Routine prostate biopsies following radiotherapy for prostate cancer: Results for 226 patients

ROUTINE PROSTATE BIOPSIES FOLLOWING RADIOTHERAPY FOR PROSTATE CANCER: RESULTS FOR 226 PATIENTS J.M. CROOK, M.D. G.A. PERRY, M.D. S. ROBERTSON, M.D. B.A. ESCHE, M.D. From the Department and the Department

ABSTRACT-Objectives. prostate outcome.

cancer

following

of Radiation Oncology, Ottawa Regional Cancer Centre, of Pathology, Ottawa General Hospital, Ottawa, Canada

To determine the time course radiotherapy (RT) and to correlate

of histologic resolution of biopsy results with clinical

Methods. Since July 1990, all patients treated with radical external beam RT for prostate cancer at the General Division of the Ottawa Regional Cancer Centre have had systematic transrectal ultrasound (TRUS) and TRUS-guided biopsies beginning 12 months after RT and then every 6 months until negative or until clinical failure. Thus, 226 patients have had 375 TRUS with four to seven specimens per examination. Stage distribution was Tl b: 32, Tl c: 1 1, T2a: 45, T2b: 82, T3: 50, and T4: 6. Median follow-up was 33 months. Results. Biopsy results were negative in 69.5% of patients by 30 months of followup. Thirty-two (14%) had local failure (Tl b: 12.5%, Tl c: 0%, T2a: 1 1 %, T2b: 15%, T3: 18%, T4: 33%). Seven (3%) had chemical failure, and 47 (21%) had biopsy-only failure. Median follow-up for the biopsy-only failure group is only 19.5 months and mean prostate-specific antigen (PSA) is 1 .O ng/mL. Thirty-nine patients, initially with biopsyonly failure, have converted to negative biopsies at a median of 26 months. Nadir PSA for patients with local failure was 3.9 ng/mL at 14 months versus 0.7 ng/mL at 23 months for those without failure. Patients with late conversion to negative biopsy results had a later nadir PSA of 1.3 ng/mL at 27.3 months. \ ~\_ \ Conclusions. Routine prostate biopsy specimens after RT in an unselected ‘population show tumor clearance that is in agreement with long-term clinical follow-up, although tumor may take more than 30 months to resolve. Nadir PSA can be used to predict outcome.

External beam radiotherapy CRT) is the most commonly used potentially curative treatment for carcinoma of the pr0state.i In general, patients selected for radiation are 5 to 10 years older than those selected for radical prostatectomy,2 have more advanced local disease, and are of unknown nodal status. Despite a clearly less favorable patient population, clinical results at 10 years are equivalent to those for surgery.lm3 Submitted accepted

624

(Rapid Communication): November (with revisions): December 5, 1994

17,

1994,

Clinical follow-up after RT may be unreliable. The use of prostate-specific antigen (PSA) is routine, but a rising PSA may signal either local or distant recurrence. Prostate biopsy is logically the gold standard for determination of local failure but both the indications and interpretation of biopsy specimens are controversial.4 Several reports on selective post-RT biopsies, often including patients who had biopsies a few months after treatment or in whom residual disease is suspected, have shown very high positive biopsy rates.5-7 Such reports cast doubts on the efficacy U ROLOGP

/APRIL 1995 I VOLUME45, NUMBER4

TABLE I.

TNM staging for prostate cancer (1 992)21

Stage

Definition

Tl

Clinically Tumor Tumor Tumor Confined Tumor Tumor Tumor Tumor Unilateral Bilateral Tumor Tumor Tumor Tumor

Tla Tlb Tic T2 T2a T2b T2c T3 T3a T3b T3c T4 T4a T4b

or Criteria

for Inclusion

inapparent tumor, not palpable nor visible by imaging an incidental histologic finding; <5% of tissue resected an incidental histologic finding; > 5% of tissue resected identified by needle biopsy (eg, because of elevated serum PSA) within the prostate involves half of a lobe or less involves more than half of a lobe but not both lobes involves both lobes extends through the prostate capsule extracapsular extension extracapsular extension invades seminal vesicle(s) is fixed or invades adjacent structures other than seminal vesicles invades bladder neck and/or external sphincter and/or rectum invades levator muscles and/or is fixed to pelvic wall

of radiation in the curative management of prostate cancer. Positive biopsy specimens are of concern because they are associated with distant failure and death from prostate cancer. 8-13However, the difficulty in interpretation of post-RT biopsy results is often understated. Histologic clearance of tumor following RT may take 18 months or longer.gv14-16Even when biopsies are performed at an appropriate interval following treatment, major pitfalls exist. Radiation atypia in benign prostate glands may be confused with residual or recurrent tumor, leading to overcall of positive biopsy resu1ts.l’ Immunohistochemical stains for high molecular weight keratin can differentiate radiation atypia from residual tumor, since the basal cell layer of benign glands stains positive, whereas malignant glands are negative. l8 Most series, however, do not document staining for high molecular weight keratin. Tumor resolution after RT may leave scattered nests of cells showing marked radiation change. Since there is no identifiable glandular morphology, these remnants would be given a high Gleason score. Recently developed immunohistochemical stains can differentiate between rapidly proliferating poorly differentiated residual tumor and degenerated nonproliferating cells. Proliferative cell nuclear antigen (PCNA) is a nonhistone nuclear protein elaborated on the nuclear membrane of actively cycling cells, but absent in those cells that are not proliferating. Levels correlate well with other indices of proliferative activity, such as in vivo 5bromodeoxyuridine (BuDR) staininglg and tumor grade.20 In this prospective study, routine transrectal ultrasound (TRUS)-guided biopsies were done systematically on 226 unselected patients to determine the time course of histologic resolution and

UROLOGY@

I APRIL 1995 I VOLUME 45, NUMBER 4

to correlate biopsy results at intervals following RT with clinical outcome. MATERIAL

AND METHODS

In July 1990, a policy of obtaining routine postRT-TRUS-guided prostate biopsy specimens was introduced at the General Division of the Ottawa Regional Cancer Centre for all prostate cancer patients treated by pelvic RT with curative intent. Biopsies were scheduled 12 months following RT, then every 6 months until negative or until the development of clinical recurrence. Patients with initial negative biopsy results at 12 months had another biopsy at 36 months. We report the results on 226 patients aged 49 to 87 years (median, 70), treated from July 1987 to February 1993, all with histologically confirmed adenocarcinoma of the prostate. Data were updated in March 1994 and analyzed with the Kwikstat 3.3 statistical data analysis program (TexaSoft). All patients had a complete history and physical examination at the time of initial consultation. Local tumor stage was determined by digital rectal examination (DRE), performed by both the radiation oncologist and the referring urologist. If hormonal intervention had been initiated before referral, the tumor stage assigned was as described by the referring urologist. Investigations included complete hlood count, renal and hepatic function tests, serum alkaline and acid phosphatases, PSA (since November 1989, Abbot IMX assay: normal range to 5.3 ng/mL), chest radiograph, technetium-99m bone scan, and pelvic computed tomography (CT) scan. Thirty-three patients (14.6%) had a staging pelvic lymphadenectomy. Staging is according to International Union Against Cancer-TNM classification of 199221 (Table I). Distribution by stage and grade is shown in Table II.

625

TABLE II.

Distribution of patients stage and grade

by

Grade Stage Tl tiTic T2a T2b T3 T4 Total

WD 14 8 30 33 14 2 101

MD 13 3 11 40 30 - 1 98

PD 5 0 1 4 4 2 16

GX 0 0 3 5 2 1 11

KEY: WD = well dif~ereerentiated (Gleason scow 2-4); MD = moderately d$Jereerentiated (Gleason scox 5-7); PD = poorly differentiated (Gleason SCDIC 8-10); GX = unknown grade.

Forty-two percent of patients (94 of 226) had hormonal intervention (median duration, 5 months; range, 1 to 60 months) before referral for definitive RT. Hormonal treatment was discontinued before RT and not reinstituted except for documented failure. Patients with orchiectomy were excluded. Patients who failed distantly in the first year of follow-up also had a biopsy if they were close to the 12-month mark. All patients were treated with 18 MV photons using a four-field box technique. The treatment volume was limited to the prostate and seminal vesicles for small well-differentiated tumors and those that were pathologically node negative. Most patients were treated to the first echelon (external and internal iliac) nodes with fields extending superiorly to the bottom of the sacroiliac joints. Treatment to the whole pelvis was rarely used. Standard fractionation of 1.8 to 2.0 Gy/day was used. Lymph nodes received 45 to 46 Gy, the prostate and seminal vesicles received 65 to 66 Gy (range, 60 to 68 Gy). The mean follow-up is 33 months (range, 12 to 78 months). Patients were seen every 3 to 4 months for the first 2 years, every 6 months until 5 years, and yearly thereafter, with a PSA determination at each visit. The first post-treatment biopsy was scheduled 12 months after RT. Prostates with residual tumor had a biopsy every 6 months until negative or until clinical evidence of failure (rising PSA values or abnormal DRE). Repeat biopsies were also performed to investigate a rising PSA or abnormal DRE. All biopsies were performed under TRUS guidance with antibiotic prophylaxis (ciprofloxacin, 500 mg every 12 hours for three doses). All quadrants of the prostate were sampled, as well as the site of the original tumor, which was targeted for an additional two or three passes. In total four to seven samples were taken per biopsy session.

626

All biopsy results were reviewed by one pathologist (S.R.) and stained for PSA and prostatic acid phosphatase. Immunohistochemical stain for high molecular weight keratin (keratin 903) was used to distinguish residual carcinoma from radiation atypia in benign glands. Biopsies were considered positive if hematoxylin and eosin staining showed any evidence of residual malignancy, regardless of the scarcity of malignant cells or the degree of radiation effect. Positive biopsies were then subdivided into “clearly positive” and “indeterminate” categories, based on the degree of radiationinduced degenerative change. Positive or indeterminate biopsies were also processed with an immunohistochemical stain for PCNA, using the NovoCastra PC10 antibody. Fixation time in formalin was limited to 12 hours. PCNA staining was expressed as the percentage of tumor cells staining positive. Patient status was considered as no evidence of disease (NED) if both serum PSA and DRE were normal and the biopsy results were negative. Positive and indeterminate biopsy results were classified as biopsy failures if both DRE and serum PSA values were normal. If either PSA or DRE was abnormal, together with a positive or indeterminate biopsy result, the patient was classified as having a local failure. An elevated PSA level, with no evidence of tumor on either DRE or a biopsy specimen was deemed a chemical failure. Patients with distant metastases were classified as distant failures only if the prostate biopsy result was normal; otherwise they were considered to have combined local and distant failure, even if the DRE was clinically normal. R&ULTS Three hundred seventy-five biopsies were performed on 226 patients: 117 patients had one biopsy, 76 patients had two biopsies, 26 patients had three biopsies, and seven patients had four biopsies. Status of the 226 patients as of March 1994 is shown in Table III. Of the 226 patients, 51% (115 of 226) had an abnormal initial biopsy result at a median time of 13 months after RT. Seventy of the 115 (61%) showed malignant cells with apparently minimal treatment effect. Of these, 24% (17 of 70) progressed to local failure at a median of 30 months, and 33% (23 of 70) converted to negative at a median of 26 months. Of 115 patients, 45 (39%) were considered indeterminate because of marked degenerative changes. Of these, only 9% (4 of 45) progressed to local failure and 29% (13 of 45) converted to negative. Delayed conversion to negative occurred in all T stages

LJ ROLOCY@ /APRIL

1995

i VOLUME

45, NUMBER

4

*Vander Wed Messing= $-Herr rs *Crook *

(present

series)

14

cox

FIGURE 1. Percent positive biopsy versus time from completion of radiation therapy [RT). [Modified from Crook et aLz2)

10

* Schellhammer

~0 3 6 9 12 15 18 21 24 27 30 33 i-5 Months post RT TABLE III. Stage

Status

NED

of patients

as of March

1994

BF

CF

LF*

Tlb Tic T2a T2b T3 T4

21 6 28 46 21 0

5 5 9 15 10 3

1 0 1 3 2 0

2 0 4 10 6 0

2 0 1 2 3 2

Total %

122 54.0

47 20.8

7 3.1

22 9.7

10 4.4

(Tlb: 6, Tic: 1, T2a: 12, T2b: 14, T3: 6). The proportion of normal and abnormal biopsies as a function of time is shown in Figure 1. By 30 months, 69.5% of patients had achieved a negative biopsy. Eleven (7%) of the 150 patients with negative posttreatment biopsy results progressed to local failure at a median time of 36 months. Overall, 14% of patients (32 of 226) had a local failure. The local failure rate by stage is 12.5% Tlb, 0% Tic, 11% T2a, 15% TZb, 18% T3, and 33% T4. Of local failures, 28% (9 of 32) were based on a positive biopsy and elevated PSA level with a normal DRE, and were thus “preclinical.” Actuarial local control is shown in Figure 2. Twenty-one percent of patients (47 of 226) remain in the biopsy failure category, with a normal PSA and DRE. This has occurred in 16% of Tlb (5 of 32), 45% of Tic (5 of ll), 20% of T2a (9 of 45), 17% of T2b (14 of 82), 20% of T3 (10 of 50),

/APRIL

1995 I VOLUME 45, NUMBER 4

to stage

LF + DF

KEY: NED = no evidence of disease; BF = biopsy failure; CF = chemicaljaihue; distant failure; L.F + DF = simultaneous local and distant failures. “Total local failures were 32 (14%). ‘Total distant failures were28 (12.5%).

UROLOGY@

according DF+

Total

1 0 2 6 8 1

32 11 45 82 50 6

18 8

226 100

LF = isolated local failure; DF = isolated

100 90 a0 70 60 %

50 40 30 20 10 0 0

I

I

I

I

I

I

I

I

6

12

18

24

30

36

42

48

Months

FIGURE 2.

Actuarial

local control

by stage.

12

18 24 TimepostRT (month)

FIGURE 3. Proliferative cell nuclear antigen (PCNA) counts (number of PCNApositive nuclei per 100 tumor nuclei) in three consecutive biopsies for 2 individual patients. (Modified from Crook et a1.24)

and 50% of T4 (3 of 6). This includes both positive and indeterminate biopsy results, regardless of PCNA status or apparent tumor viability. The median follow-up on these patients is only 19.5 months, and the median PSA level is 1.0 ng/mL. Only 30% (14 of 47) of the biopsy failure group has a PSA level of 1.5 ng/mL or greater. We classified the positive biopsies (clearly positive versus indeterminate) based on the degree of radiation effect, in order to predict eventual conversion to negative. Forty-five patients had an initial indeterminate biopsy. Of these, 14 have had only one biopsy, and 10 have continued to be classified as indeterminate after subsequent biopsies, for a total of 53% (24 of 45) who require further follow-up to determine outcome. Sixteen (36%) have resolved to negative at a mean time of 24 months after RT, and 5 (11%) have progressed to a clearly positive biopsy result at a mean of 20 months (4 with local failure). The mean PSA nadir for the study population was 2.0 ng/mL at a mean time of 19.3 months. Nadir PSA was higher in patients with local failure (mean, 3.9 ng/mL) than those without failure (mean, 0.7 ng/mL) (P
27.3 months, whereas those who progressed had a nadir of 4.7 ng/mL at 14.9 months. Staining for PCNA was attempted in all biopsies suspicious for residual tumor. For the 39 patients showing late conversion to negative, 27 (69%) of the initial post-treatment biopsy specimens were stained for PCNA and 18 of these (67%) were initially negative, indicating a loss of proliferative capability. The 9 patients who had initially PCNApositive residual tumor showed decreasing PCNA counts with subsequent biopsies (Fig. 3). All local failures that could be stained for PCNA (26 of 32) were PCNA positive, with a mean count of 13.7 cells per 100 tumor cells, confirming their biologic activity. Forty-two percent of patients (94 of 226) had hormonal treatment before RT. The effect on local outcome was examined as a function of duration of hormones. Negative biopsy rates at 18 months are 53% (56 of 106) for those who did not receive prior hormones and 74% for those treated for more than 4 months (P = 0.047). Patients with stage T2b or T3 disease showed no influence of prior hormone therapy in the rate of local failure. COMMENT Prostate cancer is a disease with a long natural history. Overall survival is not an appropriate measure of treatment efficacy in an aging population with multiple competing causes of mortality.25 Even disease-specific survival can be misleading, given the high risk of subclinical microscopic dissemination at the time of diagnosis. Efficacy of local therapy must be judged by local tumor eradication. Therefore, prostate biopsy should be the ultimate measure of treatment success. Unfortunately, the prolonged time to histologic tumor clearance and the microscopic changes caused by irradiation of nonmalignant glands can make interpretation of biopsy results difficult. Other series that may not have taken these factors into account have reported very high rates of residual tumor following RT.5-7 We have tried to eliminate selection bias by performing biopsies on all patients. Thirty-two patients (14%) have had local failures, 10 with concurrent distant metastases. Twenty-eight percent of these local failures were detected by the combination of a positive biopsy and a rising PSA level without abnormality on DRE. Thus, although our follow-up is relatively short, our criteria for local failure are strict, and failures are diagnosed earlier than they would be clinically. Surgical series reporting results for radical prostatectomy show that biochemical failure UROLOGY@ / APRK 1995 I VOWME 45, NUMBER 4

at 3 years is equivalent to clinical failure at 10 years26,27 and the same may be true for RT. The 7% false-negative rate (defined as patients with a negative post-treatment biopsy that later converts to positive) is comparable to other series9 and is likely due to sampling error. We have tried to minimize sampling error by using TRUS guidance, taking multiple samples, and preferentially targeting the original tumor site. Biopsy failures, by definition, have no supporting evidence for biologically active disease. Whether the biopsy shows apparently viable tumor or degenerated cells showing marked radiation change, continued resolution over time may occur. So far, only 18% of patients (21 of 115) with an initially positive post-treatment biopsy have developed local failure, whereas 34% (39 of 115) have demonstrated delayed conversion to negative at a mean of 26 months (range, 17 to 65 months). This rate of late tumor clearance agrees well with Scardino’s16 rate of 32% of positive 12month biopsy results becoming negative by 24 months. For patients not achieving complete tumor clearance, one can expect a continued toll of local failures. Schellhammer et al.1° have reported local failures developing in 25% of patients with a positive post-treatment biopsy by 3 years, 50% by 5 years, and 65% by 8 years. Histologic resolution correlates well with PSA nadir. Patients with local failure had a higher PSA nadir, 3.9 ng/mL, and reached the nadir earlier (14 months) than patients without failure (0.7 ng/mL at 23 months). These values are similar to those reported by Ritter et al. 28.. 2.9 ng/mL for patients with isolated local failure and 0.9 ng/mL for patients with no failure (Proschek assay: normal upper limit of 3.2 ng/mL). Our patients with delayed tumor clearance reached a nadir PSA level of 1.3 ng/mL at 27 months, an interval almost identical to the mean time for biopsy conversion. It remains unclear what constitutes a normal PSA after RT. Since the prostate has not been removed, one cannot expect the serum PSA level to become undetectable or to fall below 0.6 ng/mL.2T29 Most authors continue to define chemical failure as a PSA level above the upper limits of normal for their laboratory Levels associated with long-term disease-free survival are 1.1 to 1.2 ng/mL,13,28,29 although Stamey et al.ll reported stable values as high as 2.9 ng/mL in radiation-treated patients. In our series, the current median PSA in the patients with NED is 0.7 ng/mL. Forty-seven of our 226 patients (21%) are presently classified as biopsy failures. These patients have no clinical or PSA abnormalities. Al-

U ROLOCYO / APIUL 1995 I VOLUME 45, NUMBER 4

though some will undoubtedly fail, many of these biopsies may eventually convert to negative. The median follow-up of this group is only 19.5 months, which is short compared with the time course for tumor resolution, and the mean nadir PSA level is 1.0 ng/mL, which indicates an overall favorable prognosis. However, 14 of the 47 patients (30%) presently have a serum PSA over 1.5 ng/mL and this subgroup is at higher risk for eventual failure. PCNA was found to correlate with tumor viability after radiotherapy The absence of PCNA in residual tumor is predictive of eventual resolution, since no patient with negative PCNA had local progression. Unfortunately, the converse is not true, since PCNA positivity may decrease with time. RT causes postmitotic cell death, allowing fatally damaged prostate carcinoma cells to undergo a limited number of cell divisions.30,31 Since tumor doubling times are many months, residual proliferative activity may be seen up to 24 months after RT. All local failures showed high PCNA counts (mean, 13.7 per 100 tumor cells), indicative of retained proliferative potential. Other markers of cell proliferative capacity are available (Ki-67, MoAb, AgNor)31-33 and are currently being investigated. The correct interpretation of post-RT prostate biopsies requires a certain expertise. Pathologists should be encouraged to grade the therapy effect 34*35as this can be helpful in interpretation of the ‘report. Gleason grading should be avoided, since it was designed for untreated prostate cancer and may not be applicable to irradiated cancers where tumor gland morphology can be markedly altered. As with androgen deprivation therapy, invasive cell clusters or single cells can mimic a high Gleason grade.36 Gleason grading may, however, be appropriate in biopsies showing no apparent therapy effect. Staining for high molecular weight keratin is essential to avoid mistaking radiation atypia in benign glands for residual tumor. Such overcall may have contributed to uncertainty about the clinical relevance of positive biopsies,18 since patients with radiation atypia alone are unlikely to develop distant metastases or die of prostate cancer. There is no doubt that true residual tumor of proven biologic activity (abnormal DRE or rising PSA levels) will eventually progress and disseminate. Prestidge et a1.37 report that the most important predictive factor for the outcome of a positive post-radiation prostate biopsy is the DRE. In their series, 52% of patients with an abnormal DRE developed distant metastases compared with 26% of patients with no palpable abnormality. At

629

a median follow-up of 13.8 years, 50 of 116 patients (43%) are alive with no evidence of disease other than a positive prostate biopsy result. PSA levels for these patients were not reported. Contrary to the experience of other authors,12 rapidly rising PSA values in our population of irradiated patients were seen only in those with disseminated disease. Local failures had a mean PSA doubling time of 8.1 months compared with 3.4 months for patients with distant failure. Zagars amd Pollack38 reported mean doubling times of 6 months for metastatic disease and 10.8 months for locoregional failure. Although PSA doubling times appear to be related more to grade than to site of failure,25 higher grade tumors have a higher metastatic potential. This may account for the consistently shorter doubling times for patients with metastases rather than isolated local failure. Cox et ~1.~~ have proposed an alternative theory that, since the more rapidly dividing cells are more radiosensitive, radiation may select more slowly proliferating clones, accounting for the slower rise in PSA levels with local failure. There was a higher rate of negative biopsies at 18 months in patients who received more than 4 months of prior hormonal therapy (75%) compared with no hormonal treatment (53%) (P = 0.047). However, there was no corresponding difference in local failure rates. This should not be interpreted as a failure of cytoreductive therapy Patients receiving hormones before RT were selected for many reasons and referred for RT at times often unrelated to optimal cytoreduction. Furthermore, the hormonal therapy was not uniform. No conclusions should be drawn when such heterogeneity exists. Prospective randomized trials of hormonal cytoreduction are presently under way We do not recommend routine post-treatment biopsies in the follow-up of prostate cancer after RT. However, TRUS-guided biopsies may be useful if further local treatment is considered for local failure. A positive biopsy result after 30 months, especially one showing PCNA positivity, likely represents true local treatment failure. Although our follow-up is short, we recommend that treatment decisions should not be based on a positive biopsy result alone, in the absence of either a palpable abnormality or a rising PSA level. Currently, efforts are under way to reduce the failure rate. Fiducial markers inserted under ultrasound guidance can be used to track prostate motion during RT and ensure optimal centering of the radiation beam throughout therapy, reducing the risk of geographic miss. Conformal therapy may allow dose escalation without increasing tox-

630

icity. Finally, cytoreduction prior to radiotherapy may reduce the number of clonogenic cells and improve the dose/response relationship. CONCLUSION Prostate biopsies following radical radiotherapy may take 2 l/2 to 3 years to resolve. Staining for proliferative cell nuclear antigen can help differentiate viable from nonviable tumor cells. Although longer follow-up is required, it appears that an abnormal prostate biopsy without other evidence of disease is insufficient to make the diagnosis of local failure. J. M. Crook, M.D. Department of Radiation Oncology Ottawa Regional Cancer Centre Ottawa, Canada ACKNOWLEDGMENT. To Carolle Brazeau for secretarial assistance, and to Dr. V. Zaleski for his expertise in ultrasonography and the Ottawa Regional urologists for referral of patients, especially Dr. N. Futter, Dr. D. McKay, Dr. W. Walsh (Ottawa General Hospital), Dr. G. Chenard (Riverside Hospital), Dr. N. Saliba and Dr. S. Faddoul (Centre hospitalier de Gatineau), Dr. G. Bourdeau and Dr. C. Lajeunesse (Montfort Hospital).

REFERENCES 1. Hanks GE: External beam radiation treatment for prostate cancer: still the gold standard. Oncology 6: 79-86, 1992. 2. Perez CA, Hanks GE, Leibel SA, Zietman AL, Fuks Z, and Lee WR: Localized carcinoma of the prostate (Stages Tlb, Tic, T2, and T3). Review of management with external beam radiation therapy. Cancer 72: 31563173, 1993. 3. Bagshaw MA, Kaplan ID, and Cox RC: Radiation therapy for localized disease. Cancer 71: 939-952, 1993. 4. Dugan TC, Shipley WU, Young RH, Verhey LJ, Althausen AF, Heney NM, McManus PL, and Abraham EH: Biopsy after external beam radiation therapy for adenocarcinoma of the prostate: correlation with original histological grade and current prostate specific antigen levels. J Urol 146: 1313-1316, 1991. 5. Kabalin JN, Hodge KK, McNeal JE, Freiha FS, and Stamey TA: Identification of residual cancer in the prostate following radiation therapy: role of transrectal ultrasound guided biopsy and prostate specific antigen. J Urol 142: 326-331, 1989. 6. Scardino PT, and Wheeler TM: Prostatic biopsy after irradiation therapy for prostatic cancer. Urology 25: 39-46, 1985. 7. Nachtsheim DA Jr, McAninch JW, Stutzman RE, and Goebel JL: Latent residual tumor following external radiotherapy for prostate adenocarcinoma. J Urol 120: 312-314, 1978. 8. Fuks Z, Leibel SA, Wallner KE, Begg CB, Fair WR, Anderson LL, Hilaris BS, and Whitmore WF: The effect of local control on metastatic dissemination in carcinoma of the prostate: long-term results in patients treated with lz51 implantation. Int J Radiat Oncol Biol Phys 21: 537-547, 1991. 9. Kuban DA, and El-Mahdi AM: Local control after radiation for prostatic carcinoma: significance and assessment. Semin Radiat Oncol 3: 221-229, 1993.

U ROLOCY@ /APRIL 1995 I VOLUME 45, NUMBER 4

10. Schellhammer PF, el-Mahdi AM, Higgins EM, Schultheiss TE, Ladaga LE, and Babb TJ: Prostate biopsy after definitive treatment by interstitial ‘25iodine implant or external beam radiation therapy. J Urol 137: 897-901, 1987. 11. Stamey TA, Kabalin JN, and Ferrari M: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. III. Radiation treated patients. J Urol 141: 1084-1087, 1989. 12. Stamey TA, Ferrari MK, and Schmid HP: The value of serial prostate specific antigen determinations 5 years after radiotherapy: steeply increasing values characterize 80% of patients. J Urol 150: 1856-1859, 1993. 13. Zagars GK, vonEschenbach AC, Ayala AG, Schultheiss TE, and Sherman NE: The influence of local control on metastatic dissemination of prostate cancer treated by external beam megavoltage radiation therapy. Cancer 68: 2370-2377, 1991. 14. Cox JD, and Kline RW: The lack of prognostic significance of biopsies after radiotherapy for prostatic cancer. Semin Urol 1: 237-242, 1983. 15. Herr HW, and Whitmore WF Jr: Significance of prostatic biopsies after radiation therapy for carcinoma of the prostate. Prostate 3: 339-350, 1982. 16. Scardino PT: The prognostic significance of biopsies after radiotherapy for prostatic cancer. Semin Urol 1: 243-252,1983. 17. Bostwick

DG, Egbert BM, and Fajardo LF: Radiation injury of the normal and neoplastic prostate. Am J Surg Path01 6: 541-551, 1982. 18. Brawer MK, Nagle RB, Pitts W, Freiha F, and Gamble SL: Keratin immunoreactivity as an aid to the diagnosis of persistent adenocarcinoma in irradiated human prostates. Cancer 63: 454-460, 1989. 19. Carroll PR, Waldman FM, Rosenau W, Cohen MB, Vapnek JM, Fong P, Narayan P, and Maya11 BH: Cell proliferation in prostatic adenocarcinoma: in vitro measurement by S-bromodeoxyuridine incorporation and proliferating cell nuclear antigen expression. J Urol 149: 403-407, 1993. 20. Nemoto R, Kawamura H, Miyakawa I, Uchida K, Hattori K, Koiso K, and Harada M: Immunohistochemical detection of proliferating cell nuclear antigen (PCNA)/cyclin in human prostate adenocarcinoma. J Urol149: 165-169,1993. 21. Hermanek P, and Sobin LH (Eds): International Union Against Cancer. TNM Classification ofMalignant Turnours, 4th ed., 2nd rev. Berlin, Springer-Verlag, 1992. 22. Crook J, Robertson S, Collin G, Zaleski V, and Esche B: Clinical relevance of trans-rectal ultrasound, biopsy, and serum prostate-specific antigen following external beam radiotherapy for carcinoma of the prostate. Int J Radiat Oncol Biol Phys 27: 31-37, 1993. 23. van der Werf-Messing B: Prostatic cancer treated at Rotterdam Radiotherapy Institute. Strahlentherapie 154: 537-541, 1978. 24. Crook J, Robertson S, and Esche B: Proliferative cell nuclear antigen in post-radiotherapy prostate biopsies. Int J Radiat Oncol Biol Phys 30: 303-308, 1994. 25. Hanks GE: Treatment of early stage prostate cancer: radiotherapy, in Devita VT, Hellman S, and Rosenberg JB (Eds): Important Advances in Oncology, 1994. Philadelphia, Lippincott, 1994. 26. Fair WR, Aprikian AG, Cohen D, Sogani P, and Reuter V: Use of neoadjuvant androgen deprivation therapy in clinically localized prostate cancer. Clin Invest Med 16: 516-522, 1993. 27. Paulson DF, Moul JW, and Walther PJ: Radical prostatectomy for clinical stage Tl-ZNOMO prostatic adenocarcinoma: long-term results. J Urol 144: 1180-1184, 1990.

U ROLOCY@ / APRIL 1995 / VOLUME 45,

NUMBER

4

28. Ritter MA, Messing EM, Shanahan TG, Potts S, Chappell RJ, and Kinsella TJ: Prostate-specific antigen as a predictor of radiotherapy response and patterns of failure in localized prostate cancer. J Clin Oncol 10: 1208-1217, 1992. 29. Russell KJ, and Boileau MA: Current status of prostate-specific antigen in the radiotherapeutic management of prostatic cancer. Semin Radiat Oncol 3: 154-168, 1993. 30. Mostofi FK, Sesterhenn IA, and Davis CJ Jr: A pathologist’s view of prostatic carcinoma. Cancer 71 (suppl): 906-932,1993. 31. Mostofi

FK, Davis CJ Jr, and Sesterhenn IA: Pathology of carcinoma of the prostate. Cancer 70 (suppl): 235-253, 1992. 32. Harper ME, Glynne-Jones E, Goddard L, Wilson DW, Matenhelia SS, Conn IG, Peeling WB, and Griffiths K: Relationship of proliferating cell nuclear antigen (PCNA) in prostatic carcinomas to various clinical parameters. Prostate 20: 243-253,1992. 33. Sakr

WA, Sarkar FH, Sreepathi P, Drozdowicz S, and Crissman JD: Measurement of cellular proliferation in human prostate by AgNOR, PCNA, and SPF. Prostate 22: 147-154, 1993. 34. Dhom G, and Degro S: Therapy of prostatic cancer and histopathologic follow-up. Prostate 3: 531-542, 1982. 35. B&king A, and Auffermann W: Cytological grading of therapy-induced tumor regression in prostatic carcinoma: proposal of a new system. Diagn Cytopathol 3: 108-111, 1987. 36. Armas OA, Aprikian AG, Melamed J, Cordon-Card0 C, Cohen DW, Erlandson R, Fair WR, and Reuter VE: Clinical and pathobiological effects of neoadjuvant total androgen ablation therapy on clinically localized prostatic adenocarcinoma. Am J Surg Path01 18: 979-991, 1994. 37. Prestidge BR, Kaplan I, Cox RS, and Bagshaw MA: Predictors of survival after a positive post-irradiation prostate biopsy. Int J Radiat Oncol Biol Phys 28: 17-22, 1994. 38. Zagars GK, and Pollack A: The fall and rise of prostate-specific antigen. Kinetics of serum prostate-specific antigen levels after radiation therapy for prostate cancer. Cancer 72: 832-842,1993. 39. Cox RS,

Kaplan ID, and Bagshaw MA: Prostate-specific antigen kinetics after external beam irradiation for carcinoma of the prostate. Int J Radiat Oncol Biol Phys 28: 23-31, 1994. EDITORIAL COMMENT This study reports the results of a remarkable 6-year effort by the authors to define the time course of histologic resolution of prostate cancer after radiotherapy. Of great importance is that it was conducted prospectively in consecutive patients. Seventy percent of 226 patients achieved negative biopsy status by 30 months follow-up and the conversion from early positivity to later negative biopsies was clearly shown. Importantly, prostate-specific antigen (PSA) response predicted biopsy outcome but conversion to late negative status was frequently accompanied by a later PSA response in those patients. Rapidly rising PSAs after radiation failure were observed only in a small group of patients with documented metastatic disease. These data have established the new standard that extends and more accurately defines previous reports concerning the accuracy of postirradiation biopsies by Cox et al.’ and Scardino et nl.* They also show, along with other recent articles,3-5 that the previous statement by Kabalin et al.,‘j “Residual prostate cancer was proven by biopsy in 25 of 27 patients

631

(93%) 18 months to 12 years after completion of radiation therapy” and Stamey et al7 “...approximately 20% of the patients...can be cured by irradiation therapy. The remaining 80% of the patients in whom radiotherapy fails appear to have an accelerated growth rate...” were observations made in highly selected patient groups that do not represent the results obtained in unselected radiotherapy patients. Unfortunately, their comments have been interpreted by many physicians as representing the results of radiation therapy as a whole. Dr. Crook and her colleagues have reminded us that there is a role for institution-based clinical research where carefully planned and conducted clinical studies in large groups of consecutive patients can provide valuable information that can help to guide us in patient management. Gerald E. Hanks, M.D. Fox Chase Cancer Center 7701 Burholme Avenue Philadelphia, PA I91 I I REFERENCES 1. Cox JD, and Stoffel TJ: The significance of needle biopsy after irradiation for the stage C adenocarcinoma of the prostate. Cancer 40: 156-160, 1977.

632

2. Scardino PT: The prognostic significance of biopsies after radiotherapy for prostatic cancer. Semin Urol 1: 243-251, 1985. 3. Hancock SL, Cox RS, and Bagshaw MA: Biochemical control of prostate cancer and kinetics of prostate specific antigen in recurrences in a cohort of patients treated by external beam irradiation. Int J Radiat Oncol Biol Phys 30: 217, 1994. 4. Hanks GE, Lee WR, Corn B, and Schultheiss TE: Response to “only 20% of irradiated prostate cancer patients are cured and the remainder may be made worse”. Int J Radiat Oncol Biol Phys 30: 217, 1994. 5. Zietman AL, Shipley WU, Coen JJ, Shipley JW, and Althausen AF: Long-term outcome for men with prostate cancer treated with radiation as sole therapy. Int J Radiat Oncol Biol Phys 30: 218, 1994. 6. Kabalin JN, Hodge KK, McNeal JE, Freiha FS, and Stamey TA: Identification of residual cancer in the prostate following radiation therapy: role of transrectal ultrasound guided biopsy and prostate specific antigen. J Urol 142: 326-331, 1989. 7. Stamey TA, Ferrari MK, and Schmid H: The value of serial prostate specific antigen determinations 5 years after radiotherapy: steeply increasing values characterize 80% of patients. J Urol 150: 1856-1859, 1993.

UROLOGY@ /APRIL 1995 I VOLUME 45, NUMBER 4