from the Partin study was that tumors with established capsular penetration and negative margins were as likely to progress 4 years postoperatively as those with positive margins. Several possible explanations for these differing results ~re possible. Whereas 53% of the patients in the current study_with negative margins had high grade tumor, 70% of the patients within the study by Partin et al had a Gleason score of 7 or above. In their study the 5-year progression-free probability was 75% for tumors with a Gleason score of 5 and 6 as compared to 25% with Gleason scores of 7 or above. Only including patients who had established capsular penetration in the region of the neurovascular bundle may also have introduced a bias to their study. Because a greater amount of soft tissue may be removed from the region of the neurovascular bundle than around any other site in the prostate gland, one can still have extensive tumor spread out of the prostate in this region while still obtaining negative margins. 15 Cases with negative margins and established capsular penetration at other sites often have less extracapsular tumor, since a greater amount of extraprostatic tumor in these regions invariably results in positive margins. Just as cases with established capsular penetration h_ave a worse prognosis than those with focal capsular penetration, cases with extensive established capsular penetration in the region of the neurovascular bundle may have a worse prognosis than tumors with lesser amounts of established capsular penetration occurring elsewhere in the gland, even in tumors with negative margins. Despite these differences, our study and that by Partin et al 5 demonstrate that patients with high grade tumors with established capsular penetration and positive margins fare poorly following radical prostatectomy. The differences between the 2 studies are predominantly in the more rapid and higher rate of progression seen in the Partin study. This disparity results more from differences in the popul~tions studied than in the study designs. In general, tumors m the Partin study were more aggressive and advanced, with higher 0
Gleason grades, more extensive extracapsular spread and more widespread involved margins. CONCLUSIONS
Our study evaluated a group of patients with capsular penetration who have been followed prospectively without adjuvant radiation or endocrine therapy. In our study the local recurrence rate, using a generous definition, was only 6.5%. However, it is unclear what per cent of patients with elevated serum PSA levels alone had occult local recurrence and what per cent harbored occult distant metastases. Also, it is unknown whether radiotherapy administered immediately postoperatively would have altered the results. Therefore, it is difficult to extrapolate from our data regarding the use of immediate postoperative adjuvant therapy. Nevertheless, certain statements can be made regarding this issue. The negligible influence of positive margins in patients with focal capsular penetration has important consequences in deciding whether to administer immediate postoperative adjuvant radiotherapy or hormonal therapy. In patients whose tumors display focal capsular penetration the status of the capsular margin should not influence the decision. Low grade tumors with focal capsular penetration, and low grade tumors with established capsular penetration and negative margins have a low risk of progression. In these 2 groups it is difficult to justify the use of adjuvant therapy. Between 30% and 50% of the patients with established capsular penetration in the high and intermediate risk groups have not had progression of disease 8 years following radical prostatectomy. However, in the intermediate and high risk groups of tumors with established capsular penetration the Kaplan-Meier curves have not reached a plateau. Because our curves are parallel at 8 years of followup, there is nothing to suggest that these curves will converge with even longer followup. However, what percentage of these patients will ultimately be cured or whether disease will continually progress must await additional followup. Tumors with these features might be the ideal population to evaluate the efficacy of immediate postoperative adjuvant therapy in a randomized trial. When designing clinical trials and evaluating the efficacy of adjuvant therapy following radical prostatectomy, tumors with capsular penetration should be stratified into groups having similar risks of progression according to the extent of capsular penetration, surgical margins and grade. REFERENCES
1. McNeal, J. E., Villers, A. A., Redwine, E. A., Freiha, F. S. and
2. 3. 4.
TABLE 3.
Inclusion criteria
Current Study Clinical stage Bone scan Minimum followup if no progression (yrs.) Specimen submitted Capsular penetration
Seminal vesicles Periseminal vesicles Lymph nodes
B Neg. 2 Totally Established or focal, any site
Neg. Neg. Neg.
Partin et al 5
B Neg.
5.
1
Subtotally by protocol Established in area of neurovascular bundle with or without established capsular penetration elsewhere Pas. or neg. Pas. or neg. Neg.
6.
7. 8.
Stamey, T. A.: Capsular penetration in prostate cancer: significance for natural history and treatment. Amer. J. Surg. Path., 14: 240, 1990. Schellhammer, P. F.: Radical prostatectomy: patterns of local failure and survival in 67 patients. Urology, 31: 191, 1988. Paulson, D. F., Moul, J. W. and Walther, P. J.: Radical prostatectomy for clinical stage Tl-2NOMO prostatic adenocarcinoma: long-term results. J. Urol., 144: 1180, 1990. Stein, A., deKernion, J. B., Smith, R. B., Dorey, F. and Patel, H.: Prostate specific antigen levels after radical prostatectomy in patients with organ confined and locally extensive prostate cancer. J. Urol., 147: 942, 1992. Partin, A. W., Borland, R. N., Epstein, J. I. and Brendler, C. B.: Influence of wide excision of the neurovascular bundle(s) on prognosis in men with clinically localized prostate cancer with established capsular penetration. J. Urol., 150: 142, 1993. Epstein, J. I., CarMichael, M .. and Walsh, P. C.: Adenocarcinoma of the prostate invading the seminal vesicle: definition and relation of tumor volume, grade and margins of resection to prognosis. J. Urol., 149: 1040, 1993. Walsh, P. C., Lepor, H. and Eggleston, J.C.: Radical prostatectomy with preservation of sexual function: anatomical and pathological considerations. Prostate, 4: 473, 1983. Walsh, P. C.: Technique of radical retropubic prostatectomy with preservation of sexual function-an anatomic approach. In: Di-
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
agnosis and Management of Genitourinary Cancer. Edited by D. G. Skinner and G. Leiskovsky, Philadelphia: W. B. Saunders Co., chapt. 53, pp. 753-778, 1988. 9. Epstein, J. I., Pizov, G. and Walsh, P. C.: Correlation of pathologic findings with progression following radical retropubic prostatectomy. Cancer, in press. 10. Ayala, A. G., Ro, J. Y., Babaian, R., Troncoso, P. and Grignon, D. J.: The prostatic capsule: does it exist? Its importance in the staging and treatment of prostatic carcinoma. Amer. J. Surg. Path., 13: 21, 1989. 11. Epstein, J. I.: The evaluation of radical prostatectomy specimens: therapeutic and prognostic implications. Path. Ann., 26: 159, 1991.
141
12. Stamey, T. A., McNeal, J. E., Freiha, F. S. and Redwine, E.: Morphometric and clinical studies on 68 consecutive radical prostatectomies. J. Urol., 139: 1235, 1988. 13. Epstein, J. I.: Evaluation of radical prostatectomy capsular margins of resection: the significance of margins designated as negative, closely approaching, and positive. Amer. J. Surg. Path., 14: 626, 1990. 14. Gleason, D. F.: Histologic grading of prostate cancer: a perspective. Hum. Path., 23: 273, 1992. 15. Walsh, P. C., Epstein, J. I. and Lowe, F. C.: Potency following radical prostatectomy with wide unilateral excision of the neurovascular bundle. J. Urol., 138: 823, 1987.
THE JOURNAL OF UROLOGY
Printed in
Copyright© 1993 by AMERICAN UROLOGICAL ASSOCIATION, INC.
INFLUENCE OF CAPSULAR PENETRATION ON PROGRESSION FOLLOWING RADICAL PROSTATECTOMY: A STUDY OF 196 CASES WITH LONG-TERM FOLLOWUP JONATHAN I. EPSTEIN,* MARNE J. CARMICHAEL, GALINA PIZOVt
AND
PATRICK C. WALSH
From the Departments of Pathology and Urology, The Johns Hopkins University School of Medicine and the James Buchanen Brady Urological Institute, The Johns Hopkins Hospital, Baltimore, Maryland
ABSTRACT
We studied 196 radical prostatectomy cases performed for clinical stage B prostate cancer with capsular penetration; in all cases seminal vesicles and lymph nodes were free of tumor. The mean followup in patients who showed no evidence of progression was 5 years. Focal capsular penetration was seen in 93 cases. There was no difference in progression in this group, irrespective of whether margins were negative or positive. High grade tumors (Gleason score 7 or more) had a significantly higher risk of progression compared to lower grade tumors (p = 0.0002). Established capsular penetration was seen in 103 tumors. Cancers with established capsular penetration had a higher risk of progression than those with focal capsular penetration. Established capsular penetration tumors were stratified into 3 groups with increasing risks of progression: 1) margins were negative and grade was low, 2) margins were positive or grade was high yet both adverse features were not present or 3) margins were positive and grade was high. The differences in progression among these 3 groups were statistically significant. Because of the negligible influence of positive margins in patients with focal capsular penetration the status of capsular margins should not influence the decision on whether to administer immediate postoperative adjuvant therapy. To evaluate the efficacy of adjuvant therapy following radical prostatectomy, tumors with capsular penetration should be stratified into groups having similar risks of progression according to the extent of capsular penetration, surgical margins of resection and grade. KEY WORDS:
prostatic neoplasms, neoplasm invasiveness, prostatectomy
An often repeated truism in urology is that once prostate cancer has spread outside the gland the window of opportunity to cure the patient is lost. Studies have demonstrated that capsular penetration is related to other variables associated with aggressive tumor behavior, such as increased tumor volume, seminal vesicle invasion, lymph node metastases and poor histological grade. 1 However, only a few studies have assessed the relationship between capsular penetration and recurrence following radical prostatectomy. 2- 5 Prior studies evaluating the prognostic significance of capsular penetration have suffered from the following deficiencies: 1) prostates were not serially sectioned and embedded in their entirety, 2) radical prostatectomy specimens were not stratified by capsular margins of resection, 3) Gleason score 7 tumor was considered as intermediate grade tumor rather than high grade tumor, 4) cases when tumor extended into but not through the capsule were considered as showing capsular penetration, 5) extent of capsular penetration was not quantified, 6) pathological information was gleaned from pathology reports in which ambiguous terminology may have led to erroneous interpretation and 7) followup information antedated the use of serum prostate specific antigen (PSA) levels to detect recurrence. The current study evaluated 196 radical prostatectomy specimens with pathological stage C disease in which the seminal vesicles and pelvic lymph nodes were free of tumor. Cases were stratified according to the extent of capsular penetration, status of capsular margins and Gleason score. These parameters were correlated with progression. MATERIALS AND METHODS
Between April 1982 and the end of 1988, 600 consecutive men underwent retropubic prostatectomy by 1 surgeon (P. C. Accepted for publication December 18, 1992.
* Requests for reprints: Department of Pathology, The Johns Hopkins Hospital, Baltimore, Maryland 21205. t Current address: Hadassah University Hospital, Jerusalem, Israel.
W.) at our hospital. Preoperatively, all patients were evaluated by careful palpation of the primary lesion, serum enzymatic prostatic acid phosphatase (PAP), bone scan, pathological review of the biopsy material and, since June 1986, PSA levels. From the 600 men 196 cases satisfied the following requirements and were included in this study: 1) the patients had clinical stage B disease, 2) all patients whose disease did not progress had a minimum of 2 years of follovvup, 3) all men had at least 1 serum PSA determination postoperatively and had a PSA value at the last visit, 4) all prostates were serially sectioned, embedded in their entirety and assessable as to margins of resection, 5) all tumors demonstrated penetration through the capsule into periprostatic soft tissue and 6) in all cases seminal vesicles and lymph nodes were free of tumor. A total of 15 cases in which tumor invaded the periseminal vesicle soft tissue was also dropped from the analysis, since a prior study showed these tumors to have an intermediate prognosis between those showing capsular penetration alone and those showing seminal vesicle invasion. 6 The surgical procedure on the first 100 patients was performed as described in 1983. 7 Thereafter, once the anatomy of the pelvic fascia and neurovascular bundles was better characterized it was possible to excise the neurovascular bundle(s) widely when indicated. All subsequent procedures were performed via the technique presently in use. 8 Only 2 patients received immediate adjuvant radiotherapy to the prostatic bed and both of these men eventually experienced progression. Otherwise, no patient received adjuvant radiotherapy or hormonal therapy. All men were evaluated at 3 and 6 months postoperatively. Patients with tumor involving the surgical margins were followed semiannually, while all others were evaluated yearly. Followup examination included a complete history, rectal examination, serum acid phosphatase and, since June 1986, serum PSA. Because the patients came from 47 of
135
136
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
the 50 states in the United States it was not possible to evaluate all men at our hospital yearly. However, in those men followup information from their referring urologist was obtained by mail yearly. PSA was considered elevated if it was greater than the lower limit of detection for the reference laboratory; at our hospital using the Tandem-R PSA* solid phase 2-site immunoradiometric assay the lower limit of detection is 0.2 ng./ml. Patients with symptoms of recurrent disease or elevated serum markers were evaluated with a complete history, a repeat careful rectal examination, chest x-ray, bone scan and computerized tomography of the pelvis. Local recurrence was defined as palpable induration at the operative site; in many cases if this induration was associated with an elevated PSA level the men were classified as having local recurrence without biopsy confirmation. Distant metastases were confirmed by bone scan. Following resection the intact prostate was weighed, coated over the entire surface with india ink and fixed in 10% buffered formalin for 18 to 24 hours. Following fixation the distal and proximal thin shave urethral margins were removed for histological examination. These shave margins were taken as cross sections, with the urethra running through the center. The gland was then serially sectioned at 2 to 3 mm. intervals perpendicular to the long (apical-basal) axis of the gland. Each cross section of the prostate was further divided (that is, halves, quarters and so forth) to conform to the size of a cassette. The entire prostate was embedded and designated in a manner permitting the localization of each section within the prostate. The tumor was graded using the Gleason grading system. Capsular penetration was defined as tumor extending out of the prostate into periprostatic soft tissue. Capsular penetration was subdivided into those in which only a few neoplastic glands were present exterior to the prostate (focal capsular penetration, fig. 1, A) and cases with more extensive extraprostatic spread of tumor (established capsular penetration, fig. 1, B). Cases with focal capsular penetration revealed on only 1 or 2 slides a few cancer glands outside of the prostate, which may not have been identified had we not totally submitted the prostate for histological analysis. Capsular margins of resection were designated as negative or positive, as previously described.9 Statistics were performed using the statistics graphic data measurement software program. t End points measured as interval to progression were calculated by the Kaplan-Meier method for estimation. Differences between Kaplan-Meier curves were tested for statistical significance using the Wilcoxon-Gehan test. RESULTS
Focal capsular penetration. There were 93 cases showing focal capsular penetration: 77 men (82%) showed no evidence of progression with a mean followup of 5 years and 16 men experienced progression. The overall progression-free probability curve for tumors with focal capsular penetration is illustrated in figure 2. Margins were free of tumor in 55 of these specimens and in 38 cases the margins were focally positive. There was no difference in progression, irrespective of whether margins were negative or positive (fig. 3, A). The mean Gleason grade for tumors with focal capsular penetration was 6.1. High grade tumors had a significantly increased risk of progression compared to tumors with a Gleason score of less than 7 (p = 0.0002, fig. 3, B). The interval to failure stratified by type of recurrence is shown in table 1. There were 5 patients with focal capsular penetration who did not have a serum PSA level drawn until 2 to 3 years following radical prostatectomy; progression was noted at that time. Given that the curves for tumors with positive and negative margins are superimposed, and the marked difference in progression when tumors were stratified
* Hybritech Inc., San Diego, California.
t STATA Computing Resource, Los Angeles, California.
FIG. 1. A, several neoplastic glands immediately exterior to prostate show perineural invasion in periprostatic tissue (right side), consistent with focal capsular penetration. Reduced from X25. B, established capsular penetration resulting in irregular protuberance to capsular surface because of tumor-associated desmoplasia. Reduced from Xl8.
by grade, our results would not have been altered significantly whether these few patients had progression 1 or 2 years earlier. Established capsular penetration. A total of 103 tumors demonstrated established capsular penetration. The mean followup for patients showing no evidence of progression was 5 years, while 36 men (35%) experienced progression. The actuarial risk of progression in tumors with established capsular penetration is illustrated in figure 2. The difference in progression in cancers with established capsular penetration compared to those with focal capsular penetration just missed statistical significance (p = 0.06). Of the cases 29 had negative margins, 70 tumors had focally positive margins and 4 showed extensively positive margins. Tumors removed with negative margins had a lower likelihood of progressing (p = 0.006, fig. 4, A). The mean Gleason score for tumor with established capsular penetration was 6. 7. The difference in progression when tumors with established capsular penetration were stratified by Gleason score barely missed statistical significance (p = 0.07, fig. 4, B). Because low grade tumors (Gleason score less than 7) with positive margins had similar progression rates as high grade (Gleason score 7 or more) tumors with negative margins, these 2 groups were combined. The resulting 3 groups had significantly different progression probabilities: low risk-negative margins and low grade, intermediate risk-negative margins and high grade tumor or positive margins and low grade tumor, and high risk-positive margins and high grade (fig. 5). The
137
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
Correlation of Focal and Established Capsular Penetration with Progression '.l (7) >, _µ
(18)
l:'.:(:==9l=c::-:--u__
..... ..... .....
_j__~c6)
__j__
(4)
.75 -
D
ro
focal cap pen.
D 0
L CL
"'"'
.5 -
....L
(3)
I
C
established
0
. .-,
in
cap. pen.
.25 -
I!>
L
Ol
D
L
a.. 0 I
I
I
.4
0
8
first year of recurrence FIG. 2. Correlation of extent of capsular penetration (cap. pen.) with progression. Numbers in parentheses represent number of censored patients.
A
Focal Capsular Penetration: Correlation of Surgical ~arg1ns with Progression 1 -
>
-
(12)
+-'
C1 ll
pas. margin
(5)
rl
(4)
rl
D
ro
D
(7)
(4)
(9)
B
13)
(2)
11 1 _ ,----i1c_~ ( 7 )
(2)
(3)
.75 -
~
ro
D 0
75 -
(6)
Cll
C1l
L
CL
.5 -
Gleason ~7
0)
I
0)
£:
s:
5 -
I
If) If)
L
(31 Gleason
D
C
QJ
(5) (5)
rl
0) 0)
L
(12)
rl
margin
Q_
(10)
~
neg.
L
(11)
>-
(2)
0
Focal Capsular Penetration: Correlation of Gleason Sum with Progression
C 0 rl If)
.25 -
01 0 L
Ul Q)
25 -
L
01
Q_
0 (_
0 -
Q_
I 0
I
I
I
2
3
4
I 5
I
I
I
I
5
7
8
9
0 I
I
I
0
3
5
first year of recurrence FIG,
I
8
I
g
first year of recurrence
3. Focal capsular penetration. A, correlation of surgical margins with progression. B, correlation of Gleason sum with progression
curve for low grade tumors with negative margins decreases to a progression-free probability of zero at year 8. However, this decrease represents only 1 patient followed to 8 years who had progression. The difference in progression between the low and intermediate risk tumors was p = 0.05. High risk tumors had a higher likelihood of progression as compared to the intermediate risk tumors (p = 0.04). The difference in progression between patients with low and high risk tumors was highly significant (p = 0.004). The interval to progression in these different groups, classified by the type of recurrence, is depicted in table 2. In 14 patients the first postoperative serum PSA levels were drawn 2 or more years following radical prostatectomy, with progression noted at that time. All 14 of these cases had positive margins and 12 of these tumors were high grade. Since these men may have had progression at even an earlier date had serum PSA levels been available, the differences in tumors stratified by margins and grade may be more significant than what was calculated using the information available. Of the 14 men with delayed measurements of serum PSA 12 had high grade tumor with positive margins. Furthermore, all 3 patients
with PSA levels drawn 4 to 5 years following radical prostatecTABLE 1.
Type of Failure
Focal capsular penetration: interval to progression No. Cases
Yrs. to Failure
Gleason score less than 7 (59 pts.)
PSA Local Distant
4 0
4 (mean), 2-8 (range)*
0
Gleason score 7 or more (34 pts.)
PSA Local Local Local Distant
8 1 1 1 1
3 (mean), 2-5 (range)t PSA at 2, local at 3+ PSA at 2, local at 4 Local at 3 Distant at 2
* Delayed measurement of PSA in 2 men (2 and 3 years postoperatively) with abnormal PSA noted at first measurement. t Delayed measurement of PSA in 2 men (2 and 2 years postoperatively) with abnormal PSA noted at first measurement. Delayed measurement of PSA with abnormal PSA noted at first measurement.
+
138
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
A
B
Established Capsular Penetration: Correlation of Surgical Margins with Progression (1)
1-
1 -----=====ic_
_c_~_'19)
(3)
?;'
>,
~
(6)
(4)
rl rl
rl
:.8 ro
D
negative
.75 -
0
L
(_
Cl..
Cl..
.;:
UJ
Q) (_
.5 -
I C
positive
.s
ClJ L 01 0 L
-s
.5 -
I C
Cl) Cl)
.75 -
"'
D 0
margin
D
Q) Q)
Established Capsular Penetration: Correlation of Gleason Sum with Progression
(2)
margin
en
Cl) Q)
.25 -
Gleason ~7 .25 -
L 01 0 L
D..
Cl..
0 -
0 I
0
I
2
I
3
I
I
4
first year of recurrence
6
I
I
7
B
I
I
4
0
I
5
I
6
I
B
first year of recurrence
FIG. 4. Established capsular penetration. A, correlation of surgical margins with progression. B, correlation of Gleason sum with progression
Established Capsular Penetration: Correlation of Surgical Margins and Gleason Sum wi tn Progression (3)
1-
(4)
(2)
(2)
neg. margins, Gleason
>, .µ
..... •rl •rl
.0
ro
.75 -
.0 D L CL
w w L
pos. margins. Gleason
C3)( 2) .5 -
't-
i C 0
..... "'w (/)
L
pos. margins, Gleason~?
.25 -
CJ)
0
L
CL
0 I
0
I
2
I
3
I
4
I
5
I
6
first year of recurrence
I
7
I
8
FIG. 5. Established capsular penetration. Correlation of surgical margins and Gleason sum with progression. neg., negative. pos., positive
tomy had high grade tumors with positive margins. It is likely that at least some of these men would have been shown to have had progression at an earlier date had serum PSA levels been available immediately following radical prostatectomy. Therefore, tumors in the high risk group with positive margins and high grade tumors are probably associated with even a worse risk of progression as compared to the intermediate and lower risk groups than we were able to calculate using the available data. Also, because measurements of PSA were more often delayed in tumors with established penetration compared to cases with focal penetration, tumors with established penetration probably have a relatively worse progression rate than illustrated in figure 2. DISCUSSION
Despite the use of terminology such as capsular penetration, the prostate lacks a well defined capsule. 10 • 11 In some areas there appears to be a separation between the most peripheral nonneoplastic acini and the edge of the prostate, giving the appearance of a fibromuscular capsule. However, elsewhere nonneoplastic glands extend right up to the edge of the prostate. Towards the base of the prostate the edge of the gland is even more ill-defined with nonneoplastic glands appearing to extend
out of the prostate. Anteriorly, the edge of the gland is also poorly delimited where the anterior fibromuscular stroma of the prostate merges in with smooth muscle that extends to the pubic bone. Because one cannot distinguish where the prostatic capsule begins and the normal prostatic stroma ends, it makes no sense to classify tumors as going into but not through the prostatic capsule. Furthermore, the majority of studies have demonstrated that prognosis is adversely affected only when tumor penetrates full thickness through the prostatic capsule and extends into adjacent periprostatic soft tissue. 12 Consequently, tumors should be categorized as either being confined to the gland or extending out of the prostate. Since some physicians differentiate between capsular penetration and capsular perforation, pathologists should define their use of these terms until nomenclature is standardized. Recognizing that the prostate lacks a true discrete fibrous or fibromuscular capsule, the term prostatic capsule, nevertheless, is still convenient to denote the edge of the gland. Only a few studies have measured the. extent of capsular penetration. Stamey et al quantified capsular penetration based on the length of capsular penetration summed along different slides. 12 However, because prostate cancer often spreads along nerves parallel to the prostatic capsule one can see tumor barely
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY TABLE 2.
Established capsular penetration: interval to progression
Type of Failure
No. Cases
Yrs. to Failure
Gleason score 7 or more and neg. margins (13 pts.) PSA 1 8 Local 0 Distant 0 Gleason score less than 7 and pas. margins or Gleason score 7 or more and neg. margins (41 pts.)
PSA Local Local Local Distant
8 1 1 1
4 (mean), 2-7 (range)* PSA at 1, local at 2 Local at 4 Local at 6
0
Gleason score 7 or more and pas. margins (49 pts.)
PSA Local Local Local Local Local Distant Distant Distant Distant
12 1 2 1 2 1 2 1 1 1
4 (mean), 2-6 (range)t PSA at 3, local at 4+ Local at 4+ PSA at 3, local at 5+ Local at 5 PSA at 7, local at 8 PSA at 2, distant at 3 PSA at 2, distant at 3+ Distant at 3+ PSA at 2, local, distant at 4+
* Delayed measurement of PSA in 2 men (2 and 3 years postoperatively) with abnormal PSA noted at time of first measurement. t Delayed measurement of PSA in 5 men (2, 2, 3, 4 and 5 years postoperatively) with abnormal PSA noted at time of first measurement. Delayed measurement of PSA with abnormal PSA noted at time of first measurement.
+
out of the gland skimming along for significant distances. Also, cases with extensive spread of tumor perpendicularly to the prostatic capsule into the neurovascular bundle might be misdiagnosed as showing minimal capsular penetration if there is not significant horizontal spread of tumor as well. At our institution we have classified cases of capsular penetration as showing focal capsular penetration when only a few neoplastic glands are present in periprostatic soft tissue. 9 These glands are barely exterior to the prostate and tend to grow horizontally parallel to the prostate rather than extending away from the prostatic gland. Tumors with a greater degree of extracapsular spread are designated as having established capsular penetration. Using this definition, our study demonstrates that tumors with established capsular penetration had a higher risk of progression than those showing only focal capsular penetration. The only parameter that helped stratify tumors with focal capsular penetration was the Gleason score. A somewhat surprising finding was the virtually identical progression rate in cancers with focal capsular penetration regardless of whether capsular margins were positive or negative. There are 2 potential explanations for this disparity. First, many of the margins designated as positive in patients with focal capsular penetration represent an artifactually positive margin. In general, there exists in prostate cancer a great potential for over-diagnosing positive margins as a consequence of the limited, easily disruptable soft tissue that surrounds radical prostatectomy specimens.13 In prostate cancers when there is only minimal tumor spread out of the prostate and there is a positive margin, our study suggests that there is even a greater likelihood that these positive margins are artifactual. The other explanation is that tumors displaying only focal capsular penetration are inherently less aggressive, and even if tumor is cut across it has little potential for regrowth and progression. Progression in cases of established penetration was adversely influenced by positive margins and high grade tumor. We were able to stratify tumors into 3 groups associated with increasing risks of progression based on whether 1) margins were negative and grade was low, 2) margins were either positive or grade was
139
high and yet both adverse features were not present, or 3) margins were positive and grade was high. It is difficult to compare the current findings with those of prior studies. Paulson et al compared 74 tumors that had established capsular penetration and negative margins (specimen confined) to 123 tumors that had established capsular penetration and positive margins (margin positive). 3At 10 years postoperatively the disease-free probabilities in these 2 groups were 70% and 40%, respectively. In their study progression was determined by an elevated PAP level. Other differences between our study and theirs were: in their study the inclusion of cases with positive seminal vesicles and stage A disease, a significant number of patients received postoperative radiotherapy although this was not found to influence failure and cases were subdivided into Gleason scores of 7 or less versus a Gleason score of more than 7. We and others have demonstrated that Gleason score 7 tumor behaves more closely to high grade tumor than intermediate grade tumor. 9• 14 Schellhammer analyzed a group of patients with pathological stage C disease, including cases with positive seminal vesicles, in whom recurrence was determined by biopsy. 2 There was no evidence of disease postoperatively in 77% and 54% of the men at 5 and 10 years, respectively. Data from his study were obtained by review of pathology reports rather than by direct interaction with pathologists. As he quotes, "Terminology used to describe involvement of the capsule is also imprecise. The terms involvement, invasion, permeation, perforation, or penetration are often used interchangeably and do not identify specifically whether the tumor involved the inner most layer of the fibrous capsule or extends through the capsule to the pericapsular tissue." This statement reflects that the terminology used to describe the relationship of tumor to the capsule is often ambiguous, and attempts to glean data from pathology reports regarding the status of the capsule are fraught with inaccuracies. The importance of having direct involvement by a study pathologist was recently demonstrated at our institution; when urologists initially reviewed pathology reports generated by different pathologists the status of the margins of resection was misinterpreted in a number of cases. Just as the terminology used to describe the prostatic capsule is often open to interpretation, margins are also described using a variety of terms. Pathologists often use the terms extending to, involving and reaching the margin synonymously to denote a positive margin. In contrast, terms describing a tumor approaching or extending close to a margin implies that tumor goes near the margin but tumor is not cut across. Difficulties in assessing the distal margin of the prostate, as discussed in detail elsewhere, may also lead to ambiguities regarding the margins of resection. Articles in which much of the data concern the pathology of the radical prostatectomy specimen should include among the authors a pathologist who has reviewed the material and helped verify the accuracy of the data. One of the few studies to assess the relationship of capsular penetration to progression using serum PSA levels as a means of detecting recurrences is that by Stein et al from UCLA. 4 The 5 and 10-year disease-free progression rates for tumors with capsular penetration were 79% and 62%, respectively. In this study tumors that "penetrated into or through" the capsule were considered as showing pathological stage C disease. Adjuvant radiotherapy was administered to some of these men, although the authors claim that this did not affect their results. Gleason score 7 tumors were considered to be moderately differentiated and cases were not stratified by margins. In addition, the data were obtained by review of pathology reports without the direct input of a pathologist. Despite differing methodologies, all of these studies in conjunction with the current work suggest that not all prostate cancers with capsular penetration experience progression 8 to 10 years postoperatively. Another recent study to address the significance of capsular
140
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
penetration is that by Partin et al from our institution analyzing a different group of patients (table 3). 5 In that study, of which the lead author of the current paper is also a coauthor, particular attention was focused on tumors that invaded the region of the neurovascular bundle. 5 The lower incidence of positive margins found by Partin et al_ is attributable to_ their requiring established capsular penetration to be present m the region of the neurovascular bundle. This is the site where additional soft tissue may be removed to increase the likelihood of negative surgical margins. Tumors with positive margins progressed at a more rapid rate and with greater frequ_ency than in the current study, probably as a result of a higher proportion of extensive positive margins_. Only 5% ?~ the patients with established capsular penetration and positive margins had extensively positive margins in the current_study, as compared to 19% in the group studied by Partin et al. Another explanation is that some patients with positive margins in our study did not have serum PSA levels drawn for se~eral years following prostatectomy; these men may have experienced progression earlier that went undetected. A m?re surpri~ing result from the Partin study was that tumors with established capsular penetration and negative margins were as likely to progress 4 years postoperatively as those with positive margins. Several possible explanations for these differing results ~re possible. Whereas 53% of the patients in the current study_with negative margins had high grade tumor, 70% of the patients within the study by Partin et al had a Gleason score of 7 or above. In their study the 5-year progression-free probability was 75% for tumors with a Gleason score of 5 and 6 as compared to 25% with Gleason scores of 7 or above. Only including patients who had established capsular penetration in the region of the neurovascular bundle may also have introduced a bias to their study. Because a greater amount of soft tissue may be removed from the region of the neurovascular bundle than around any other site in the prostate gland, one can still have extensive tumor spread out of the prostate in this region while still obtaining negative margins. 15 Cases with negative margins and established capsular penetration at other sites often have less extracapsular tumor, since a greater amount of extraprostatic tumor in these regions invariably results in positive margins. Just as cases with established capsular penetration h_ave a worse prognosis than those with focal capsular penetration, cases with extensive established capsular penetration in the region of the neurovascular bundle may have a worse prognosis than tumors with lesser amounts of established capsular penetration occurring elsewhere in the gland, even in tumors with negative margins. Despite these differences, our study and that by Partin et al 5 demonstrate that patients with high grade tumors with established capsular penetration and positive margins fare poorly following radical prostatectomy. The differences between the 2 studies are predominantly in the more rapid and higher rate of progression seen in the Partin study. This disparity results more from differences in the popul~tions studied than in the study designs. In general, tumors m the Partin study were more aggressive and advanced, with higher 0
Gleason grades, more extensive extracapsular spread and more widespread involved margins. CONCLUSIONS
Our study evaluated a group of patients with capsular penetration who have been followed prospectively without adjuvant radiation or endocrine therapy. In our study the local recurrence rate, using a generous definition, was only 6.5%. However, it is unclear what per cent of patients with elevated serum PSA levels alone had occult local recurrence and what per cent harbored occult distant metastases. Also, it is unknown whether radiotherapy administered immediately postoperatively would have altered the results. Therefore, it is difficult to extrapolate from our data regarding the use of immediate postoperative adjuvant therapy. Nevertheless, certain statements can be made regarding this issue. The negligible influence of positive margins in patients with focal capsular penetration has important consequences in deciding whether to administer immediate postoperative adjuvant radiotherapy or hormonal therapy. In patients whose tumors display focal capsular penetration the status of the capsular margin should not influence the decision. Low grade tumors with focal capsular penetration, and low grade tumors with established capsular penetration and negative margins have a low risk of progression. In these 2 groups it is difficult to justify the use of adjuvant therapy. Between 30% and 50% of the patients with established capsular penetration in the high and intermediate risk groups have not had progression of disease 8 years following radical prostatectomy. However, in the intermediate and high risk groups of tumors with established capsular penetration the Kaplan-Meier curves have not reached a plateau. Because our curves are parallel at 8 years of followup, there is nothing to suggest that these curves will converge with even longer followup. However, what percentage of these patients will ultimately be cured or whether disease will continually progress must await additional followup. Tumors with these features might be the ideal population to evaluate the efficacy of immediate postoperative adjuvant therapy in a randomized trial. When designing clinical trials and evaluating the efficacy of adjuvant therapy following radical prostatectomy, tumors with capsular penetration should be stratified into groups having similar risks of progression according to the extent of capsular penetration, surgical margins and grade. REFERENCES
1. McNeal, J. E., Villers, A. A., Redwine, E. A., Freiha, F. S. and
2. 3. 4.
TABLE 3.
Inclusion criteria
Current Study Clinical stage Bone scan Minimum followup if no progression (yrs.) Specimen submitted Capsular penetration
Seminal vesicles Periseminal vesicles Lymph nodes
B Neg. 2 Totally Established or focal, any site
Neg. Neg. Neg.
Partin et al 5
B Neg.
5.
1
Subtotally by protocol Established in area of neurovascular bundle with or without established capsular penetration elsewhere Pas. or neg. Pas. or neg. Neg.
6.
7. 8.
Stamey, T. A.: Capsular penetration in prostate cancer: significance for natural history and treatment. Amer. J. Surg. Path., 14: 240, 1990. Schellhammer, P. F.: Radical prostatectomy: patterns of local failure and survival in 67 patients. Urology, 31: 191, 1988. Paulson, D. F., Moul, J. W. and Walther, P. J.: Radical prostatectomy for clinical stage Tl-2NOMO prostatic adenocarcinoma: long-term results. J. Urol., 144: 1180, 1990. Stein, A., deKernion, J. B., Smith, R. B., Dorey, F. and Patel, H.: Prostate specific antigen levels after radical prostatectomy in patients with organ confined and locally extensive prostate cancer. J. Urol., 147: 942, 1992. Partin, A. W., Borland, R. N., Epstein, J. I. and Brendler, C. B.: Influence of wide excision of the neurovascular bundle(s) on prognosis in men with clinically localized prostate cancer with established capsular penetration. J. Urol., 150: 142, 1993. Epstein, J. I., CarMichael, M .. and Walsh, P. C.: Adenocarcinoma of the prostate invading the seminal vesicle: definition and relation of tumor volume, grade and margins of resection to prognosis. J. Urol., 149: 1040, 1993. Walsh, P. C., Lepor, H. and Eggleston, J.C.: Radical prostatectomy with preservation of sexual function: anatomical and pathological considerations. Prostate, 4: 473, 1983. Walsh, P. C.: Technique of radical retropubic prostatectomy with preservation of sexual function-an anatomic approach. In: Di-
CAPSULAR PENETRATION AND PROGRESSION AFTER RADICAL PROSTATECTOMY
agnosis and Management of Genitourinary Cancer. Edited by D. G. Skinner and G. Leiskovsky, Philadelphia: W. B. Saunders Co., chapt. 53, pp. 753-778, 1988. 9. Epstein, J. I., Pizov, G. and Walsh, P. C.: Correlation of pathologic findings with progression following radical retropubic prostatectomy. Cancer, in press. 10. Ayala, A. G., Ro, J. Y., Babaian, R., Troncoso, P. and Grignon, D. J.: The prostatic capsule: does it exist? Its importance in the staging and treatment of prostatic carcinoma. Amer. J. Surg. Path., 13: 21, 1989. 11. Epstein, J. I.: The evaluation of radical prostatectomy specimens: therapeutic and prognostic implications. Path. Ann., 26: 159, 1991.
141
12. Stamey, T. A., McNeal, J. E., Freiha, F. S. and Redwine, E.: Morphometric and clinical studies on 68 consecutive radical prostatectomies. J. Urol., 139: 1235, 1988. 13. Epstein, J. I.: Evaluation of radical prostatectomy capsular margins of resection: the significance of margins designated as negative, closely approaching, and positive. Amer. J. Surg. Path., 14: 626, 1990. 14. Gleason, D. F.: Histologic grading of prostate cancer: a perspective. Hum. Path., 23: 273, 1992. 15. Walsh, P. C., Epstein, J. I. and Lowe, F. C.: Potency following radical prostatectomy with wide unilateral excision of the neurovascular bundle. J. Urol., 138: 823, 1987.