- Email: [email protected]
Surgical Treatment of Locally Advanced (T3) Prostatic Carcinoma: Early Results
0022-534 7/87 /1384-0816$02.00/0
THE
Vol. 138, October
JOURNAL OF UROLOGY
Copyright© 1987 by The Williams & Wilkins Co.
Printed in U.S.A.
SURGICAL TREATMENT OF LOCALLY ADVANCED (T3) PROSTATIC CARCINOMA: EARLY RESULTS RUUD J. L. H. BOSCH, KARL H. KURTH
AND
FRITZ H. SCHROEDER
From the Department of Urology, Erasmus University Rotterdam, Rotterdam, The Netherlands
ABSTRACT
The fate of 48 patients with clinical stage T3 prostatic carcinoma after attempted curative surgical management was studied. In 23 of these patients positive frozen sections of the lymph nodes were found at pelvic lymphadenectomy and orchiectomy was performed. The median interval to progression was 61 months. Radical prostatectomy was performed in the remaining 25 patients. In 4 of these patients positive lymph nodes were found on paraffin sections but no additional treatment was given. Over-all, total tumor removal as defined by negative lymph nodes and negative margins of resection could be achieved in 14 of the 48 patients (29 per cent). During the same period 34 patients with clinical stage T<3 prostatic carcinoma were treated in a similar manner. Orchiectomy was done in 4 patients because of positive frozen sections of the lymph nodes and radical prostatectomy was done in 30, including 1 in whom positive paraffin sections of the lymph nodes were found but no additional treatment was given. An attempt was made to study the impact of several prognostic factors by comparing the probability of progression between patients with stage pT3 disease with (T3pT3NO) or without (T<3pT3NO) extracapsular tumor growth as determined by preoperative rectal examination (36 versus 27 per cent progression at 3 years), with or without positive margins of resection (45 versus 20 per cent progression at 3 years) and with or without involvement of the seminal vesicles (47 versus 18 per cent progression at 3 years). Our results suggest that a certain proportion of patients with clinical stage T3 disease will benefit from radical prostatectomy. This is to be expected especially in patients with stage T3pT3NO cancer and negative margins. (J. Ural., 138: 816-822, . 1987) The results of radical prostatectomy in patients with pathological stage T3 prostatic carcinoma, mostly with unknown lymph node status, have been reported by several investigators.1-4 In most of the series on stage pT3 disease treated by radical prostatectomy the inclusion of such patients resulted from clinical understaging. Patients believed to have an intracapsular tumor at clinical examination actually may have extraprostatic extension in up to 68 per cent of the cases. 5 Local overstaging of clinical T3 cancer is equally well known in up to 50 per cent of the cases. 6 In these patients abstinence from surgical treatment would result if clinical stage T3 prostatic carcinoma is not considered an indication for radical prostatectomy and a large group of curable patients would lose a chance for cure. Fair results have been obtained in patients with stage pT3 disease treated by radical prostatectomy as far as survival and especially local control are concerned. 1-4 We evaluated the possibility of surgical management of clinical stage T3 prostatic carcinoma. The fate of patients with surgically staged clinical T3 prostatic carcinoma is studied. Patients with T3pN + disease were treated by early orchiectomy. The median interval to progression in these patients is reported. Stage pNO cancer was treated by radical prostatectomy. The impact of high clinical stage was studied by comparing the probability of progression in cases of T3pT3NO and T<3pT3NO disease. Furthermore, we attempted to study the impact on progression of positive margins of resection in stage pT3 cancer treated by radical prostatectomy. Also, the impact on progression of seminal vesicle invasion was studied by comparing the probability of progression between patients with or without seminal vesicle invasion. Our results suggest that at least patients with T3pT3NO cancer and negative margins may benefit from radical prostatectomy. Unfortunately, it is difficult to identify this group preoperatively. Accepted for publication January 23, 1987.
MATERIALS AND METHODS
From 1977 to 1985, 82 prostatic carcinoma patients considered to be potential candidates for radical prostatectomy underwent staging by pelvic lymphadenectomy. Mean patient age was 63 years, with a range of 47 to 74 years. Clinical staging included a chest x-ray, excretory urogram, bone scan, and determination of serum alkaline and acid phosphatase levels. If no evidence of distant spread was found bipedal lymphangiography and computerized tomography (CT) of the abdomen were performed, combined with fine needle aspiration biopsy of suspicious lymph nodes in some cases. The local lesion in the prostate was staged by rectal examination by at least 2 clinicians. Prostatic core biopsies always were taken from both lobes. Clinical T stages were assigned according to the tumor, nodes and metastasis system,7 and they were not changed retrospectively. A comparison between this system and the Whitmore classification8 is shown in table 1. When doubt existed on whether a tumor should be staged as T3 or T4, cystoscopy and bimanual palpation with the patient under general anesthesia were performed. Transrectal ultrasound was performed to determine the baseline prostatic volume but it was not used to judge resectability of the tumor. It should be stressed that a stage T3 lesion is well defined in the tumor, nodes and metastasis system. The definition adheres to anatomical boundaries. If there is bladder neck invasion, fixation to the pelvic walls or fixation to the rectum the lesion is staged as T4. There is no point in making a segregation of large and small T3 lesions when resectability is concerned, since this is judged by the anatomical spread and not by size. However, a large tumor more often will be stage T4 instead of T3 than will a small tumor. Tumors were graded according to the Mostofi system 9 as either well (grade 1), moderately (grade 2) or poorly (grade 3) differentiated.
816
LOCALLY ADVANCED
48 had clinicai T3 and 3Li had clinical Fwzen sections of 6 Tl and 20 T2) the lymph nodes were for carcinoma at the time of lymphadenectomy in 23 of 48 patients with stage T3 and in 4 of 34 with stage T<3 disease. These patients were treated by early orchiectomy. Radical prostatectomy was performed in the remaining 55 patients (25 with T3 and 30 with T<3 cancer). Definite paraffin sections of the lymph nodes were positive in 4 of 25 patients with stage T3 and in 1 of 30 with stage T<3 cancer who were treated by radical prostatectomy. All 5 patients were followed without hormonal treatment or adjuvant irradiation. Of 21 patients with T3pNO cancer 5 (24 per cent) had pathological downstaging to pT2NO tumor, while 15 of 29 (52 per cent) with stage T<3 disease had upstaging to pT3NO. The final stages in patients with negative lymph nodes after histopathological examination of the prostate were stage T3pT<3 in 5, stage T3pT3 in 16, stage T<3pT3 in 15 and stage T<3pT<3 in 14 (fig. 1). Two patients (1 with stage T3pT3 and 1 with stage T<3pT3 tumor) were excluded from analysis of the time to progression, since death not related to cancer TABLE 1.
Comparison between the tumor, nodes and meta.stasis, and Whitmore staging systems Tumor, Nodes and Metastasis 7
Whitmore 8 Stage
Description
T Category
Description
-~-~--
A
Bl
Not palpable on rectal examination Al-focal Ca (:J chips) Unilobar <2 cm.
TOpTl-3
No palpable tumor
Tl
Tumor intracapsular, surrounded by palpably normal gland Tumor confined to gland, smooth nodule deforming contour but lat. sulci and seminal vesicles not involved Tumor extending beyond capsule with or without involvement of lat. sulci and/or seminal
B2 B3
Unilobar >2 cm. All other intracapsular
T2
C
Extending through capsule Cl-sulcus or sulci not free C2-base of seminal vesicles involved with or without sulci C:J-more than base of seminal vesicles with or without other adjacent
T3
T4
Tumor fixed or involving adjacent structures
PROSTATIC CARCINOMA
817
occurred 'sVithin 6 weeks "'~")]'"'',_, 29 ~~-"'-""~ with pT3NO tumor for analysis. Follovvup ranged from 9 to 105 months (mean 43 months) in the pT3NO group and 8 to 84 months (mean 41 months) in the T3pN+ group. Followup consisted of physical examination, and serum alkaline and acid phosphatase determinations 1 and 3 months after the initial operation, and at 3-month intervals thereafter. Chest x-rays and bone scans were performed at 6 to 12-month intervals. In the T3pN+ group treated by orchiectomy the prostatic volume was determined at regular intervals by transrectal ultrasonography. Progression of disease was defined as the appearance of biopsy proved local recurrence in radical prostatectomy patients or a 25 per cent or more increase in prostatic volume measured by transrectal ultrasonography in those who underwent orchiectomy. Furthermore, the appearance of distant metastases on chest x-rays or bone scans, or elevation of previously normal acid phosphatase levels on 2 consecutive followup examinations were considered to be proof of progressive disease. To study the impact of high clinical stage, positive margins of resection and seminal vesicle invasion on the probability of progression, the 29 patients with stage pT3NO disease were subdivided into 15 with stage T3pT3 and 14 with stage T<3pT3 cancer, 12 with and 17 without positive margins of resection, and 12 with and 17 without seminal vesicle invasion, respectively. Statistical analysis. To determine the comparability of the T3pT2 versus T<3pT3 groups, the positive versus negative pT3 margin groups, and the negative and positive seminal vesicle groups, the possible difference in frequency distribution of prognostic factors among the groups compared was tested for statistical significance by the chi-square or Wilcoxon test (table 2). N stages at pelvic lymphadenectomy and tumor grades in 23 patients with stage T3pN + disease treated by orchiectomy are listed in table 3. For the T3pN+ group and for all patients with pT3NO cancer, a curve representing the interval to treatment failure was constructed according to the Kaplan-Meier method. 10 The probability of progression ±2 standard errors for a given interval was calculated according to the method of Greenwood. Similarly, Kaplan-Meier curves were constructed for the T3pT3 versus T<3pT3, positive versus negative pT3 margin and negative versus positive seminal vesicle groups. Statistical significance of possible differences in interval to progression was tested by the Mantel-Cox test. RESULTS
For 23 patients with stage T3pN+ disease who underwent orchiectomy after lymphadenectomy, a time to progres-
structures
i
32 pN+ pts.
/\ 5 T3 pT<3
115 T3 pT3 I 114 T<3 pT3 l
14 T<3 pT<3
2 excluded
29 pT3 FIG. 1. Staging results of 82 patients considered to be potential candidates for radical prostatectomy because of clinically localized prostatic carcinoma (stages TO to T3). Of 2 excluded patients 1 originally had stage T3pT3 and 1 had stage T<3pT3 disease.
818
BOSCH, KURTH AND SCHROEDER TABLE 2.
Frequency distribution of prognostic factors in the groups compared pT3 Margin
No. pts. Mean mos. followup (range) Amount of tumor (No.): Large (>10 gm.) Medium (3-10 gm.) Small (<3 gm.) Grade of differentiation (No.): 2 3 Pos. margins of resection (No.) Seminal vesicle invasion (No.) T3 clinical stage (No.) *p
pT3 Seminal Vesicles
T3pT3
T<3pT3 Pos.
Neg.
Free of Disease
Invasion
15 31 (9-77)
14 55 (16-105)
12 42 (16-105)
17 43 (9-91)
17 40 (9-105)
12 47 (16-103)
8 7
10 3 1
11
9 6 6 7
6 8 6 5
5 7 6 5 6
1
7 9* 1 10
7 9 7 9
10 6 1
8 4
10 7 7
5 7 5
8
7
= 0.028. All other comparisons were not statistically significant. Pelvic lymphadenectomy staging and tumor grade in 23 patients with stage T3 pN+ prostatic cancer who underwent orchiectomy
TABLE 3.
Total No. Pts./ No. With Progression Total No. pts. Pelvic lymphadenectomy: Stage:
Nl N2 N3 N4 Tumor grade: 1 2 3
23/10 4/1 13/5
-/6/4
-/7/2 16/8
sion curve was constructed according to the method of Kaplan and Meier (fig. 2). Table 3 lists N stage at pelvic lymphadenectomy, tumor grade and incidence of progressive disease in this group. Progression occurred in 10 of 23 patients at an interval of 6 to 61 months. Of these patients 7 died of carcinoma after bone metastases had developed. The median interval to progression was 61 months. A total of 13 operations was necessary in these 10 patients, including transurethral resection of the prostate in 9, repeat transurethral resection of the prostate in 2, cutaneous ureterostomy and colostomy in 1, and urinary diversion in 1. Of 4 patients in the T3pT3N+ group treated by radical prostatectomy (representing the false negative frozen section cases) without hormonal or radiotherapy, the survival free of disease to date is 89, 55, 24 and 9 months, respectively. The prognostic factors for 15 patients with stage T3pT3NO and 14 with stage T<3pT3NO disease treated by radical prostatectomy without additional therapy are shown in table 2. Followup was shorter in the former group (mean 31 months) than in the T<3pT3NO group (mean 55 months). In regard to the frequency distribution of the prognostic factors between the groups no significant differences were noted: tumor grade (p = 0.5814), amount of tumor (reported in 3 categories of large >10 gm., medium 3 to 10 gm. and small <3 gm., p = 0.5045), seminal vesicle invasion (p = 0.8250) and positive margins of resection (p = 1.000). The probability of progression at 3 years was slightly higher in the T3pT3NO group than in the T<3pT3NO group (36 versus 27 per cent) (fig. 3). The difference observed was not statistically significant (p = 0.2964). The presumed site of spread at clinical examination was correlated with the findings at histopathological examination of the prostate in 15 patients with clinical stage T3pT3 disease treated by radical prostatectomy. None of these 15 patients had presumed invasion of the seminal vesicles. In 10 patients tumor growth beyond the capsule on 1 side was presumed at clinical examination and it was confirmed by histopathological findings. In addition, 4 of these 10 patients had invasion of 1 seminal vesicle. Positive margins of resection were found in 1 of the 4 patients. No seminal vesicle invasion was demonstrated
in 6 patients although positive margins of resection were found in 3 (all at the apex of the prostate). In 5 patients tumor growth beyond the capsule on both sides was presumed at clinical examination and was confirmed in all 5. In addition, 3 of the 5 patients had invasion of both seminal vesicles, including 1 who had positive margins of resection. No seminal vesicle invasion could be demonstrated in 2 patients, although positive margins of resection were found in 1. The margins of resection were positive in 12 of the 29 patients (41 per cent) with stage pT3NO disease. The impact on progression of positive margins of resection was studied by dividing the 29 patients into those with positive (12) and negative (17) margins (table 2). The followup interval was comparable. For the frequency distribution of prognostic factors between the groups no significant differences were noted in regard to tumor grade (p = 0.5938), T3 clinical stage (p = 1.000) and seminal vesicle invasion (p = 1.000). However, in the positive margin group significantly more patients with larger amounts of tumor were found than in the negative margin group (p = 0.0280). The probability of progression at 3 years was higher in those with positive than in those with negative margins (45 versus 20 per cent) (fig. 4). This difference was not statistically significant (p = 0.1879). Seminal vesicle invasion was present in 12 of the 29 patients studied (41 per cent). The impact on progression of seminal vesicle invasion was studied by dividing the 29 patients into those with (12) and those without (17) seminal vesicle invasion (table 2). Followup was shorter in the patients without seminal vesicle invasion (mean 40 months) than in those with seminal vesicle invasion (mean 47 months). In regard to the frequency distribution of prognostic factors between the groups, no significant differences were seen in tumor grade (p = 0.5938), amount of tumor (p = 0.6788), T3 clinical stage (p = 0.8250) and positive margins of resection (p = 1.000). The probability of progression was higher in the pT3 seminal vesicle invasion group than in the pT3 group without seminal vesicle invasion (47 versus 18 per cent) (fig. 5). The difference observed was not statistically significant (p = 0.2275). · Among the total group of patients with stage pT3 disease the probability of progression at 3 years was 30 per cent (fig. 6). When the state of margins of resection and the presence of seminal vesicle invasion were considered, it appeared that both parameters, when positive, contributed independently to progression and that the presence of both is ominous: 3 of 5 patients with positive margins and seminal vesicle invasion (60 per cent) and 1 of 10 with negative margins of resection without seminal vesicle invasion (10 per cent) had progression (table 4). All patients with progression had poorly differentiated tumors. Of 48 patients with clinical stage T3NX carcinoma with no signs of lymph node metastases on lymphangiography or CT total tumor removal, defined as negative lymph nodes at pelvic lymphadenectomy and negative margins of resection, could be achieved in 14 (29 per cent): 5 with T3pT2NO and 9 with T3pT3NO disease and positive margins.
SURGICAL TREATMENT OF LOCALLY ADVANCED
PROSTATIC CARCINOivi:A
819
100---..1...-------------1,. ____ _, L_ -~
% not failing
L.1
·------: L ___ - - - - - - - - - - - -
'L ________ - - - -----:
1___ -
-,
'
50
' '------------...
' ' ''
·-------------- - --1
'
''
1___ - - -
- -------------- -
---·
0
12
24
36
60
48
84
72
months follow-up
FIG. 2. Interval to treatment failure in 23 patients with stage T3pN+ disease who underwent orchiectomy. Dotted lines indicate probability of progression ±2 standard errors. Ticks on curve represent censored cases. Of 23 patients 8 had progression before 5 years of followup and 3 were followed for 5 years or longer. 100
''
L--~
% not failing
'--'-,
:., ~J._!J __ - ___ !, ___ JJ __ - J. ~ - - - - __
_j
T 3PT 3 (N=15)
50
IP=0.2964! 0
12
0
24
36
60
48
72
84
96
10B
months follow up
FIG. 3. Interval to treatment failure in 15 patients with stage T3pT3NO and 14 with stage T<3pT3NO prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 1 and 5, respectively, were followed for 5 years or longer. 100
I
I I
% not failing
L-
--"l
L- - - ,
'
LL __
SO
1
pT
J. __ __
3 margin
' l__ - --
-DL - - - - - - - ;
negative ( N =17)
~~;~i~-:_~~s~t~~; (~ :,-21 ---'
j
IP=0.1879
I
oL-~-~~ 0
12
24
36
48
60
72
84
96
108
months follow-up
FIG. 4. Interval to treatment failure in patients with margins positive (12) and negative (17) for stage pT3 prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 2 and 4, respectively, were followed 5 years or longer. DISCUSSION
Patients with positive lymph nodes (pN+) have systemic disease and should be treated accordingly. 11 In our study patients with clinical stage T3 disease and positive pelvic lymph nodes on frozen section were treated by early orchiectomy. The median interval to progression of 61 months is better than that reported in most series with varied treatment modalities. 12 A possible impact of this treatment on survival remains to be determined. If radical prostatectomy is restricted to patients
with stages T<3 cancer, up to 68 per cent 5 with a pT3 lesion will undergo an operation. Retrospective studies of treatment results with radical prostatectomy in pT3 cancer patients, mostly with unknown lymph node status, have been reported. 1- 4 In many patients hormonal treatment or adjuvant radiotherapy was used. Ten-year survival rates in pT3 cancer patients with unknown lymph node status have been reported to be 361, 29 2 and 39 per cent. 4 However, many patients received additional hormonal treatment in these series. The 10-year survival rates in pT3 cancer patients without lymph node involvement seem
820
BOSCH, KURTH AND SCHROEDER
100
,,
pT3 sem. vesic. free (N
I
Cl
= 17)
U---U---l
C:
I
I
L---------L------~~-~ pT3 sem. vesic. invasion (N = 12)
~
0 50 C:
cf!
p=0.2275
0
48
36
24
12
84
72
60
96
108
months follow up FIG. 5. Interval to treatment failure in 12 patients with and 17 without seminal vesicle invasion of stage pT3 disease after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 3 and 3, respectively, were followed for 5 years or longer.
.- -----~-- -- --, _-. ~~__, .,.,
100
I L.--1
% not failing
---------------------------------------
......
-~~~~~--~~~~-----~--~
-------,
L---------------------------------------
50
0
0
12
24
48
36
60
84
72
96
108
months follow-up FIG. 6. Interval to treatment failure in 29 patients with stage pT3 prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Dotted lines indicate probabUity of progression ±2 standard errors. Of patients 8 had progression before 5 years of foliowup and 6 were followed for 5 years or longer. TABLE
4. Impact on progression rate of state of margins of resection and seminal vesicle invasion in patients with stage pT3NO disease treated by
pelvic lymphadenectomy and radical prostatectomy Seminal Vesicle Invasion
Local recurrence Distant recurrence Total progression
No Seminal Vesicle Invasion
Resection Margins Pos. No./Total (%)
Resection Margins Neg. No,/Total (%)
Resection Margins Pos. No./Total (%)
Resection Margins Neg. No./Total (%)
2/5 (40) 1/5 (20) 3/5 (60)
1/7 (14) 1/7 (14) 2/7 (28)
1/7 (14) 1/7 (14) 2/7 (28)
0/10 (0) 1/10 (10) 1/10 (10)
to be considerably better. Flocks and associates reported a 67 per cent 10-year survival free of disease but these patients had interstitial instillations of 198gold at the time of radical prostatectomy.3 Zincke and associates reported a 10-year actuarial survival rate of more than 70 per cent, although only a small proportion of the patients actually was followed for that period and many patients received hormonal treatment. 13 Therefore, 10 and 15-year survival rates for patients with stage T3 disease and negative lymph nodes treated by radical prostatectomy who did not receive additional hormonal therapy or radiotherapy are not, in fact, available. Several reasons speak for the use of radical prostatectomy in patients with stage T3 cancer. There is the possibility of clinical overstaging of T3 tumors in up to 50 per cent. 6 In our series overstaging of clinical T3 tumors was found in 24 per cent of the patients. This percentage might be lowered by the use of transrectal ultrasound. These patients would have been excluded from radical prostatectomy on the basis of incorrect staging if one had not considered stage T3 prostatic carcinoma to be an indication for radical prostatectomy. There is evidence that stage T3 prostatic carcinoma can be cured in a certain proportion of the patients, even with poorly differentiated
tumors. Schroder noted in a series of 27 patients with stage T3 grade 3 disease treated by radical prostatectomy (lymph node status unknown) that 16 had recurrent carcinoma with an average survival of 4.5 years, while 11 others were alive or died without recurrence at an average of 12.6 years. 14 The latter group can be considered to be cured. Pelvic lymphadenectomy may help to select these patients. There may be an important palliative benefit for many patients after radical prostatectomy as far as local control is concerned. This benefit must be weighed against the risk of incontinence, which may be slightly higher in patients with locally advanced tumors. A high local progression rate (10 of 23 patients) was noted in the patients with positive lymph nodes who underwent orchiectomy in this study, as opposed to 4 of 29 patients with stage pT3 disease who were treated by radical prostatectomy. Local recurrence rates in stage C carcinoma patients treated by radical prostatectomy with or without adjuvant treatment have been reported to vary from 4.3 to 30 per cent. Schroeder and Belt reported a 12. 7 per cent local recurrence rate at 15 years in patients with pathological stage C disease treated by radical prostatectomy.1 In that series about half of the patients received additional hormonal therapy. Radiotherapy is failing
SURGICAL TREATMENT OF LOCALLY ADVANCED
to sterilize the local tumor xnore often as the clinical stage becomes more advancedo 15 · 16 In a recent review Scardino noted that positive post-irradiation prostatic biopsies are found in 45 to 74 per cent of the patients with clinical stage C cancer who underwent various forms of radiotherapy, 17 with a probability of local disease progression of 60 per cent at 5 yearso 16 In our study local recurrence was found in 4 of the 29 operated patients (14 per cent) with a mean interval to recurrence of only 9 months, while the mean followup was 43 months for the total group of patients with stage pT3 tumoL Gibbons and associates made a similar observation and reported a 30 per cent local progression rate with a mean interval to progression of 51 months, while the mean followup was 902 years in 23 patients with clinical stage C disease treated by radical prostatectomyo 18 Flocks and associates found a 4A per cent local recurrence rate among 335 patients treated by radical prostatectomy and 198gold interstitial instillations followed for 5 years, while in 69 who were followed for 15 years the local recurrence rate was identicaL'1 These figures suggest that local recurrence, if it does occur, happens early in followup in most caseso The cutoff point may be at approximately 5 yearso Do positive margins of resection have the same ominous meaning for the patient as positive post-irradiation prostatic biopsies? Zincke and associates reported on 10 patients with positive marginso 13 Of these patients 70 per cent had disease progression and 30 per cent died of cancer at an average followup of 5 yearso Of 3 patients who were free of disease 2 received adjuvant radiotherapyo Gibbons and associates reported on 6 patients with positive margins, of whom 1 had local and 3 had distant failureo 18 Failures occurred at between 4 and 190 months of followupo Two patients with distant failure had adjuvant radiotherapyo In our report the probability of progression at 3 years in patients with positive margins was 45 per cent compared to 20 per cent in those with negative marginso In the former group 3 local and 2 distant failures occurred, compared to 1 local and 2 distant failures in the latter groupo In 29 per cent of the patients who presented with a clinical stage T3 lesion, complete removal of tumor as defined by negative lymph nodes and negative margins of resection could be achievedo In light of these results it is suggested that although the numbers of patients studied is small a certain proportion of patients with clinical stage T3 disease will benefit from radical prostatectomyo This is to be expected, especially in those with stage T3pT3NO cancer and negative marginso Our patients with positive margins of resection had significantly larger amounts of tumor than those with negative marginso Preoperative estimation of the tumor amount could aid in the selection of these patientso REFERENCES L Schroeder, Fo R and Belt, K: Carcinoma of the prostate: a study of 213 patients with stage C tumors treated by total perinea! prostatectomyo Jo UroL, 114: 257, 19750 20 Boxer, R Jo, Kaufman, Jo Jo and Goodwin, WOK: Radical prostatectomy for carcinoma of the prostate: 1951-19760 A review of 329 patientso Jo UroL, 117: 208, 19770 30 Flocks, R ff, O'Donoghue, K Po No, Millerman, L Ao and Culp, Do Ao: Management of stage C prostatic carcinomao UroL Clino No AmeL, 2: 163, 19750 40 Elder, Jo So, Jewett, Ho Jo and Walsh, Po Co: Radical perinea! prostatectomy for clinical stage B2 carcinoma of the prostateo Jo UroL, 127: 704, 19820 50 Eggleston, Jo Co and Walsh, Po Co: Radical prostatectomy with preservation of sexual function: pathological findings in the first 100 caseso Jo UroL, 134: 1146, 19850 60 Catalona, WO Jo and Stein, Ao Jo: Staging errors in clinically localized prostatic carcinomao Jo UroL, 127: 452, 19820 70 Harmer, Mo R: TNM Classification of Malignant Tumourso Geneva: International Union Against Cancer, 19820 80 Whitmore, WO Fo, JL: Natural history and staging of prostate cancero UroL Clino No Amero, 11: 205, 19840
PROSTATIC CARCINOMP
9. Mostofi, Fo K: Grading of prostatic carcinomao Cancer ChernotheL Repo, 59: 111, 19750 100 Kaplan, K L and Meier, Po: Nonparametric estimation from incomplete observationso J. AmeL Stato Asso, 53: 457, 19580 1L Elder, Jo so and Catalona, WO Jo: Management of newly diagnosed metastatic carcinoma of the prostateo UroL Clino No AmeL, 11: 283, 19840 120 van Aubel, 00 Go Jo Mo, Hoekstra, WO Jo and Schroder, Fo R: Early orchiectomy for patients with stage D 1 prostatic carcinomao J 0 UroL, 134: 292, 19850 130 Zincke, ff, Utz, Do C, Benson, R C, JL and Patterson, Do K: Bilateral pelvic lymphadenectomy and radical retropubic prostatectomy for stage C adenocarcinoma of prostateo Urology, 24: 532, 19840 140 Schroder, Fo R: Prostatic carcinoma: comments on radical surgical treatmento Scando Jo UroL NephroL, suppL 55, po 181, 19800 150 Freiha, Fo So and Bagshaw, Mo Ao: Carcinoma of the prostate: results of post-irradiation biopsyo Prostate, 5: 19, 1984. 160 Scardino, Po To, Frankel, Jo Mo, Wheeler, To Mo, Meacham, R R, Hoffman, Go So, Seale, Co, Wilbanks, Jo R, Easley, Jo and Carlton, Co K, Jro: The prognostic significance of post-irradiation biopsy results in patients with prostatic canceL Jo UroL, 135: 510, 19860 170 Scardino, Po To: The treatment of localized prostatic cancero Scando Jo UroL NephroL, 20: 1, 19860 180 Gibbons, R Po, Cole, R So, Richardson, R Go, Correa, R Jo, JL, Brannen, Go K, Mason, Jo To, Taylor, WO Jo and Hafermann, Mo Do: Adjuvant radiotherapy following radical prostatectomy: results and complicationso Jo UroL, 135: 65, 19860 EDITORIAL COMMENT These investigators have analyzed carefully the pathological findings of prostates removed from patients believed to have clinical stage C disease in an attempt to detect what patients with this stage of disease might benefit from total prostatectomyo Possible benefactors have been defined as patients with negative lymph nodes and negative margins of resectiono However, they have shown clearly that seminal vesicle involvement also is a significant prognostic factor, with 47 per cent of the patients with seminal vesicle invasion having progression at 3 yearso Therefore, if one adds the presence of seminal vesicle invasion to negative lymph nodes and negative margins of resection, only 7 of 48 clinical stage 3 cancer patients (15 per cent) could be potentially "cured" with total prostatectomyo The problem obviously is how to exclude the other 85 per cent without surgeryo Transrectal ultrasonography probably will prove to be helpful in the staging of patients with clinical stage C disease who otherwise might be candidates for an operationo Scardino and associates showed that in 19 patients with clinical stage C disease on rectal examination 4 of the 5 whose disease appeared to be confined on rectal ultrasound proved to have confined disease at operationo 1 Despite better evaluation of the primary by rectal ultrasound, approximately 50 per cent of the patients with clinical stage C disease will have positive lymph nodes, and I agree with these investigators that patients with positive lymph nodes have systemic diseaseo To make a judgment regarding the palliative benefit (local control) of total prostatectomy performed on stage C cancer patients it is necessary to know the morbidity associated with the operation, and this information is not availableo A 2 per cent mortality rate is notedo Local control was achieved in 86 per cent of the patients in this series, which essentially is the same as can be expected with external beam radiation therapyo 2 This is an important feasibility study that is being performed carefully and analyzed at a major university center by an experienced group of scholarly investigators well versed in the study of prostatic canceL This study ultimately will provide important information on long-term survival rates of stage C cancer patients with negative lymph nodes treated by radical prostatectomy who did not receive adjuvant therapyo The number of patients followed to date is small, the followup is short and none of the observations in this paper has statistical significanceo The information in this preliminary report currently should not be used to justify the use of radical prostatectomy in patients with clinical stage C adenocarcinoma of the prostateo R. P Gibbons Section of Urology and Renal Transplantation Virginia Mason Medical Center Seattle, Washington
THE
Vol. 138, October
JOURNAL OF UROLOGY
Copyright© 1987 by The Williams & Wilkins Co.
Printed in U.S.A.
SURGICAL TREATMENT OF LOCALLY ADVANCED (T3) PROSTATIC CARCINOMA: EARLY RESULTS RUUD J. L. H. BOSCH, KARL H. KURTH
AND
FRITZ H. SCHROEDER
From the Department of Urology, Erasmus University Rotterdam, Rotterdam, The Netherlands
ABSTRACT
The fate of 48 patients with clinical stage T3 prostatic carcinoma after attempted curative surgical management was studied. In 23 of these patients positive frozen sections of the lymph nodes were found at pelvic lymphadenectomy and orchiectomy was performed. The median interval to progression was 61 months. Radical prostatectomy was performed in the remaining 25 patients. In 4 of these patients positive lymph nodes were found on paraffin sections but no additional treatment was given. Over-all, total tumor removal as defined by negative lymph nodes and negative margins of resection could be achieved in 14 of the 48 patients (29 per cent). During the same period 34 patients with clinical stage T<3 prostatic carcinoma were treated in a similar manner. Orchiectomy was done in 4 patients because of positive frozen sections of the lymph nodes and radical prostatectomy was done in 30, including 1 in whom positive paraffin sections of the lymph nodes were found but no additional treatment was given. An attempt was made to study the impact of several prognostic factors by comparing the probability of progression between patients with stage pT3 disease with (T3pT3NO) or without (T<3pT3NO) extracapsular tumor growth as determined by preoperative rectal examination (36 versus 27 per cent progression at 3 years), with or without positive margins of resection (45 versus 20 per cent progression at 3 years) and with or without involvement of the seminal vesicles (47 versus 18 per cent progression at 3 years). Our results suggest that a certain proportion of patients with clinical stage T3 disease will benefit from radical prostatectomy. This is to be expected especially in patients with stage T3pT3NO cancer and negative margins. (J. Ural., 138: 816-822, . 1987) The results of radical prostatectomy in patients with pathological stage T3 prostatic carcinoma, mostly with unknown lymph node status, have been reported by several investigators.1-4 In most of the series on stage pT3 disease treated by radical prostatectomy the inclusion of such patients resulted from clinical understaging. Patients believed to have an intracapsular tumor at clinical examination actually may have extraprostatic extension in up to 68 per cent of the cases. 5 Local overstaging of clinical T3 cancer is equally well known in up to 50 per cent of the cases. 6 In these patients abstinence from surgical treatment would result if clinical stage T3 prostatic carcinoma is not considered an indication for radical prostatectomy and a large group of curable patients would lose a chance for cure. Fair results have been obtained in patients with stage pT3 disease treated by radical prostatectomy as far as survival and especially local control are concerned. 1-4 We evaluated the possibility of surgical management of clinical stage T3 prostatic carcinoma. The fate of patients with surgically staged clinical T3 prostatic carcinoma is studied. Patients with T3pN + disease were treated by early orchiectomy. The median interval to progression in these patients is reported. Stage pNO cancer was treated by radical prostatectomy. The impact of high clinical stage was studied by comparing the probability of progression in cases of T3pT3NO and T<3pT3NO disease. Furthermore, we attempted to study the impact on progression of positive margins of resection in stage pT3 cancer treated by radical prostatectomy. Also, the impact on progression of seminal vesicle invasion was studied by comparing the probability of progression between patients with or without seminal vesicle invasion. Our results suggest that at least patients with T3pT3NO cancer and negative margins may benefit from radical prostatectomy. Unfortunately, it is difficult to identify this group preoperatively. Accepted for publication January 23, 1987.
MATERIALS AND METHODS
From 1977 to 1985, 82 prostatic carcinoma patients considered to be potential candidates for radical prostatectomy underwent staging by pelvic lymphadenectomy. Mean patient age was 63 years, with a range of 47 to 74 years. Clinical staging included a chest x-ray, excretory urogram, bone scan, and determination of serum alkaline and acid phosphatase levels. If no evidence of distant spread was found bipedal lymphangiography and computerized tomography (CT) of the abdomen were performed, combined with fine needle aspiration biopsy of suspicious lymph nodes in some cases. The local lesion in the prostate was staged by rectal examination by at least 2 clinicians. Prostatic core biopsies always were taken from both lobes. Clinical T stages were assigned according to the tumor, nodes and metastasis system,7 and they were not changed retrospectively. A comparison between this system and the Whitmore classification8 is shown in table 1. When doubt existed on whether a tumor should be staged as T3 or T4, cystoscopy and bimanual palpation with the patient under general anesthesia were performed. Transrectal ultrasound was performed to determine the baseline prostatic volume but it was not used to judge resectability of the tumor. It should be stressed that a stage T3 lesion is well defined in the tumor, nodes and metastasis system. The definition adheres to anatomical boundaries. If there is bladder neck invasion, fixation to the pelvic walls or fixation to the rectum the lesion is staged as T4. There is no point in making a segregation of large and small T3 lesions when resectability is concerned, since this is judged by the anatomical spread and not by size. However, a large tumor more often will be stage T4 instead of T3 than will a small tumor. Tumors were graded according to the Mostofi system 9 as either well (grade 1), moderately (grade 2) or poorly (grade 3) differentiated.
816
LOCALLY ADVANCED
48 had clinicai T3 and 3Li had clinical Fwzen sections of 6 Tl and 20 T2) the lymph nodes were for carcinoma at the time of lymphadenectomy in 23 of 48 patients with stage T3 and in 4 of 34 with stage T<3 disease. These patients were treated by early orchiectomy. Radical prostatectomy was performed in the remaining 55 patients (25 with T3 and 30 with T<3 cancer). Definite paraffin sections of the lymph nodes were positive in 4 of 25 patients with stage T3 and in 1 of 30 with stage T<3 cancer who were treated by radical prostatectomy. All 5 patients were followed without hormonal treatment or adjuvant irradiation. Of 21 patients with T3pNO cancer 5 (24 per cent) had pathological downstaging to pT2NO tumor, while 15 of 29 (52 per cent) with stage T<3 disease had upstaging to pT3NO. The final stages in patients with negative lymph nodes after histopathological examination of the prostate were stage T3pT<3 in 5, stage T3pT3 in 16, stage T<3pT3 in 15 and stage T<3pT<3 in 14 (fig. 1). Two patients (1 with stage T3pT3 and 1 with stage T<3pT3 tumor) were excluded from analysis of the time to progression, since death not related to cancer TABLE 1.
Comparison between the tumor, nodes and meta.stasis, and Whitmore staging systems Tumor, Nodes and Metastasis 7
Whitmore 8 Stage
Description
T Category
Description
-~-~--
A
Bl
Not palpable on rectal examination Al-focal Ca (
TOpTl-3
No palpable tumor
Tl
Tumor intracapsular, surrounded by palpably normal gland Tumor confined to gland, smooth nodule deforming contour but lat. sulci and seminal vesicles not involved Tumor extending beyond capsule with or without involvement of lat. sulci and/or seminal
B2 B3
Unilobar >2 cm. All other intracapsular
T2
C
Extending through capsule Cl-sulcus or sulci not free C2-base of seminal vesicles involved with or without sulci C:J-more than base of seminal vesicles with or without other adjacent
T3
T4
Tumor fixed or involving adjacent structures
PROSTATIC CARCINOMA
817
occurred 'sVithin 6 weeks "'~")]'"'',_, 29 ~~-"'-""~ with pT3NO tumor for analysis. Follovvup ranged from 9 to 105 months (mean 43 months) in the pT3NO group and 8 to 84 months (mean 41 months) in the T3pN+ group. Followup consisted of physical examination, and serum alkaline and acid phosphatase determinations 1 and 3 months after the initial operation, and at 3-month intervals thereafter. Chest x-rays and bone scans were performed at 6 to 12-month intervals. In the T3pN+ group treated by orchiectomy the prostatic volume was determined at regular intervals by transrectal ultrasonography. Progression of disease was defined as the appearance of biopsy proved local recurrence in radical prostatectomy patients or a 25 per cent or more increase in prostatic volume measured by transrectal ultrasonography in those who underwent orchiectomy. Furthermore, the appearance of distant metastases on chest x-rays or bone scans, or elevation of previously normal acid phosphatase levels on 2 consecutive followup examinations were considered to be proof of progressive disease. To study the impact of high clinical stage, positive margins of resection and seminal vesicle invasion on the probability of progression, the 29 patients with stage pT3NO disease were subdivided into 15 with stage T3pT3 and 14 with stage T<3pT3 cancer, 12 with and 17 without positive margins of resection, and 12 with and 17 without seminal vesicle invasion, respectively. Statistical analysis. To determine the comparability of the T3pT2 versus T<3pT3 groups, the positive versus negative pT3 margin groups, and the negative and positive seminal vesicle groups, the possible difference in frequency distribution of prognostic factors among the groups compared was tested for statistical significance by the chi-square or Wilcoxon test (table 2). N stages at pelvic lymphadenectomy and tumor grades in 23 patients with stage T3pN + disease treated by orchiectomy are listed in table 3. For the T3pN+ group and for all patients with pT3NO cancer, a curve representing the interval to treatment failure was constructed according to the Kaplan-Meier method. 10 The probability of progression ±2 standard errors for a given interval was calculated according to the method of Greenwood. Similarly, Kaplan-Meier curves were constructed for the T3pT3 versus T<3pT3, positive versus negative pT3 margin and negative versus positive seminal vesicle groups. Statistical significance of possible differences in interval to progression was tested by the Mantel-Cox test. RESULTS
For 23 patients with stage T3pN+ disease who underwent orchiectomy after lymphadenectomy, a time to progres-
structures
i
32 pN+ pts.
/\ 5 T3 pT<3
115 T3 pT3 I 114 T<3 pT3 l
14 T<3 pT<3
2 excluded
29 pT3 FIG. 1. Staging results of 82 patients considered to be potential candidates for radical prostatectomy because of clinically localized prostatic carcinoma (stages TO to T3). Of 2 excluded patients 1 originally had stage T3pT3 and 1 had stage T<3pT3 disease.
818
BOSCH, KURTH AND SCHROEDER TABLE 2.
Frequency distribution of prognostic factors in the groups compared pT3 Margin
No. pts. Mean mos. followup (range) Amount of tumor (No.): Large (>10 gm.) Medium (3-10 gm.) Small (<3 gm.) Grade of differentiation (No.): 2 3 Pos. margins of resection (No.) Seminal vesicle invasion (No.) T3 clinical stage (No.) *p
pT3 Seminal Vesicles
T3pT3
T<3pT3 Pos.
Neg.
Free of Disease
Invasion
15 31 (9-77)
14 55 (16-105)
12 42 (16-105)
17 43 (9-91)
17 40 (9-105)
12 47 (16-103)
8 7
10 3 1
11
9 6 6 7
6 8 6 5
5 7 6 5 6
1
7 9* 1 10
7 9 7 9
10 6 1
8 4
10 7 7
5 7 5
8
7
= 0.028. All other comparisons were not statistically significant. Pelvic lymphadenectomy staging and tumor grade in 23 patients with stage T3 pN+ prostatic cancer who underwent orchiectomy
TABLE 3.
Total No. Pts./ No. With Progression Total No. pts. Pelvic lymphadenectomy: Stage:
Nl N2 N3 N4 Tumor grade: 1 2 3
23/10 4/1 13/5
-/6/4
-/7/2 16/8
sion curve was constructed according to the method of Kaplan and Meier (fig. 2). Table 3 lists N stage at pelvic lymphadenectomy, tumor grade and incidence of progressive disease in this group. Progression occurred in 10 of 23 patients at an interval of 6 to 61 months. Of these patients 7 died of carcinoma after bone metastases had developed. The median interval to progression was 61 months. A total of 13 operations was necessary in these 10 patients, including transurethral resection of the prostate in 9, repeat transurethral resection of the prostate in 2, cutaneous ureterostomy and colostomy in 1, and urinary diversion in 1. Of 4 patients in the T3pT3N+ group treated by radical prostatectomy (representing the false negative frozen section cases) without hormonal or radiotherapy, the survival free of disease to date is 89, 55, 24 and 9 months, respectively. The prognostic factors for 15 patients with stage T3pT3NO and 14 with stage T<3pT3NO disease treated by radical prostatectomy without additional therapy are shown in table 2. Followup was shorter in the former group (mean 31 months) than in the T<3pT3NO group (mean 55 months). In regard to the frequency distribution of the prognostic factors between the groups no significant differences were noted: tumor grade (p = 0.5814), amount of tumor (reported in 3 categories of large >10 gm., medium 3 to 10 gm. and small <3 gm., p = 0.5045), seminal vesicle invasion (p = 0.8250) and positive margins of resection (p = 1.000). The probability of progression at 3 years was slightly higher in the T3pT3NO group than in the T<3pT3NO group (36 versus 27 per cent) (fig. 3). The difference observed was not statistically significant (p = 0.2964). The presumed site of spread at clinical examination was correlated with the findings at histopathological examination of the prostate in 15 patients with clinical stage T3pT3 disease treated by radical prostatectomy. None of these 15 patients had presumed invasion of the seminal vesicles. In 10 patients tumor growth beyond the capsule on 1 side was presumed at clinical examination and it was confirmed by histopathological findings. In addition, 4 of these 10 patients had invasion of 1 seminal vesicle. Positive margins of resection were found in 1 of the 4 patients. No seminal vesicle invasion was demonstrated
in 6 patients although positive margins of resection were found in 3 (all at the apex of the prostate). In 5 patients tumor growth beyond the capsule on both sides was presumed at clinical examination and was confirmed in all 5. In addition, 3 of the 5 patients had invasion of both seminal vesicles, including 1 who had positive margins of resection. No seminal vesicle invasion could be demonstrated in 2 patients, although positive margins of resection were found in 1. The margins of resection were positive in 12 of the 29 patients (41 per cent) with stage pT3NO disease. The impact on progression of positive margins of resection was studied by dividing the 29 patients into those with positive (12) and negative (17) margins (table 2). The followup interval was comparable. For the frequency distribution of prognostic factors between the groups no significant differences were noted in regard to tumor grade (p = 0.5938), T3 clinical stage (p = 1.000) and seminal vesicle invasion (p = 1.000). However, in the positive margin group significantly more patients with larger amounts of tumor were found than in the negative margin group (p = 0.0280). The probability of progression at 3 years was higher in those with positive than in those with negative margins (45 versus 20 per cent) (fig. 4). This difference was not statistically significant (p = 0.1879). Seminal vesicle invasion was present in 12 of the 29 patients studied (41 per cent). The impact on progression of seminal vesicle invasion was studied by dividing the 29 patients into those with (12) and those without (17) seminal vesicle invasion (table 2). Followup was shorter in the patients without seminal vesicle invasion (mean 40 months) than in those with seminal vesicle invasion (mean 47 months). In regard to the frequency distribution of prognostic factors between the groups, no significant differences were seen in tumor grade (p = 0.5938), amount of tumor (p = 0.6788), T3 clinical stage (p = 0.8250) and positive margins of resection (p = 1.000). The probability of progression was higher in the pT3 seminal vesicle invasion group than in the pT3 group without seminal vesicle invasion (47 versus 18 per cent) (fig. 5). The difference observed was not statistically significant (p = 0.2275). · Among the total group of patients with stage pT3 disease the probability of progression at 3 years was 30 per cent (fig. 6). When the state of margins of resection and the presence of seminal vesicle invasion were considered, it appeared that both parameters, when positive, contributed independently to progression and that the presence of both is ominous: 3 of 5 patients with positive margins and seminal vesicle invasion (60 per cent) and 1 of 10 with negative margins of resection without seminal vesicle invasion (10 per cent) had progression (table 4). All patients with progression had poorly differentiated tumors. Of 48 patients with clinical stage T3NX carcinoma with no signs of lymph node metastases on lymphangiography or CT total tumor removal, defined as negative lymph nodes at pelvic lymphadenectomy and negative margins of resection, could be achieved in 14 (29 per cent): 5 with T3pT2NO and 9 with T3pT3NO disease and positive margins.
SURGICAL TREATMENT OF LOCALLY ADVANCED
PROSTATIC CARCINOivi:A
819
100---..1...-------------1,. ____ _, L_ -~
% not failing
L.1
·------: L ___ - - - - - - - - - - - -
'L ________ - - - -----:
1___ -
-,
'
50
' '------------...
' ' ''
·-------------- - --1
'
''
1___ - - -
- -------------- -
---·
0
12
24
36
60
48
84
72
months follow-up
FIG. 2. Interval to treatment failure in 23 patients with stage T3pN+ disease who underwent orchiectomy. Dotted lines indicate probability of progression ±2 standard errors. Ticks on curve represent censored cases. Of 23 patients 8 had progression before 5 years of followup and 3 were followed for 5 years or longer. 100
''
L--~
% not failing
'--'-,
:., ~J._!J __ - ___ !, ___ JJ __ - J. ~ - - - - __
_j
T 3PT 3 (N=15)
50
IP=0.2964! 0
12
0
24
36
60
48
72
84
96
10B
months follow up
FIG. 3. Interval to treatment failure in 15 patients with stage T3pT3NO and 14 with stage T<3pT3NO prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 1 and 5, respectively, were followed for 5 years or longer. 100
I
I I
% not failing
L-
--"l
L- - - ,
'
LL __
SO
1
pT
J. __ __
3 margin
' l__ - --
-DL - - - - - - - ;
negative ( N =17)
~~;~i~-:_~~s~t~~; (~ :,-21 ---'
j
IP=0.1879
I
oL-~-~~ 0
12
24
36
48
60
72
84
96
108
months follow-up
FIG. 4. Interval to treatment failure in patients with margins positive (12) and negative (17) for stage pT3 prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 2 and 4, respectively, were followed 5 years or longer. DISCUSSION
Patients with positive lymph nodes (pN+) have systemic disease and should be treated accordingly. 11 In our study patients with clinical stage T3 disease and positive pelvic lymph nodes on frozen section were treated by early orchiectomy. The median interval to progression of 61 months is better than that reported in most series with varied treatment modalities. 12 A possible impact of this treatment on survival remains to be determined. If radical prostatectomy is restricted to patients
with stages T<3 cancer, up to 68 per cent 5 with a pT3 lesion will undergo an operation. Retrospective studies of treatment results with radical prostatectomy in pT3 cancer patients, mostly with unknown lymph node status, have been reported. 1- 4 In many patients hormonal treatment or adjuvant radiotherapy was used. Ten-year survival rates in pT3 cancer patients with unknown lymph node status have been reported to be 361, 29 2 and 39 per cent. 4 However, many patients received additional hormonal treatment in these series. The 10-year survival rates in pT3 cancer patients without lymph node involvement seem
820
BOSCH, KURTH AND SCHROEDER
100
,,
pT3 sem. vesic. free (N
I
Cl
= 17)
U---U---l
C:
I
I
L---------L------~~-~ pT3 sem. vesic. invasion (N = 12)
~
0 50 C:
cf!
p=0.2275
0
48
36
24
12
84
72
60
96
108
months follow up FIG. 5. Interval to treatment failure in 12 patients with and 17 without seminal vesicle invasion of stage pT3 disease after pelvic lymphadenectomy and radical prostatectomy. Of patients 5 and 3, respectively, had progression before 5 years of followup, and 3 and 3, respectively, were followed for 5 years or longer.
.- -----~-- -- --, _-. ~~__, .,.,
100
I L.--1
% not failing
---------------------------------------
......
-~~~~~--~~~~-----~--~
-------,
L---------------------------------------
50
0
0
12
24
48
36
60
84
72
96
108
months follow-up FIG. 6. Interval to treatment failure in 29 patients with stage pT3 prostatic carcinoma after pelvic lymphadenectomy and radical prostatectomy. Dotted lines indicate probabUity of progression ±2 standard errors. Of patients 8 had progression before 5 years of foliowup and 6 were followed for 5 years or longer. TABLE
4. Impact on progression rate of state of margins of resection and seminal vesicle invasion in patients with stage pT3NO disease treated by
pelvic lymphadenectomy and radical prostatectomy Seminal Vesicle Invasion
Local recurrence Distant recurrence Total progression
No Seminal Vesicle Invasion
Resection Margins Pos. No./Total (%)
Resection Margins Neg. No,/Total (%)
Resection Margins Pos. No./Total (%)
Resection Margins Neg. No./Total (%)
2/5 (40) 1/5 (20) 3/5 (60)
1/7 (14) 1/7 (14) 2/7 (28)
1/7 (14) 1/7 (14) 2/7 (28)
0/10 (0) 1/10 (10) 1/10 (10)
to be considerably better. Flocks and associates reported a 67 per cent 10-year survival free of disease but these patients had interstitial instillations of 198gold at the time of radical prostatectomy.3 Zincke and associates reported a 10-year actuarial survival rate of more than 70 per cent, although only a small proportion of the patients actually was followed for that period and many patients received hormonal treatment. 13 Therefore, 10 and 15-year survival rates for patients with stage T3 disease and negative lymph nodes treated by radical prostatectomy who did not receive additional hormonal therapy or radiotherapy are not, in fact, available. Several reasons speak for the use of radical prostatectomy in patients with stage T3 cancer. There is the possibility of clinical overstaging of T3 tumors in up to 50 per cent. 6 In our series overstaging of clinical T3 tumors was found in 24 per cent of the patients. This percentage might be lowered by the use of transrectal ultrasound. These patients would have been excluded from radical prostatectomy on the basis of incorrect staging if one had not considered stage T3 prostatic carcinoma to be an indication for radical prostatectomy. There is evidence that stage T3 prostatic carcinoma can be cured in a certain proportion of the patients, even with poorly differentiated
tumors. Schroder noted in a series of 27 patients with stage T3 grade 3 disease treated by radical prostatectomy (lymph node status unknown) that 16 had recurrent carcinoma with an average survival of 4.5 years, while 11 others were alive or died without recurrence at an average of 12.6 years. 14 The latter group can be considered to be cured. Pelvic lymphadenectomy may help to select these patients. There may be an important palliative benefit for many patients after radical prostatectomy as far as local control is concerned. This benefit must be weighed against the risk of incontinence, which may be slightly higher in patients with locally advanced tumors. A high local progression rate (10 of 23 patients) was noted in the patients with positive lymph nodes who underwent orchiectomy in this study, as opposed to 4 of 29 patients with stage pT3 disease who were treated by radical prostatectomy. Local recurrence rates in stage C carcinoma patients treated by radical prostatectomy with or without adjuvant treatment have been reported to vary from 4.3 to 30 per cent. Schroeder and Belt reported a 12. 7 per cent local recurrence rate at 15 years in patients with pathological stage C disease treated by radical prostatectomy.1 In that series about half of the patients received additional hormonal therapy. Radiotherapy is failing
SURGICAL TREATMENT OF LOCALLY ADVANCED
to sterilize the local tumor xnore often as the clinical stage becomes more advancedo 15 · 16 In a recent review Scardino noted that positive post-irradiation prostatic biopsies are found in 45 to 74 per cent of the patients with clinical stage C cancer who underwent various forms of radiotherapy, 17 with a probability of local disease progression of 60 per cent at 5 yearso 16 In our study local recurrence was found in 4 of the 29 operated patients (14 per cent) with a mean interval to recurrence of only 9 months, while the mean followup was 43 months for the total group of patients with stage pT3 tumoL Gibbons and associates made a similar observation and reported a 30 per cent local progression rate with a mean interval to progression of 51 months, while the mean followup was 902 years in 23 patients with clinical stage C disease treated by radical prostatectomyo 18 Flocks and associates found a 4A per cent local recurrence rate among 335 patients treated by radical prostatectomy and 198gold interstitial instillations followed for 5 years, while in 69 who were followed for 15 years the local recurrence rate was identicaL'1 These figures suggest that local recurrence, if it does occur, happens early in followup in most caseso The cutoff point may be at approximately 5 yearso Do positive margins of resection have the same ominous meaning for the patient as positive post-irradiation prostatic biopsies? Zincke and associates reported on 10 patients with positive marginso 13 Of these patients 70 per cent had disease progression and 30 per cent died of cancer at an average followup of 5 yearso Of 3 patients who were free of disease 2 received adjuvant radiotherapyo Gibbons and associates reported on 6 patients with positive margins, of whom 1 had local and 3 had distant failureo 18 Failures occurred at between 4 and 190 months of followupo Two patients with distant failure had adjuvant radiotherapyo In our report the probability of progression at 3 years in patients with positive margins was 45 per cent compared to 20 per cent in those with negative marginso In the former group 3 local and 2 distant failures occurred, compared to 1 local and 2 distant failures in the latter groupo In 29 per cent of the patients who presented with a clinical stage T3 lesion, complete removal of tumor as defined by negative lymph nodes and negative margins of resection could be achievedo In light of these results it is suggested that although the numbers of patients studied is small a certain proportion of patients with clinical stage T3 disease will benefit from radical prostatectomyo This is to be expected, especially in those with stage T3pT3NO cancer and negative marginso Our patients with positive margins of resection had significantly larger amounts of tumor than those with negative marginso Preoperative estimation of the tumor amount could aid in the selection of these patientso REFERENCES L Schroeder, Fo R and Belt, K: Carcinoma of the prostate: a study of 213 patients with stage C tumors treated by total perinea! prostatectomyo Jo UroL, 114: 257, 19750 20 Boxer, R Jo, Kaufman, Jo Jo and Goodwin, WOK: Radical prostatectomy for carcinoma of the prostate: 1951-19760 A review of 329 patientso Jo UroL, 117: 208, 19770 30 Flocks, R ff, O'Donoghue, K Po No, Millerman, L Ao and Culp, Do Ao: Management of stage C prostatic carcinomao UroL Clino No AmeL, 2: 163, 19750 40 Elder, Jo So, Jewett, Ho Jo and Walsh, Po Co: Radical perinea! prostatectomy for clinical stage B2 carcinoma of the prostateo Jo UroL, 127: 704, 19820 50 Eggleston, Jo Co and Walsh, Po Co: Radical prostatectomy with preservation of sexual function: pathological findings in the first 100 caseso Jo UroL, 134: 1146, 19850 60 Catalona, WO Jo and Stein, Ao Jo: Staging errors in clinically localized prostatic carcinomao Jo UroL, 127: 452, 19820 70 Harmer, Mo R: TNM Classification of Malignant Tumourso Geneva: International Union Against Cancer, 19820 80 Whitmore, WO Fo, JL: Natural history and staging of prostate cancero UroL Clino No Amero, 11: 205, 19840
PROSTATIC CARCINOMP
9. Mostofi, Fo K: Grading of prostatic carcinomao Cancer ChernotheL Repo, 59: 111, 19750 100 Kaplan, K L and Meier, Po: Nonparametric estimation from incomplete observationso J. AmeL Stato Asso, 53: 457, 19580 1L Elder, Jo so and Catalona, WO Jo: Management of newly diagnosed metastatic carcinoma of the prostateo UroL Clino No AmeL, 11: 283, 19840 120 van Aubel, 00 Go Jo Mo, Hoekstra, WO Jo and Schroder, Fo R: Early orchiectomy for patients with stage D 1 prostatic carcinomao J 0 UroL, 134: 292, 19850 130 Zincke, ff, Utz, Do C, Benson, R C, JL and Patterson, Do K: Bilateral pelvic lymphadenectomy and radical retropubic prostatectomy for stage C adenocarcinoma of prostateo Urology, 24: 532, 19840 140 Schroder, Fo R: Prostatic carcinoma: comments on radical surgical treatmento Scando Jo UroL NephroL, suppL 55, po 181, 19800 150 Freiha, Fo So and Bagshaw, Mo Ao: Carcinoma of the prostate: results of post-irradiation biopsyo Prostate, 5: 19, 1984. 160 Scardino, Po To, Frankel, Jo Mo, Wheeler, To Mo, Meacham, R R, Hoffman, Go So, Seale, Co, Wilbanks, Jo R, Easley, Jo and Carlton, Co K, Jro: The prognostic significance of post-irradiation biopsy results in patients with prostatic canceL Jo UroL, 135: 510, 19860 170 Scardino, Po To: The treatment of localized prostatic cancero Scando Jo UroL NephroL, 20: 1, 19860 180 Gibbons, R Po, Cole, R So, Richardson, R Go, Correa, R Jo, JL, Brannen, Go K, Mason, Jo To, Taylor, WO Jo and Hafermann, Mo Do: Adjuvant radiotherapy following radical prostatectomy: results and complicationso Jo UroL, 135: 65, 19860 EDITORIAL COMMENT These investigators have analyzed carefully the pathological findings of prostates removed from patients believed to have clinical stage C disease in an attempt to detect what patients with this stage of disease might benefit from total prostatectomyo Possible benefactors have been defined as patients with negative lymph nodes and negative margins of resectiono However, they have shown clearly that seminal vesicle involvement also is a significant prognostic factor, with 47 per cent of the patients with seminal vesicle invasion having progression at 3 yearso Therefore, if one adds the presence of seminal vesicle invasion to negative lymph nodes and negative margins of resection, only 7 of 48 clinical stage 3 cancer patients (15 per cent) could be potentially "cured" with total prostatectomyo The problem obviously is how to exclude the other 85 per cent without surgeryo Transrectal ultrasonography probably will prove to be helpful in the staging of patients with clinical stage C disease who otherwise might be candidates for an operationo Scardino and associates showed that in 19 patients with clinical stage C disease on rectal examination 4 of the 5 whose disease appeared to be confined on rectal ultrasound proved to have confined disease at operationo 1 Despite better evaluation of the primary by rectal ultrasound, approximately 50 per cent of the patients with clinical stage C disease will have positive lymph nodes, and I agree with these investigators that patients with positive lymph nodes have systemic diseaseo To make a judgment regarding the palliative benefit (local control) of total prostatectomy performed on stage C cancer patients it is necessary to know the morbidity associated with the operation, and this information is not availableo A 2 per cent mortality rate is notedo Local control was achieved in 86 per cent of the patients in this series, which essentially is the same as can be expected with external beam radiation therapyo 2 This is an important feasibility study that is being performed carefully and analyzed at a major university center by an experienced group of scholarly investigators well versed in the study of prostatic canceL This study ultimately will provide important information on long-term survival rates of stage C cancer patients with negative lymph nodes treated by radical prostatectomy who did not receive adjuvant therapyo The number of patients followed to date is small, the followup is short and none of the observations in this paper has statistical significanceo The information in this preliminary report currently should not be used to justify the use of radical prostatectomy in patients with clinical stage C adenocarcinoma of the prostateo R. P Gibbons Section of Urology and Renal Transplantation Virginia Mason Medical Center Seattle, Washington