REPEAT SCREENING FOR PROSTATE CANCER AFTER 1-YEAR FOLLOWUP IN 984 BIOPSIED MEN: CLINICAL AND PATHOLOGICAL FEATURES OF DETECTED CANCER

0022-5347/98/1606-2121$03.00/0

THE JOURNALOF UROLCGY Copyright 0 1998 by AMERICAN UROLOGICAL ASSOCIATION, INC.

Vol. 160, 2121-2125, December 1998 Printed in U . S A

REPEAT SCREENING FOR PROSTATE CANCER AFTER 1-YEAR FOLLOWUP IN 984 BIOPSIED MEN: CLINICAL AND PATHOLOGICAL FEATURES OF DETECTED CANCER JOHN B. W. RIETBERGEN, ART0 E. BOEKEN KRUGER, ROBERT F. HOEDEMAEKER, CHRIS H. BANGMA, WIM J. KIRKELS AND FRITZ H. SCHRODER From the Departments of Urology and Pathology, Erasmw University and Academic Hospital, Rotterdam, The Netherlands

ABSTRACT

Purpose: We describe the yield of a repeat examination and biopsy procedure 1year after initial biopsy was negative. We also assessed the parameters responsible for the failure to diagnose these cancers at the primary screening. Materials and Methods: We screened 8,103 men randomized to the screening arm of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer using prostate specific antigen measurement, digital rectal examination and transrectal ultrasound. At the primary screening biopsy of 1,875 men was positive for prostate cancer in 374. Of the remaining 1,501 men 984 underwent repeat screening. Results: Biopsy at repeat screening diagnosed prostate cancer in 49 of 442 men (ll%), a rate significantly lower than the 19.9% true positive biopsy rate at the primary screening. Pathological characteristics of the tumors diagnosed were not significantly different in the 2 groups. However, prostate volume in men diagnosed with prostate cancer was significantly greater at repeat versus primary screening (mean 42.6 versus 34.9 cc, p = 0.003). The clinical characteristics were more favorable because of a n increased proportion of stage T1C tumors. Prostate volume in men with stage T1C cancer was significantly greater than in those with palpable or visible tumors in whom prostate specific antigen values were in the same range. Conclusions: The most important factor responsible for the failure to diagnose these cancers at the primary screening was significantly greater prostate volume. Tumor characteristics were not significantly different in the groups. If prostate cancer screening were to become a routine health care policy, efforts would have to be made to improve the chances of diagnosing prostate cancer in larger prostates by repeat biopsy or by increasing the number of cores obtained. KEYWORDS:prostate, prostatic neoplasms, prostate-specific antigen Although it is controversial due to a lack of evidence in randomized screening studies, secondary prevention of prostate cancer through early detection and treatment is widespread. For detecting prostate cancer in asymptomatic cases serum prostate specific antigen (PSA) determination, digital rectal examination and transrectal ultrasound have led to an increasing number of prostate biopsies and clinically diagnosed cases of prostate cancer. However, biopsy results are negative in the majority of men who undergo biopsy because of abnormal test results. It is well recognized that cases of prostate cancer are missed at biopsy.'.' The introductionby Hodge et a1 of the practice of obtaining 6 random systematic core biopsies3 has increased accuracy when diagnosing prostate cancer. Recently Eskew et a1 described the method of systematic 5 region prostate biopsy, which may further increase the diagnostic yield.4 Others have recommended repeat biopsy for such specific indications as persistently elevated serum PSA5 and abnormal histological findings in the initial biopsy specimen6-*to decrease the proportion of prostate cancer missed by initial biopsy. In this context it is interesting to know the yield of a selective repeat examination procedure and whether there

are any parameters that may explain why cancer diagnosed at repeat screening was not diagnosed by the primary biopsy. We describe prostate cancer detection in relation to such clinical features as PSA, prostate volume, and the clinical and pathological characteristics of cancer detected after repeat examination of 984 men in whom biopsy results were negative on the initial population based screening study. Evidently our study comprises a highly select group of men and results are not applicable to the issue of repeat screening in a population based setting. METHODS

Population. Between October 1991 and July 1996, 17,006 men 55 to 76 years old responded to a letter of invitation to be entered in the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer. The goal of this large general population based study is to randomize 190,000 men, including 40,000 in Rotterdam, The Netherlands. The only exclusion criterion in the study was a previous diagnosis of prostate cancer. Written informed consent was obtained from all study participants. Of those who responded 8,593 and 8,413 were randomized to the screening and control arms, respectively. Men enrolled in the control arm were not Accepted for publication June 12, 1998. Supported by grants from The Dutch Cancer Foundation, The offered any testing. Techniques. Of the 8,593 men in the screening arm 8,013 hevention Fund, Hybritech, Inc. and Europe Against Cancer. underwent determination of serum PSA concentration using Editor's Note: This article is the fourth of 5 publishedin this issue for which category 1 CME credits can be earned. In- the Hybritech Tandem-E* PSA immunoenzymetric assay,

structions for obtaining credits are given with the questions on pages 2206 and 2207. 2121

* Hybritech, Inc., San Diego, California.

2122

REPEAT SCREENING FOR PROSTATE CANCER

detection and 19.9%true positive biopsy rates). At PSA levels of less than 4 ng./ml. 100 cases of prostate cancer were diagnosed from 943 systematic sextant biopsies (10.6% true examination and transrectal ultrasound the screening team positive biopsy rate). At PSA 4 ng./ml. or greater 932 biopsies was blinded to PSA results. Prostate nodularity, induration resulted in the diagnosis of 274 cases of prostate cancer and asymmetry were considered abnormal digital rectal ex- (29.3% true positive biopsy rate). Repeat screening. Population: A total of 1,501men in whom amination results. Biplanar transrectal ultrasonography was performed using a Bruel & Kjaer model 1846 mainframe and initial biopsy results were negative were invited to undergo a 7 MHz. endorectal transducer. We used the sonographic repeat screening 1 year later. To date screening of 984 men criteria for prostate cancer described by Lee et al.' Prostate has resulted in 498 indications for biopsy. The remaining 486 volume was obtained by planimetric measurement using a men did not undergo biopsy due to abnormalities on digital 0.5 cm. step section technique. Beginning in October 1995 rectal examination and/or transrectal ultrasound that could digital rectal examination and transrectal ultrasound were not be reproduced in 280, serum PSA less than 1 ng./ml. in 136,a decrease in PSA to less than 4 ng./ml., normal digital not performed in men with PSA below 1 ng./ml.lo Transrectal ultrasound guided systematic sextant biopsy rectal examination and transrectal ultrasound in 60, and was performed by a resident urologist using a Pro-mag 2.2* withdrawal from the study in 10. Cancer Detection ARer Repeat Examination in 984 Men biopsy gun and an 18 gauge biopsy needle. Hypoechoic lesions were also sampled. In cases of high grade prostatic Who Underwent Biopsy (table 1): In 632 men PSA was less intraepithelial neoplasia or atypia biopsy was repeated than 4 ng./ml. Of the 146 men in whom digital rectal examwithin 3 months. Men in whom results of the primary or ination and/or transrectal ultrasound was abnormal 123 unrepeat biopsy were negative were invited to undergo repeat derwent biopsy and 12 were diagnosed with prostate cancer examination after 1 year, according to the screening algo- (9.8% true positive biopsy rate). Of the 352 men in whom PSA rithm. The screening team was blinded to the results of the was 4 ng./ml. or greater 319 underwent biopsy and 37 were primary screening diagnosed with prostate cancer (11.6% true positive biopsy Pathological grade and clinical stage. Tumor volume of the rate). Overall 49 cases of prostate cancer were diagnosed in radical prostatectomy specimens processed at our hospital 442 biopsies (11% true positive biopsy rate). was determined using a W O O Kontront digital morphometComparison of the primary and repeat screening uisits. ric analysis imaging system. Tumor grading was done ac- Cancer Detection: The true positive biopsy rate (percentage cording to the M. D. Anderson grading system because patho- of cancer cases detected per number of biopsies done) was not logical evaluation of the radical prostatectomy specimen was significantly different in men in whom PSA was less than 4 performed at the hospital in which the patient was treated, ng./ml. and who underwent biopsy at primary versus repeat and not all of those institutions use the Gleason score. The screening (10.6 versus 9.8%, p = 0.8, table 1). In men in 1992 TNM classification was used to describe clinical stage." whom PSA was 4 ng./ml. or greater the difference was sigTumor stage was determined at the time of screening, while nificant (29.4versus 11.6%, p
digital rectal examination and transrectal ultrasound. The PSA test cutoff point was 4.0 ng./ml. and any value greater than 3.9 ng./ml. was considered elevated. At digital rectal

TABLE1. Number of men screened, biopsied and diagnosed with prostate cancer, and true positive biopsy indications, as determined by screening and repeat screening PSA Less Than 4 NgJM1. Primary Screening

No. pts. screened No. pts. biopsied (5%) No.pta. with prostate Ca (%) % h e pos. biopsy rate

6,949 943 (13.6) 100 (1.43) 10.6

Repeat Screening

632 123 (19.5) 12 (1.9) 9.8

PSA 4 Ng./Ml. or Greater

Primary Screening 1,064 932 (87.6) 274 (25.8) 29.4

Repeat Screening

352 319 (90.6) 37 (10.5) 11.6

2123

REPEAT SCREENING FOR PROSTATE CANCER +

0

benign

0

12

6

while in the repeat screening group it was 46.6 cc (median 41.5,range 22.7 to 116.4).However, PSA values were not significantly different (p = 0.4). Mean PSA at primary screening was 6.5 ng./ml. (median 5.6,range 0.4 to 31.81, while at repeat screening it was 6.2 ngjml.(median 6.2,range 1.0 to 13.2). Tumor volume was known in 71 men in whom prostate cancer was diagnosed at PSA 4 ngJml. or greater and who were treated with radical prostatectomy at our institution. The diagnosis was made at primary and repeat screeningin 60 and 11 cases, respectively, and tumor volume did not differ significantly (p = 0.14).Mean tumor volume in men diagnosed at primary and repeat screening was 1.6 cc (median 1.0,range 0.1 to 13.5)and 0.8 cc (median 0.7,range 0.04 to 1.91,respectively. There was no significant difference in PSA between the 2 groups (p = 0.87).

cancer

18

24

30

DISCUSSION

PSA primary usit

The frequency of screening and subsequent performance of biopsy procedures is a trade-off between unnecessary testing and the risk of missing potentially curable cancer that may be a threat to the well-being of the host within his life span. Recently Stroumbakis et a1 showed that repeat biopsy was primary screening, mean prostate volume was 48.3cc (medi- negative in 20% of 89 men with biopsy proved prostate canan 42.6,range 21.1to 116.4)in those diagnosed with prostate cer.’ In a study by Keetch et al repeat biopsy was done cancer at repeat screening. This difference was statistically because of a persistent elevated PSA level combined with significant as well (p = 0.003).In the 19 men diagnosed at abnormal digital rectal examination and/or transrectal ultrarepeat screening with clinical stage T1C prostate cancer sound findings in men who initially underwent biopsy for the mean prostate volume was 58.9 cc (median 53.2,range 28.2 same reasons (19% true positive biopsy rate).6In our series to 116.7),while it was 46.2 cc (median 36.4,range 21.0 to PSA of 4 ng./ml. or greater was an indication for biopsy 121.9)in men with clinical stages T2 and T3 prostate cancer regardless of the digital rectal examination and transrectal in whom PSA was 4 ng./ml. or greater. This difference was ultrasound findings, resulting in an 11.6% true positive biopsy rate. Comparable to the findings of Keetch et al, in our statistically significant (p = 0.018 Mann-Whitney U test). Clinical Stage and Grade Distribution (table 2): At the study the true biopsy rate was 15.7% in men who underwent primary screening 77 of the 374 men (20.6%)diagnosed with biopsy above the PSA threshold with abnormal digital rectal prostate cancer did not have organ confined disease, includ- examination and/or transrectal ultrasound findings. We found no significant difference in true positive biopsy ing 7 with bone (4)or lymph node metastasis (3).In 2 of the 49 cases (4%) of prostate cancer diagnosed at repeat screen- rates at PSA levels below 4 ng./ml. in the primary and repeat ing disease was not confined to the prostate (stage T3 dis- screening groups (table 1).A considerable number of men did ease). No metastases were noted. The difference in clinical not undergo biopsy at repeat screening since abnormal findstage distribution for organ confined versus advanced dis- ings on digital rectal examination and transrectal ultrasound ease was statistically significant (p = 0.005).Prostate cancer at the primary screening could not be reproduced (table 1). diagnosed at the primary screening was well differentiated in The mechanisms that may have had a role were high inter54% of the cases, whereas at repeat screening disease was examiner variability during digital rectal examination,12 well differentiated in 69%. However, this difference was not which was most important, and the fact that reversible conditions, such as prostatitis, may cause abnormal digital recstatistically significant (p = 0.06). Patients Treated With Radical Prostatectomy (table 3):Of tal examination and transrectal ultrasound findings. Howthe primary and repeat screening groups 149 (40%) and 20 ever, the selection of men in whom digital rectal examination (41%) men were treated with radical prostatectomy, respec- and transrectal ultrasound findings are abnormal seems to tively. Table 3 shows pathological stage and grade distribu- maintain the true positive biopsy rate at repeat screening at tions of the radical prostatectomy specimens in men diag- almost the same level as at primary screening. PSA distribution did not differ significantly in patients diagnosed at the primary and repeat screenings. Stage distribution was similar in each group treated with radical nosed at primary and repeat screening (p = 0.38).As measured prostatectomy (approximately 40% locally advanced cancer, initially, mean prostate volume in men with prostate cancer p = 0.64).In the primary screening group 47% of the cases was significantly greater in the repeat than in the primary and comprised well differentiated disease, while in the repeat screening group (48.3versus 40.2 cc, p = 0.003).UZZO’~ et al reported that prostate size as well as tumor screening group 55% were well differentiated. However, this Dane~hgari’~ difference was not statistically significant (p = 0.5).Mean characteristics may influence cancer detedion. In the subset of prostate volume in men treated with radical prostatectomy men treated with radical prostatectomy in our series the pathodiffered significantly in the primary and repeat screening logical characteristics of tumors diagnosed at primary and regroups (p = 0.038).Mean prostate volume in the primary peat screening were identical. In each group capsular penetrascreening group was 37.2cc (median 32.8,range 15.4to 99.7), tion was present in approximately 40% of the patients. While

PSA at primary versus repeat screening in 4.42 men who underwent biopsy. Line represents line of equality of PSA measured at primary and repeat screenings.

TABLE2. Clinical stage and grade distribution in patients with prostate cancer diagnosed during primary and repeat screening No. Grade (%)

No. Stage (%) TxMO

TlMO

T2MO

T3MO

T4MO

$:

L i

Totals

X

1

2

3

Totals

F’rimary screening 2 (0.5) 94 (25.1) 201 (53.7) 68 (18.2) 2 (0.5) 4 (1.1) 3 (0.8) 374 (100) 2 (0.5) 203 (51.3) 131 (35) 38 (10.2) 1.122 (100) 147(100) 34 (69) 14(29) 1 (2) 0 49(100) 0 0 lg(38.8) 28(57.1) 2 (4.1) 0 Repeat SCTeening 0

2124

REPEAT SCREENING FOR PROSTATE CANCER TABLE3. Radical pmstotectomy pathological stage and grade distribution, as determined by primary and repeat screening No. Grade (96) No. Staae (96)

tumor volume did not differ significantly (p = 0.14), there was a trend toward smaller tumor volume at repeat screening, which may not have been significant due to the low number of men in that group. PSA distribution did not differ significantly, although median PSA in the repeat screening group was slightly greater. However, prostate volume was significantly greater in the repeat screening group (p = 0.003), which must have decreased the chance of making the diagnosis at the primary screening. This finding confirms that of Eskew et al, who showed that a volume based biopsy pattern enhances the yield of biopsy in terms of number of prostate cancer cases diagnosed! Despite the identical distribution of pathological stage in the 2 groups the distribution of clinical stage in the men with prostate cancer diagnosed at repeat screening was significantly more favorable than in the primary screening group. The percentage of stage T1C disease nearly doubled and only 4% of the men had clinical signs of extracapsular growth. Oe~terling'~ and Epstein" et a1 independently diagnosed stage T1C prostate cancer in prostates of significantly higher volume, which may explain why these tumors are impalpable. Our findings at repeat screening confirmed their observations. Prostates harboring stage T1C tumors were significantly larger than those harboring stage T2 or T3 disease (median 53.2 versus 36.4 cc) when PSA was 4 ng./ml. or greater. Half of the 20 men treated with radical prostatectomy had stage T1C cancer, including 4 in whom it was pathologically organ confined. Nine men had stage T2 disease, including 7 with organ confined disease. One patient had clinical stage T3A disease, which was reflective of the radical prostatectomy specimen. The fact that higher prostate volume leads to the underestimation of extracapsular disease in stage T1C prostate cancer and the percentage of stage T1C disease nearly doubled due to the greater prostate volumes in the repeat screening group resulted in a more favorable clinical than pathological stage distribution. An additional explanation for the diagnosis of prostate cancer in men in whom biopsy was negative at the primary screening may be the presence of small but rapidly growing tumors, which would enhance the detection chance at repeat screening. Carter et a1 reported that serial measurement indicated that the rate of serum PSA increase was higher in patients with prostate cancer than in those with benign prostate condition^.'^ They suggested a PSA velocity cutoff point of 0.75 ng./ml. In our series of 442 men who underwent repeat biopsy the mean increase in PSA in those with prostate cancer versus those with negative biopsy results was 0.8 versus 0.6 ng./ml. However, this trend was not statistically significant (p = 0.083). An increase of 0.75 ng./ml. or more was noted in 34.3% of the men with a negative biopsy and in 44.9% of those with prostate cancer. Again, this difference was not significant (p = 0.14). The figure shows this overlap in a scatterplot. Pearson and Carter demonstrated that cancer cases showed an early linear phase of PSA increase followed by an exponential phase." Since the increase in PSA in men with and without prostate cancer did not differ significantly, it may safely be assumed that the diagnosis at repeat screening was not made possible by tumors with a short PSA doubling time. As suggested by Carter et al, a longer observation period of 2 years may be necessary to study the value of PSA velocitv in this situation.17 ~~.~ ...... ~

~~~

CONCLUSIONS

Significantly greater prostate volume seems to have a key role in the diagnosis of prostate cancer a t repeat screening as well as in the increase of the proportion of stage T1C disease. Our results indicate that the chance of diagnosing prostate cancer in men with a large prostate is decreased. There was a similar pathological stage distribution in patients in the primary and repeat screening groups who were treated with radical prostatectomy. While tumor volume did not differ significantly between the groups, a t this time it cannot be excluded that the trend toward smaller tumors in the repeat screening group will become significant with increasing numbers of cases. If early detection and treatment decrease prostate cancer specific mortality, the diagnosis of cancer in larger prostates should be improved by repeat biopsy or an increase in the number of cores in the initial biopsy in men with a high prostate volume, since the characteristics are often unfavorable in terms of the proportion of extracapsular tumors. Mr. G. J. de Zwart, Mrs. M. Roobol-Bouts, Ms. W. Roobol, Mrs. A. Smit,Mrs. X van der DriRTetteroo and Mrs. C F'ranken-Raab provided technical support, and Dr. B. Blijenberg and h b . I. Eman provided clinical laboratory support. REFERENCES

1. Stroumbakis, N., Cookson, M. S., Reuter, V. E. and Fair, W. R.: Clinical sigmficance of repeat sextant biopsies in prostate cancer patients. Urology, suppl. 3A,4 9 113, 1997. 2. Coplen, D. E., Andriole, G. L., Yuan, J . J., Jr. and Catalona, W. J.:The ability of systematic transrectal ultrasound guided biopsy to detect prostate cancer in men with the clinical diagnosis of benign prostatic hyperplasia. J. Urol., 146 75,1991. 3. Hodge, K. K, McNeal, J. E., Terris, M. K. and Stamey, T. A,: Random systematic versus directed ultrasound guided core biopsies of the prostate. J. Urol., 142 74,1989. 4. Eskew, L. A.,Bare, R. L. and McCullough, D. L.: Systematic 5 region prostate biopsy is superior to sextant method for diagnosing carcinoma of the prostate. J. Urol., 157: 199,1997. 5. Keetch, D. W.,Catalona, W. J . and Smith, D. S.: Serial prostatic biopsies in men with persistently elevated serum prostate specific antigen values. J. Urol., 151: 1571, 1994. 6. Ellis, W. J. and Brawer, M. K.: Repeat prostate needle biopsy: who needs it? J . Urol., 153 1496,1995. 7. Wills, M. L.,Hamper, U. M., Partin, A. W. and Epstein, J . I.: Incidence of high-grade prostatic intraepithelial neoplasia in sextant needle biopsy specimens. Urology, 4 9 367,1997. 8. Bostwick, D. G., Qian, J . and Frankel, R:The incidence of high grade prostatic intraepithelial neoplasia in needle biopsies. J. Urol., 1W.1791,1995. 9. Lee, F., Gray, J. M., McLeary, R. D., Meadows, T. R., Kumasaka, G. H., Borlaza, G. S., Straub, W. H., Lee, F., Jr., Solomon, M. H. and McHugh, T. A.:Transrectal ultrasound in the diagnosis of prostate cancer: location, echogenicity,histopathology, and staging. Prostate, 7:117, 1985. 10. Rietbergen, J. B., Kranse, R., Kirkels, W. J., De Koning, H. J . and Schroder, F. H.: Evaluation of prostate-specific antigen, digital rectal examination and transrectal ultrasonography in population-based screening for prostate cancer: improving the efficiency of early detection. Brit. J . Urol., suppl. 2, 79: 57, 1997. 11. Schroder, F. H., Hennanek, P., Denis. L.. Fair. W. R., Gospodarowicz, M. K. and Pavone-Macaluso, M.: The TNM classification of prostate cancer. Prostate, suppl., 4: 129, 1992. 12. Smith, D. S. and Catalona, W. J.: Interexaminer variability of

REPEAT SCREENING FOIR PROSTATE CANCER digital rectal examination in detecting prostate cancer. Urology, 4 5 70, 1995.

13. Uzzo, R. G., Wei, J. T., Waldbaum, R. S., Perlmutter, A. P., Byme, J. C. and Vaughan, E. D., Jr.: The influence of prostate size on cancer detection. Urology, 48:831, 1995. 14. Daneshgari, F., Taylor, G. D., Miller, G. J . and Crawford, E. D.: Computer simulation of the probability of detecting low volume carcinoma of the prostate with six random systematic core biopsies. Urology, 4 5 604, 1995. 15. Oesterling, J . E., Suman, V. J., Zincke, H. and Bostwick, D. G.: PSA-detected (clinical stage T l c or BO) prostate cancer. Pathologically sigruficant tumors. Uml. Clin. N. Amer., u):687,1993.

2125

16. Epstein, J. I., Walsh, P. C. and Brendler, C.B.: Radical prostatectomy for impalpable prostate cancer: the Johns Hopkins experience with tumors found on transurethral resection (stages T1A and T1B) and on needle biopsy (stage TlC). J. Urol., part 2, 152 1721, 1994. 17. Carter, H. B.,Pearson, J. D., Metter, E. J., Brant, L. J., Chan, D. W., Andres, R., Fozard, J . L. and Walsh, P. C.: Longitudinal evaluation of prostate-specific antigen levels in men with and without prostate disease. J.A.M.A., 267: 2215, 1992. 18. Pearson, J. D. and Carter, H. B.: Natural history of changes in prostate specific antigen in early stage prostate cancer. J. Urol., part 2, 152 1743, 1994.