Original Articles: Prostate Cancer: Variable Histology of Anastomotic Biopsies With Detectable Prostate Specific Antigen After Radical Prostatectomy

0022-5347/95/1533-1011$03.00/0

WEJOURNALOF UROLOGY Copyright 0 1995 by AMENCAN UROLOCICAL ASSOCIATION, INC.

Vol. 153,1011-1014, March 1995 Printed in U S A

VARIABLE HISTOLOGY OF ANASTOMOTIC BIOPSIES WITH DETECTABLE PROSTATE SPECIFIC ANTIGEN AFTER RADICAL PROSTATECTOMY JACKSON E. FOWLER, JR., JAY BROOKS, PRABHAKAR PANDEY From the Division

of

AND

LINDA E. SEAVER

Urology, University of Mississippi Medical Center, Jackson, Mississippi

ABSTRACT

Progressive elevation of the prostate specific antigen (PSA) level after radical prostatectomy for adenocarcinoma is generally considered as irrefutable evidence of recurrent tumor. We assessed the results of 62 biopsies of the vesicourethral anastomosis in 41 men who had 3 or more consecutive PSA levels of 0.4ng./ml. or greater after radical prostatedomy and no evidence of metastatic disease. The median PSA at the time of the first biopsy was 2.2 ng./ml. (range 0.4 to 50).Histological confirmation of recurrent cancer was established after 1biopsy procedure in 39% of the patients and after 1or more biopsy procedures in 59%. Biopsy was positive in 78% of 23 patients with a n abnormal digital rectal examination, 40% of 5 with an abnormal transrectal ultrasound only, and 23%of 13 with a normal digital rectal examination and ultrasound. Among the patients with and without biopsy proved tumor recurrence there were no s i g d i c a n t differences between the pathological stage or histological grade of the primary tumors, the month after surgery of the first detectable PSA level, the PSA doubling time, the month after surgery of the positive biopsy or the last negative biopsy, and the PSA level at the time of the positive biopsy or the last negative biopsy. In 6 cases benign prostatic tissue only was recovered from 1or more biopsy specimens. This experience demonstrates that in patients with a detectable PSA after radical prostatectomy recurrent cancer may be difficult to document by biopsy of the vesicourethral anastomosis. KEYWORDS:prostatic neoplasms, prostatectomy, antigens Prostate specific antigen (PSA), a serine protease synthesized by benign and malignant prostatic epithelium, is a sensitive serum marker for residual cancer after radical prostatectomy. Experience indicates that clinically manifest local tumor recurrence or metastatic disease is almost always preceded by a detectable PSA.1-7 However, few reports have detailed the results of biopsies of the vesicourethral anastomosis or prostatic bed in patients with a detectable postoperative PSA. We analyze the results of 62 anastomotic biopsies in 41 men with a detectable PSA after radical prostatectomy. MATERIALS AND METHODS

Patients. Between 1980 and 1991, 138 men underwent radical prostatectomy at the Veterans AfFairs Medical Center, Jackson, Mississippi for clinical stage A2 or B prostate cancer. Of these patients 11 received adjuvant radiation therapy and 1 was given androgen deprivation because of unsuspected pelvic lymph node metastases or seminal vesicle invasion, whereas the remainder received no anticancer therapy until there was clinical evidence of recurrent tumor. Four patients died of metastatic cancer and 5 died of intercurrent disease before the availability of an assay for PSA at the Veterans Affairs Hospital in April 1987.The remaining 129 patients, including 9 who received adjuvant radiation therapy, have been followed for 9 to 171 months (median 65 months) since surgery and for 4 to 87 months (median 59

of each seminal vesicle where it joins the prostate. The remaining gland is then divided into right and left halves, and 4 to 6 serial transverse sections are taken from each side. Additional sections may be taken from tumor bearing portions of the gland. Key pathological parameters include the presence or absence of capsular invasion, capsular perforation or seminal vesicle invasion. The tumors are graded using the system described by Gaeta et al.' The patients were generally seen by a urologist at 3 to 6-month intervals during the first 2 years after treatment and at 6 to 12-month intervals thereafter. A digital rectal examination, serum acid phosphatase determination and, during the last 7 years, a PSA measure were performed at each visit. Before 1990 surveillance bone scans were usually obtained every 1 to 2 years. Since 1990 bone scans have generally been obtained only when there is progressive PSA elevation. Of the 129 study patients a detectable PSA (2 or more consecutivelevels 0.4 ng./ml. or greater) was noted in 45 after surgery. Of the 45 patients with a detectable PSA 41 had no evidence of distant metastases on physical examination, chest x-ray or radionuclide bone scan, and they underwent 1 or more biopsies of the vesicourethral anastomosis. Of these 41 patients 38 had had no additional treatment aRer surgery and 3 received adjuvant radiation therapy. Of the remaining 4 patients with a detectablePSA metastatic cancer developed before biopsy in 2 and 2 died of intercurrent disease before a

1012

BIOPSIES AFTER RADICAL PROSTATECTOMY

within 3 to 20 months (median 6 months) before the first PSA level of 0.4 ng./ml. or greater. In 9 cases the first PSA level was obtained at 22 to 130 months (median 47 months) aRer surgery and ranged from 0.6 to 8.5 ngJml. (median 4.9). Of the 3 patients who received adjuvant radiation therapy the PSA level began to increase at 24 to 64 months after beginning treatment. At the time of the first biopsy the PSA ranged from 0.4 to 50 ng./ml. (median 2.2) and 23 of the patients (56%)had palpable abnormalities adjacent to the vesicourethral anastomosis. The magnitude of the palpable abnormalities was not graded but ranged from subtle induration to a small nodule. In 28 cases transrectal ultrasonography was performed by 1 of us (P. P.) before each biopsy. Of these 28 patients 15 had a hypoechoic, mixed hypoechoic-hyperechoic or hyperechoic mass adjacent to the tip of the funneled bladder outlet or adjacent to the external urinary sphincter above the genitourinary diaphragm. In none of the patients with a normal digital rectal examination or a normal ultrasound who had more than 1 biopsy was a digital rectal examination or ultrasound abnormal in the interval between the first and the most recent biopsy. The correlation between the findings on digital rectal examination and ultrasonography is shown in table 1. Ultrasound was abnormal in 10 of 11 patients (91%)with an abnormal digital rectal examination and 5 of 17 (29%)with a normal digital rectal examination. A total of 62 biopsies was performed on the 41 patients. For the purpose of this report each biopsy specimen was reexamined by 1 of us (J.B.). In preparation for the procedure the patients were generally treated with a fluoroquinolone antibiotic for 1 day and received a Fleet's enema immediately before the biopsy. A total of 42 biopsies was taken with ultrasound guidance using an 18 gauge automated biopsy needle and 20 were performed with digital guidance using a 16 gauge Tru-Cut* needle. When the procedure was performed with ultrasound guidance 1 to 2 biopsies were taken from each side of the vesicourethral anastomosis and, if present, 1 to 2 biopsies were taken from a visible abnormal lesion. The biopsy specimens were not designated by site of procurement and were submitted together as 1 tissue specimen. There were no episodes of fever, chills or significant bleeding after the biopsies, although self-limiting gross hematuria was not uncommon. PSA assays and PSA doubling time determinations. A polyclonal radioimmunoassayf was used to quantitate the serum PSA between July 1987 and April 1991. A monoclonal radioimmunoassayS has been used since April 1991. For the purpose of data analysis the PSA levels determined with the polyclonal assay were divided by 1.85 to correspond to the level of the monoclonal assay.g To calculate PSA doubling times the PSA levels were transformed to log 10 and plotted against time in months. A fitted regression line was determined with a least squares linear regression analysis and the PSA doubling time was calculated by dividing log 2 by the slope of the regression line. All PSA levels 0.4 ng./ml. or greater were included in the analyses. The significance of differences between sample means and medians was assessed with the 2-tailed t test and the Mann-Whitney test, respectively. A p value <0.05 was considered to indicate statistical significance. All results are reported as means -C 1 standard deviation unless indicated otherwise.

TABLE1. Findings on digital rectal examination and transrectal ultrasonography in 41 patients Ultrasonography Digital Rectal Examination Normal Abnormal Not Done Normal 12 5 1 Abnormal 1 10 12

histology of sequential biopsies in patients who underwent more than 1 procedure. Overall 30 biopsies revealed fibromuscular tissue only, 24 carcinoma, 7 benign prostatic hyperplasia (BPH) and 1 seminal vesicle epithelium. In 1 case the specimens from 1 biopsy procedure revealed BPH and carcinoma. On the basis of 1 or more biopsies 24 patients (59%)had carcinoma and 17 (41%)had no documented residual cancer. Of the latter patients 6 had 1 or more biopsies that revealed BPH, 1 seminal vesicle epithelium and 10 no documented residual glandular or epithelial tissue. Of the 23 patients with an abnormal digital rectal examination, 5 with an abnormal ultrasound only and 13 with a normal digital rectal examination and normal ultrasound 1 or more biopsies revealed carcinoma in 19 (78%),2 (40%)and 3 (23%),respectively. The interval between the first and last biopsy in the 8 patients with documented carcinoma who underwent more than 1 biopsy ranged from 4 to 31 months (median 17.5) and the interval between the first and the most recent biopsy in the 7 patients who had more than 1 biopsy done but no documented residual carcinoma ranged from 5 to 39 months (median 12, p >0.05). The interval PSA increase in the former and latter patients ranged from 0.1 to 37.1 ng./ml. (median 4.0) and from 0.1 to 49.4 ng./ml. (median 1.2), respectively (p >0.05). Characteristics of the 38 patients who did not receive adjuvant radiation therapy are shown in table 3. There were no statistically significant differences between these parameters in the patients with documented recurrent cancer and in those without recurrent cancer when considered as a group or when stratified by the histological findings of BPH, fibromuscular tissue or seminal vesicle epithelium. DISCUSSION

This study supports the concepts that recurrent tumor after radical prostatectomy is preceded by the developmentof a detectable PSA and that a detectable PSA may be the only manifestation of recurrent cancer. However, the data also show that suspected tumor recurrence may be difficult to document by biopsy even if the digital rectal examination suggests recurrent disease. In 6 patients the biopsies revealed benign prostatic tissue only. Although this phenomenon, which has been reported also by Foster et a1,6 raises the possibility that residual benign tissue resulting from unintentional disruption of the prostatic capsule during surgery may be responsible for a detectable postoperative PSA, several observations indicate that undetected carcinoma probably coexists with the benign tissue. In the cases in which benign prostatic tissue and no carcinoma was recovered from biopsy specimens the PSA at the time of the last biopsy ranged from 0.6 to 4.8 ng./ml. (median 1.6).However, Stamey et a1 have shown that every gram of benign prostatic tissue produces an average 0.31 RESULTS The results of the 62 prostate biopsies relative to the find- ng./ml. increase in the serum PSA level' and it seems unings on digital rectal examination and ultrasonography are likely that 1.9 to 15.5 gm. of benign prostatic tissue could be shown in table 2. The data are organized to demonstrate the left in situ after surgery. Also, the PSA doubling time in these cases ranged from 7.3 to 100 months (median 21.9) and increased in an exponential manner, However, PSA elevation * Travenol Laboratories, Deerfield, Illinois. in men with progressive BPH is thought to be linear in t Pros-Check, Yang Laboratories, Bellevue, Washington t. Tandem-R, Hybritech, San Diego, California. nature and is only about 0.09 ng./ml. annually." Finally, in

1013

BIOPSIES AFTER RADICAL PROSTATECTOMY TABLE2. Sequential biopsy results relative to findings on digital rectal examination and ultrasonography Biopsy Results No. Pts. Biopsy 2

Biopsy 1 Ca Fibromuscular tissue Fibromuscular tissue BPH BPH Fibromusculartissue Fibromuscular tissue Fibromuscular tissue Fibromuscular tissue Fibromuscular tissue Fibromuscular tissue

Abnormal Digital Rectal Examination

Biopsy 3

Abnormal Ultrasound Only

Ca Fibmmuscular tissue Ca

Neither

Total

2 1

16 4 3 1 4 1 1 1 6 2 2

-

-

-

3 -

BPH Fibromuscular tissue Seminal vesicle epithelium

-

-

Fibromuscular tissue Fibromuscular tissue

3 2 2

Fibromuscular tissue

TABLE3. Patient characteristics relative to biovsv result

No.pts.* Pathological stage: Organ confined Capsular perforation Seminal vesicle invasion Tumor grade:

2 3

4 Mo. PSA detectable (range)

Ca

No Ca

BPH

Fibmmuscular Tissue Only in 1 or More Biopsies

Biopsy Revealing Seminal Vesicle Epithelium

23

15

6

8

1

11

6

4

6

7 2

2

2 5 1

7 6 2

4 2

6

9 13 1 29 c 30

45 f 39 (0-86) (&122) ( N = 16) (N = 13) PSA doubling time (mos.l (range) 33 2 63 27 f 27 (4.1-300) (2.1-100) Mo. biopsy (rangelt 66 2 35 68 f 51 112-1371 (4-179) PSA at biopsy Irange)S 9.42 13.5 4.5 f 7.9 (0.6-50) (0.4-32) * Three patients who received adjuvant radiation therapy are excluded. t Month after surgery of last biopsy or positive biopsy. f PSA at last biopsy or positive biopsy.

1 case the first biopsy revealed benign prostatic tissue but a second biopsy demonstrated carcinoma, and in 1 case the specimens from 1biopsy procedure revealed BPH and carcinoma. However, the recovery of benign prostatic tissue from anastomotic biopsies demonstrates that violation of the prostatic capsule with incomplete removal of the entire prostate gland is a cause for persistent cancer after radical prostatectomy. We are also reluctant to assign significance to the recovery of fibromuscular tissue in patients with a detectable PSA after radical prostatectomy because 7 of our 23 patients (30%)with documented residual cancer had 1or more earlier biopsies that revealed fibromuscular tissue only. Others have reported that 58 to 65% of patients with a detectable PSA after radical prostatectomy will have an unrevealing initial biopsy,"-7 and Foster et a1 have shown that 47% of such patients will have carcinoma documented with subsequent biopsies.fi Biopsies that show fibromuscular tissue only probably result from sampling errors but tumor recurrence at other sites in the pelvis or occult distant metastases are reasonable alternative explanations. The imprecision of anastomotic biopsies for confirming residual cancer in men with a detectable PSA after radical prostatectomy raises questions about the necessity of the procedure for patient management. On the one hand, histological documentation of recurrent cancer before further treatment is undertaken is a sound oncological practice. On the other hand, logic dictates that it is virtually impossible for a man to have a rapidly escalating PSA level after radical prostatectomy without persistent cancer, and experience to

-

-

40 f 37 (0-109)

(N = 6 ) 33 f 34 (7.%100) 59 f 48 (4-145) 2.2 f 1.6 (O.M.8)

3 4 1 56 2 42 10-122) iN = 6 ) 24 2 23 (5.2-75) 80 f 55 (30-179) 2.7 f 2.2 (0.4-6.1)

-

1 10 2.1 22 32

date suggests that the overall chance of documenting recurrent cancer with 1biopsy procedure is less than 50%.When the digital rectal examination was normal the likelihood of detecting carcinoma with 1biopsy procedure was only 11% in our study and ranged from 27 to 31% in other experience^."^ We are persuaded that the rate of PSA elevation as measured by the PSA doubling time is the most useful parameter for determining the malignant potential of recurrent tumor after radical prostatectomy, and we use this information rather than biopsy of the vesicourethral anastomosis to assess the necessity and nature of subsequent treatment. In an analysis of 50 men who had a detectable PSA after radical prostatectomy we have shown that the mean PSA doubling time in those who did and did not have distant metastases before the institution of additional treatment was 4.7 months (range 0.8 to 7.9 months) and 23.3 months (range 2.1 to 300), respectively." Danella et a1 have made similar observations about the doubling times in patients in whom distant metastases develop after surgery.'* These data suggest that a PSA doubling time of less than 6 months is a harbinger of rapidly progressive disease and that attempts to obtain histological confirmation of persistent cancer may not be warranted. Conversely, the data also suggest that the growth rate of recurrent tumor after radical prostatectomy can be exceedingly slow. This phenomenon and the absence of convincing evidence that adjuvant radiation therapy or androgen deprivation prolongs survival provide a compelling rationale for observation only in the patient with a long PSA doubling time who is asymptomatic and has no palpable evidence of local tumor recurrence or clinical evidence of distant metas-

1014

BIOPSIES AFTER RADICAL PROSTATECTOMY

tases. Regardless of one's philosophy about the management of patients with slow elevation of the PSA after radical prostatectomy, a biopsy seems inappropriate if the patient is well informed about the nature of further therapy and does not express an interest in pursuing additional treatment.

part 2,147:952,1992. 6. Foster, L. S., Jajodia, P., Fournier, G., Jr., Shinohara,

K,

Carroll, P. and Narayan, P.: The value of prostate specific antigen and transrectal ultrasound guided biopsy in detecting prostatic fossa recurrences following radical prostatectomy. J. Urol., 149 1024,1993. 7. Lightner, D. J., Lange, P. H., Reddy, P. K and Moore, L.: ProsREFERENCES tate specific antigen and local recurrence after radical prostatectomy. J. Urol., 144:921,1990. 1. Stamey, T. A, Yang.N., Hay, A. R., McNeal, J. E., Freiha, F. S. 8. Gaeta, J. F., Asirwatham, J . E., Miller, G. and Murphy, G. P.: and Redwine, E.: Prostate-specific antigen as a serum marker Histologicgrading of primary prostatic cancer: a new approach for adenocarcinoma of the prostate. New Engl. J. Med., 317: to an old problem. J. Urol., 123 689,1980. 909, 1987. 9. Hortin, G. L., Bahnson, R. R., Daft, M., Chan, K-M., Catalona, 2. Partin, A. W., Pound, C. T., Clemens, J. Q., Epstein, J. I. and W. J . and Ladenson, J . H.: Differences in values obtained with Walsh, P. C.: Serum PSA after anatomic radical prostatec2 assays of prostate specific antigen. J. Urol., 139 762,1988. tomy. The Johns Hopkins experience after 10years. Urol. Clin. 10. Carter, H. B., Pearson, J . D., Metter, E. J., Brant, L. J., Chan, N. Amer., 20 713, 1993. D. W., Andres, R., Fozard, J. L. and Walsh, P. C.: Longitudinal 3. Stamey, T. A, Graves, H. C. B., Wehner, N., Ferrari, M. and evaluation of prostate specific antigen levels in men with and Freiha, F. S.: Early detedion of residual prostate cancer &r without prostate disease. J.A.M.A., 267:2215,1992. radical prostatectomy by an ultrasensitive assay for prostate 11. Fowler, J. E.,Jr., Pandey, P., Braswell, N. T. and Seaver, L.: speciiic antigen. J. Urol., 149 787, 1993. Prostate specific antigen progression rates &r radical pros4. Takayama, T. K,Vessella, R. L., Brawer, M. K, Notebroom, J. tatectomy or radiation therapy for localized prostate cancer. and Lange, P. H.: The enhanced detection of persistent disease Surgery, 116 302,1994. after prostatectomy with a new prostate specific antigen im12. Danella, J., Steckel, J., Dorey, F., Smith, R. B. and deKernion, munoassay. J. Urol., 150:374, 1993. J. B.: Detectable prostate specific antigen levels followingrad5. Abi-Aad, A. S., Macfarlane, M. T., Stein, A. and deKernion, J . B.: ical prostatectomy: relationship of doubling time to clinical Detection of local recurrence after radical prostatectomy by outcome. J. Urol., part 2,149 447A, abstract 938, 1993. prostate specsc antigen and transrectal ultrasound. J. Urol.,