- Email: [email protected]
Significance of high-grade prostatic intraepithelial neoplasia in needle biopsy specimens
ADULT UROLOGY CME ARTICLE
SIGNIFICANCE OF HIGH-GRADE PROSTATIC INTRAEPITHELIAL NEOPLASIA IN NEEDLE BIOPSY SPECIMENS WILHELM PRANGE, ANDREAS ERBERSDOBLER, PETER HAMMERER, MARKUS GRAEFEN, STEFAN H. HAUTMANN, RICHARD E. HAUTMANN, HARTWIG HULAND, AND ROLF-P. HENKE
ABSTRACT Objectives. To examine the significance of high-grade prostatic intraepithelial neoplasia (HGPIN) in biopsy specimens. Methods. We performed sextant biopsies on a series of 83 cystoprostatectomy specimens removed for bladder cancer. For each case the number of foci and volume of both HGPIN and prostate cancer were assessed in the prostatectomy specimens and compared with the number of biopsy specimens involved by HGPIN. Results. We identified HGPIN in 82 (99%) of 83 prostatic glands, whereas prostate cancer was found in 41 cases (49%). Corresponding sextant biopsies harbored both HGPIN and prostate cancer in 6 cases (7%), whereas only HGPIN was diagnosed in 29 sextant biopsies (35%). There was a positive correlation between the number of biopsy specimens containing HGPIN and the volume and multifocality of HGPIN in the corresponding prostatic glands. Prostates with HGPIN on sextant biopsy contained prostate cancer significantly more often when compared to cases with no HGPIN on sextant biopsy. Frequency of concurrent prostate cancer was higher in cases with two or more biopsy specimens containing HGPIN than in cases with only one such biopsy specimen, but case numbers of these categories were too small to render this difference statistically significant. Conclusions. The presence of HGPIN in sextant biopsies is a significant predictor of concurrent prostate cancer. Multifocality of HGPIN is a useful parameter in assessing the extent of HGPIN in the corresponding prostates. Its value in predicting a significantly increased risk of concurrent prostate cancer needs to be further investigated in larger case studies. UROLOGY 57: 486–490, 2001. © 2001, Elsevier Science Inc.
H
igh-grade prostatic intraepithelial neoplasia (HGPIN) is considered the most likely precursor of prostate carcinoma.1 At present, biopsy represents the only definitive method for detection of HGPIN and early invasive cancer.2 The predictive value of a diagnosis of HGPIN on needle biopsy is apparently in evolution. Although HGPIN was previously reported to provide the highest risk ratio of all known predictive factors,3 recent studies4,5 found its ability to predict concurrent prosFrom the Institute of Pathology and Department of Urology, University Hospital Eppendorf, University of Hamburg, Hamburg, Germany; and Department of Urology, University Hospital Ulm, University of Ulm, Ulm, Germany Reprint requests: Wilhelm Prange, M.D., Institute of Pathology, University of Cologne, Joseph-Stelzmann-Str. 9, 50931 Cologne, Germany Submitted: May 15, 2000, accepted (with revisions): October 10, 2000
486
© 2001, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
tate cancer much less dramatic. However, most researchers still recommend repeat biopsy and close surveillance of patients with isolated HGPIN in their initial biopsy.2,6 – 8 The reported prevalence of HGPIN in contemporary sextant biopsies varies considerably between 0.7% and 25%.6,9 –11 Furthermore, adenocarcinoma was found in between 35% and 100% of subsequent biopsies of patients with the diagnosis of isolated HGPIN in their first biopsy.3,7,12,13 Although the prevalence of HGPIN is thought to depend mainly on the population of men under study, the differences regarding the reported rates of prostate cancer in subsequent biopsies can be explained in part by the circumstance that previous studies had been designed as retrospective or prospective studies, based on the results of follow-up biopsies, and had the disadvantage of an unknown percentage of missed prostate cancers. Therefore, the significance of HGPIN on 0090-4295/01/$20.00 PII S0090-4295(00)01010-4
sextant biopsy should be further characterized by new strategies. In their study on cystoprostatectomy specimens, Wiley et al.14 showed that the risk of concurrent prostate cancer depends more on the volume of HGPIN than on its absolute presence. Nevertheless, in sextant biopsies, up to now, only the presence of HGPIN has been proven a significant predictor of concurrent prostate cancer, whereas data are poor concerning the significance of extent and multifocality of HGPIN in sextant biopsies. To investigate the significance of HGPIN in sextant biopsies of patients without clinical evidence of prostate cancer, we performed sextant biopsies on a series of cystoprostatectomy specimens before complete sectioning of the prostates. Findings in biopsy specimens concerning HGPIN and prostate cancer were compared with findings in corresponding prostatic glands. MATERIAL AND METHODS A total of 87 consecutive cystoprostatectomy specimens and sextant biopsies of the prostate from male patients with transitional cell carcinoma (TCC) of the urinary bladder removed at the Department of Urology, University of Ulm, and at the Department of Urology, University of Hamburg, were examined. Patients were included in this study only if they had a serum prostate-specific antigen (PSA) below 4.0 ng/mL and a normal digital rectal examination prior to surgery. Systematic sextant biopsies were performed with 18-gauge biopsy needles driven by the spring-loaded Biopty gun (Bard Urological, Covington, Ga). Biopsy cores were taken from the apex, middle, and basis of the prostate bilaterally as described by Hodge et al.15 but without ultrasonographic guidance because sextant biopsy was performed after the cystoprostatectomy on the specimen itself. As previously reported,16 prostatic glands were separated from the bladder at the level of the bladder neck, resulting in a specimen consisting of the prostate and portions of the right and left seminal vesicles. The whole specimen was fixed in 4.0% formalin, and the surface was inked. The seminal vesicles, apex, and base were amputated and serially sectioned separately from the remainder of the prostate, which was serially blocked at 4-mm intervals along transversal planes parallel to the initial apical and basal sections (ie, perpendicular to the rectal surface). Sections were thoroughly examined by routine light microscopy following hematoxylin-eosin staining. HGPIN was assigned according to the modified two-grade system as reviewed by Bostwick.17 To determine the volume of HGPIN and cancer of the prostate, the areas of the two entities were encircled on each slice. The area size was then assessed with a digitizing tablet (HDG 1212, Hitachi, Tokyo, Japan) connected to an 80-386 DX personal computer. The scanning software was SigmaScan, version 3.9 (Jandel Scientific, Corte Madera, Calif). The sum of the areas of each slice was multiplied by the thickness of the slice, and the sum of these volumes was multiplied by 1.5 to account for formalin shrinkage. Areas of HGPIN smaller than 0.01 cm2 were not outlined but recorded as a focus with a standardized volume of 0.0025 cm3. As in the study of Qian et al.,18 an area of HGPIN or prostate cancer was considered a separate focus if it was separated by 2 mm or more from the nearest adjacent focus of HGPIN or prostate cancer, respectively. Additional recordings were zonal distribution of prostate cancer, tumor category according to the tumor-node-metastasis (TNM) classification of the International Union against Cancer,19 volume, UROLOGY 57 (3), 2001
and histologic grading according to Gleason.20 Furthermore, multifocality of HGPIN in sextant biopsies was assessed by counting the number of biopsy specimens containing HGPIN. The pathologist who compared the sextant biopsies with the prostatic glands had no knowledge as to source and identity of the specimens. Results were correlated with number of foci of HGPIN and total volume of HGPIN in the corresponding prostatic glands. Significance of differences between groups with nonparametric distribution of data was analyzed with the Mann-Whitney U test. Spearman rank correlations were used to assess the degree of correlation between variables. Twosided P values less than 0.05 were considered significant.
RESULTS The ages of the patients ranged from 38 to 83 years (median 64). Four of the 87 cases were excluded from the study because of infiltration of more than 50% of the prostatic parenchyma by the TCC. Sextant biopsies from all remaining 83 cases were free of TCC, as well as 48 (58%) of the corresponding prostatic glands. Seven prostatic glands (8%) showed TCC in the periprostatic fatty tissue, and 18 (22%) harbored noninvasive TCC in the prostatic part of the urethra and in prostatic ducts. Intraductal TCC with invasion of prostatic stroma and TCC infiltrating the prostatic gland through the capsule were observed in 5 cases each (6%). In these cases, parenchyma infiltration was 1% of the prostate volume in 7 cases and 2%, 3%, and 6% in 1 case, respectively. Incidental prostate cancer was found in 41 (49%) of 83 prostatic glands with a volume range from 0.001 to 6.2 cm3 (median 0.06). Only 4 of these cases contained high-grade areas with Gleason pattern 4, measuring from 0.06 to 3.0 cm3 (median 0.49). Gleason pattern 5 was not found. In 34 cases (82%) prostate carcinoma was located only in the peripheral zone, in 2 cases (5%) it was confined to the transition zone, and in 5 prostates (13%) both peripheral and transition zones were involved. Sextant biopsies of 29 (35%) of 83 cases harbored HGPIN but no prostate cancer, whereas in 6 cases (7%) both HGPIN and prostate cancer were found on sextant biopsy. In the latter cases prostate cancer was found in one and two biopsy specimens in 3 cases each. CORRELATION OF HGPIN IN BIOPSIES AND PROSTATIC GLANDS HGPIN was identified in 82 (99%) of 83 prostatic glands and was multicentric in 79 cases (95%). The one specimen that did not harbor HGPIN was found to be free of prostate cancer as well. Number of foci of HGPIN ranged from 0 to 80 (median 20). Total volume of HGPIN ranged from 0 to 3.45 cm3 (median 0.17). Sextant biopsies of 48 (58%) of 83 cases contained no HGPIN, whereas HGPIN was found in one, two, and three biopsy specimens of sextant 487
TABLE I. Frequency of concurrent prostate cancer according to the number of biopsy specimens involved by HGPIN Cases No. of Biopsy Specimens Involved by HGPIN 0 1 2 3
With Prostate Cancer (%) 17 11 9 4
(35) (58) (82) (80)
Without Prostate Cancer (%) 31 8 2 1
(65) (42) (18) (20)
KEY: HGPIN ⫽ high-grade prostatic intraepithelial neoplasia.
HGPIN on sextant biopsy (35%) to cases with one biopsy specimen involved by HGPIN (58%) and cases with more than one such biopsy specimen (82%), whereas there was no recognizable increase between cases with two and three biopsy specimens containing HGPIN (Table I). With respect to the frequency of concurrent prostate cancer, the difference between the 48 cases without HGPIN on sextant biopsy and the 35 cases with one to three biopsy specimens containing HGPIN was significant (P ⫽ 0.003), whereas the difference between cases with one and more than one biopsy specimen containing HGPIN did not reach significance level (P ⫽ 0.14). COMMENT
FIGURE 1. Number of foci of HGPIN (A) and volume of HGPIN (B) in prostatic glands according to number of biopsy specimens involved by HGPIN. Central stripe indicates median, and central square indicates intraquartile range. Brackets indicate smallest and biggest result within the 1.5-fold of the intraquartile range to each side of the median, respectively, whereas additional results within and outside the threefold of the intraquartile range are indicated by open boxes and asterisks, respectively. R refers to the Spearman rank correlation coefficient.
biopsies in 19 (23%), 11 (13%), and 5 (6%) cases, respectively. There was a positive correlation between the number of biopsy specimens containing HGPIN and the number of foci of HGPIN (R ⫽ 0.58; P ⬍0.001) as well as the total volume of HGPIN (R ⫽ 0.49; P ⬍0.001) in the corresponding prostatic glands (Fig. 1). RELATIONSHIP BETWEEN HGPIN AND PROSTATE CANCER The percentage of prostates involved by prostate cancer increased stepwise from cases with no 488
In this comparative study on HGPIN and concurrent prostate cancer in sextant biopsies and corresponding prostatic glands, we identified HGPIN in 82 (99%) of 83 prostatic glands, whereas prostate cancer was found in 41 cases (49%). Corresponding sextant biopsies harbored both HGPIN and prostate cancer in 6 cases (7%), whereas HGPIN in the absence of cancer was diagnosed in 29 sextant biopsies (35%). The number of biopsy specimens containing HGPIN was positively correlated with the number of foci of HGPIN and the total volume of HGPIN in the corresponding prostatic glands. Prostates with HGPIN on sextant biopsy significantly more often contained prostate cancer when compared to cases with no HGPIN on sextant biopsy. The frequency of concurrent prostate cancer was higher in cases with two or more biopsy specimens containing HGPIN than in cases with only one such biopsy specimen, but the difference did not reach significance level. Although the frequency of incidental prostate cancer (49%) in our study is within the range reported previously for white men with a mean age of 65 years,21,22 the frequency of HGPIN for both sextant biopsies (35%) and prostatic glands (99%) is markedly higher than in most previous studies: Autopsy studies reported a prevalence of HGPIN in UROLOGY 57 (3), 2001
white men with a mean age of 65 years between 50% and 63%.22,23 The reported prevalence of HGPIN in contemporary sextant biopsies of patients with clinical suspicion of prostate cancer because of elevated serum PSA or an abnormal digital rectal examination varies between 0.7% and 25%3,6,11,24 and is considered to depend on the population of men under study.2 In this context the unusually high frequency of HGPIN in sextant biopsies of patients with a normal serum PSA as well as a normal digital rectal examination in our study is remarkable. There are no obvious racial or regional factors that could have led to the increased frequency of HGPIN in our study, whereas methodical differences might well have contributed to the increase. In contrast to other reports in which prostatic glands were sectioned at 5 to 8-mm intervals,14,18 the mean slice thickness in our study was 4 mm. So, in our study more slides were examined per case, thereby increasing the probability of detecting HGPIN. However, a previous study25 on cystoprostatectomy specimens found a markedly increased prevalence (89%) of HGPIN in the prostates of patients with bladder cancer as well. Like the frequency of HGPIN reported in our study, this prevalence is within the range reported for radical prostatectomy specimens of patients with prostate cancer (85% to 100%).18,26,27 These results seem to indicate that the risk of developing HGPIN in patients with bladder cancer is higher than in the general population and that the high frequency of HGPIN in sextant biopsies in our study just reflects the high prevalence of HGPIN in the corresponding prostatic glands. At present, an increased risk of concurrent prostate cancer for patients with TCC of the bladder has not yet been documented, but larger scale studies should further investigate this topic. The predictive value of HGPIN on needle biopsy is apparently in evolution as several recent studies4,5 found the ability of HGPIN to predict carcinoma on repeat biopsy to be much less dramatic than reported earlier. We found concurrent prostate carcinoma significantly more often in prostates with HGPIN on sextant biopsy than in prostates without HGPIN on sextant biopsy, thereby confirming numerous previous reports3,6,7,12 that showed that the presence of HGPIN on needle biopsy significantly increases the likelihood of concurrent prostate cancer. The special feature of the present study is the new study design we have introduced here that enabled us to obtain the true prevalence of prostate cancer in the biopsied population by having the entire prostates available for histologic examination. However, our study is obviously limited in that the risk of developing HGPIN of the population used in the present study is apparently different from that of the general population. Therefore, our results may not prove true if simply UROLOGY 57 (3), 2001
transferred to populations other than the studied one. Another limitation is that the sampling scheme used in the present study differed from the “classical” systematic sextant biopsy in that for ethical reasons, biopsies had to be performed after cystoprostatectomy, that is, on the specimen itself. This sampling scheme precluded an ultrasonographic guidance, but apart from that, efforts had been made to copy the normal biopsy procedure as closely as possible. We found multifocality of HGPIN, assessed by the number of biopsy cores containing HGPIN, to be positively correlated with the number of HGPIN foci and volume of HGPIN in corresponding prostates. Multifocality of HGPIN on sextant biopsy, therefore, seems to be a useful tool for predicting the extent of HGPIN in a given prostate by means of a sextant biopsy. This predictability could be of clinical significance, as Wiley et al.14 reported the risk of concurrent prostate cancer to depend more on the extent of HGPIN than on its absolute presence, and the extent of HGPIN is reported to be positively correlated to the volume of prostate cancer in early stages.18 However, the significance of multifocality and extent of HGPIN in sextant biopsies is poorly investigated. In their retrospective study based on the results of follow-up biopsies, Davidson et al.3 reported the presence but not the amount of HGPIN in sextant biopsies to be a significant predictor of concurrent prostate cancer. In agreement with finding the multifocality of HGPIN on needle biopsy to correlate with the extent of HGPIN in the corresponding prostatic gland and in further support of the hypothesis of Wiley et al.14 that the risk of incidental prostate cancer is related to the extent of HGPIN, we found that the likelihood of concurrent prostate cancer in the specimen is higher in cases with two or more biopsy cores containing HGPIN than in cases with only one involved core. This finding is still of preliminary nature, as numbers of cases with HGPIN on needle biopsy were too small to render the observed differences significant. Because there are no other reports on this topic to date, the predictive value of multifocality of HGPIN in sextant biopsies should be further investigated in larger scale studies. CONCLUSIONS The presence of HGPIN in sextant biopsies is a significant predictor of concurrent prostate cancer. The number of biopsy specimens positive for HGPIN in a sextant biopsy is positively correlated with the extent of HGPIN in the corresponding prostate. The relationship between multifocality of HGPIN in sextant biopsies and the risk of concurrent prostate cancer should be investigated in further studies. 489
REFERENCES 1. Bostwick DG: Prostatic intraepithelial neoplasia (PIN): current concepts. J Cell Biochem 16: 10 –19, 1992. 2. Bostwick DG: Prostatic intraepithelial neoplasia is a risk factor for cancer. Semin Urol Oncol 17: 187–198, 1999. 3. Davidson D, Bostwick DG, Qian J, et al: Prostatic intraepithelial neoplasia is a risk factor for adenocarcinoma: predictive accuracy in needle biopsies. J Urol 154: 1295–1299, 1995. 4. Fowler JE Jr, Bigler SA, Miles D, et al: Predictors of first repeat biopsy cancer detection with suspected local stage prostate cancer. J Urol 163: 813– 818, 2000. 5. O’Dowd GJ, Miller MC, Orozco R, et al: Analysis of repeated biopsy results within 1 year after a noncancer diagnosis. Urology 55: 553–558, 2000. 6. Brawer MK, Bigler SA, Sohlberg OE, et al: Significance of prostatic intraepithelial neoplasia on prostate needle biopsy. Urology 38: 103–107, 1991. 7. Weinstein MH, and Epstein JI: Significance of highgrade prostatic intraepithelial neoplasia on needle biopsy. Hum Pathol 24: 624 – 629, 1993. 8. Chan TY, and Epstein JI: Follow-up of atypical prostate needle biopsies suspicious for cancer. Urology 53: 351–355, 1999. 9. Wills ML, Hamper UM, Partin AW, et al: Incidence of high-grade prostatic intraepithelial neoplasia in sextant biopsy specimens. Urology 49: 367–373, 1997. 10. Hoedemaeker RF, Kranse R, Rietbergen JB, et al: Evaluation of prostate needle biopsies in a population-based screening study: impact of borderline lesions. Cancer 85: 145– 152, 1999. 11. Feneley MR, Green JSA, Young MPA, et al: Prevalence of prostatic intraepithelial neoplasia (PIN) in biopsies from hospital practice and pilot screening: clinical implications. Prostate Cancer Prostatic Diseases 1: 79 – 83, 1997. 12. Ellis WJ, and Brawer MK: Repeat needle biopsy: who needs it? J Urol 153: 1496 –1498, 1995. 13. Deliveliotis C, Louras G, Raptidis G, et al: Evaluation of needle biopsy in the diagnosis of prostatic carcinoma in men with prostatic intraepithelial neoplasia. Scand J Urol Nephrol 32: 107–110, 1998. 14. Wiley EL, Davidson P, McIntire DD, et al: Risk of concurrent prostate cancer in cystoprostatectomy specimens is related to volume of high-grade prostatic intraepithelial neoplasia. Urology 49: 692– 696, 1997.
490
15. Hodge KK, McNeal JE, Terris MK, et al: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol 142: 71–75, 1989. 16. Ayala AG, Ro JY, Babaian R, et al: The prostatic capsule: does it exist? Am J Surg Pathol 13: 21–27, 1989. 17. Bostwick DG: High grade prostatic intraepithelial neoplasia. The most likely precursor of cancer. Cancer 75(suppl): 1823–1836, 1995. 18. Qian J, Wollan P, and Bostwick DG: The extent and multicentricity of high-grade prostatic intraepithelial neoplasia in clinically localized prostatic adenocarcinoma. Hum Pathol 28: 143–148, 1997. 19. Union Internationale contre le Cancer: Prostate, in Sobin LH, and Wittekind CH (Eds): TNM Classification of Malignant Tumors, 5th ed. New York, Wiley, 1997, pp 170 –173. 20. Gleason DF, and the Veterans Administration Cooperative Urological Research Group: Histological grading and clinical staging of prostate carcinoma, in Tannenbaum M (Ed): Urologic Pathology: The Prostate. Philadelphia, Lea & Febinger, 1977, pp 171–197. 21. Montie JE, Wood DP, Pontes JE, et al: Adenocarcinoma of the prostate in cystoprostatectomy specimens removed for bladder cancer. Cancer 63: 381–385, 1989. 22. Sakr WA, Grignon DJ, Crissman JD, et al: High-grade prostatic intraepithelial neoplasia (HGPIN) and prostatic adenocarcinoma between the ages of 20 – 69: an autopsy study of 249 cases. In Vivo 8: 439 – 443, 1994. 23. Sakr WA, Grignon DJ, Haas GP, et al: Epidemiology of high-grade prostatic intraepithelial neoplasia. Pathol Res Pract 191: 838 – 841, 1995. 24. Bostwick DG, Qian J, and Frankel K: The incidence of high-grade prostatic intraepithelial neoplasia in needle biopsies. J Urol 154: 1791–1794, 1995. 25. Troncoso P, Babaian RJ, Ro JY, et al: Prostatic intraepithelial neoplasia and invasive adenocarcinoma in cystoprostatectomy specimens. Urology 24(suppl): 52–56, 1989. 26. Quinn BD, Cho KR, and Epstein JI: Relationship of severe dysplasia to stage B adenocarcinoma of the prostate. Cancer 65: 2328 –2333, 1990. 27. De la Torre M, Haggman M, Brandstedt S, et al: Prostatic intraepithelial neoplasia and invasive carcinoma in total prostatectomy specimens: distribution, volume and DNA ploidy. Br J Urol 72: 207–213, 1993.
UROLOGY 57 (3), 2001
SIGNIFICANCE OF HIGH-GRADE PROSTATIC INTRAEPITHELIAL NEOPLASIA IN NEEDLE BIOPSY SPECIMENS WILHELM PRANGE, ANDREAS ERBERSDOBLER, PETER HAMMERER, MARKUS GRAEFEN, STEFAN H. HAUTMANN, RICHARD E. HAUTMANN, HARTWIG HULAND, AND ROLF-P. HENKE
ABSTRACT Objectives. To examine the significance of high-grade prostatic intraepithelial neoplasia (HGPIN) in biopsy specimens. Methods. We performed sextant biopsies on a series of 83 cystoprostatectomy specimens removed for bladder cancer. For each case the number of foci and volume of both HGPIN and prostate cancer were assessed in the prostatectomy specimens and compared with the number of biopsy specimens involved by HGPIN. Results. We identified HGPIN in 82 (99%) of 83 prostatic glands, whereas prostate cancer was found in 41 cases (49%). Corresponding sextant biopsies harbored both HGPIN and prostate cancer in 6 cases (7%), whereas only HGPIN was diagnosed in 29 sextant biopsies (35%). There was a positive correlation between the number of biopsy specimens containing HGPIN and the volume and multifocality of HGPIN in the corresponding prostatic glands. Prostates with HGPIN on sextant biopsy contained prostate cancer significantly more often when compared to cases with no HGPIN on sextant biopsy. Frequency of concurrent prostate cancer was higher in cases with two or more biopsy specimens containing HGPIN than in cases with only one such biopsy specimen, but case numbers of these categories were too small to render this difference statistically significant. Conclusions. The presence of HGPIN in sextant biopsies is a significant predictor of concurrent prostate cancer. Multifocality of HGPIN is a useful parameter in assessing the extent of HGPIN in the corresponding prostates. Its value in predicting a significantly increased risk of concurrent prostate cancer needs to be further investigated in larger case studies. UROLOGY 57: 486–490, 2001. © 2001, Elsevier Science Inc.
H
igh-grade prostatic intraepithelial neoplasia (HGPIN) is considered the most likely precursor of prostate carcinoma.1 At present, biopsy represents the only definitive method for detection of HGPIN and early invasive cancer.2 The predictive value of a diagnosis of HGPIN on needle biopsy is apparently in evolution. Although HGPIN was previously reported to provide the highest risk ratio of all known predictive factors,3 recent studies4,5 found its ability to predict concurrent prosFrom the Institute of Pathology and Department of Urology, University Hospital Eppendorf, University of Hamburg, Hamburg, Germany; and Department of Urology, University Hospital Ulm, University of Ulm, Ulm, Germany Reprint requests: Wilhelm Prange, M.D., Institute of Pathology, University of Cologne, Joseph-Stelzmann-Str. 9, 50931 Cologne, Germany Submitted: May 15, 2000, accepted (with revisions): October 10, 2000
486
© 2001, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
tate cancer much less dramatic. However, most researchers still recommend repeat biopsy and close surveillance of patients with isolated HGPIN in their initial biopsy.2,6 – 8 The reported prevalence of HGPIN in contemporary sextant biopsies varies considerably between 0.7% and 25%.6,9 –11 Furthermore, adenocarcinoma was found in between 35% and 100% of subsequent biopsies of patients with the diagnosis of isolated HGPIN in their first biopsy.3,7,12,13 Although the prevalence of HGPIN is thought to depend mainly on the population of men under study, the differences regarding the reported rates of prostate cancer in subsequent biopsies can be explained in part by the circumstance that previous studies had been designed as retrospective or prospective studies, based on the results of follow-up biopsies, and had the disadvantage of an unknown percentage of missed prostate cancers. Therefore, the significance of HGPIN on 0090-4295/01/$20.00 PII S0090-4295(00)01010-4
sextant biopsy should be further characterized by new strategies. In their study on cystoprostatectomy specimens, Wiley et al.14 showed that the risk of concurrent prostate cancer depends more on the volume of HGPIN than on its absolute presence. Nevertheless, in sextant biopsies, up to now, only the presence of HGPIN has been proven a significant predictor of concurrent prostate cancer, whereas data are poor concerning the significance of extent and multifocality of HGPIN in sextant biopsies. To investigate the significance of HGPIN in sextant biopsies of patients without clinical evidence of prostate cancer, we performed sextant biopsies on a series of cystoprostatectomy specimens before complete sectioning of the prostates. Findings in biopsy specimens concerning HGPIN and prostate cancer were compared with findings in corresponding prostatic glands. MATERIAL AND METHODS A total of 87 consecutive cystoprostatectomy specimens and sextant biopsies of the prostate from male patients with transitional cell carcinoma (TCC) of the urinary bladder removed at the Department of Urology, University of Ulm, and at the Department of Urology, University of Hamburg, were examined. Patients were included in this study only if they had a serum prostate-specific antigen (PSA) below 4.0 ng/mL and a normal digital rectal examination prior to surgery. Systematic sextant biopsies were performed with 18-gauge biopsy needles driven by the spring-loaded Biopty gun (Bard Urological, Covington, Ga). Biopsy cores were taken from the apex, middle, and basis of the prostate bilaterally as described by Hodge et al.15 but without ultrasonographic guidance because sextant biopsy was performed after the cystoprostatectomy on the specimen itself. As previously reported,16 prostatic glands were separated from the bladder at the level of the bladder neck, resulting in a specimen consisting of the prostate and portions of the right and left seminal vesicles. The whole specimen was fixed in 4.0% formalin, and the surface was inked. The seminal vesicles, apex, and base were amputated and serially sectioned separately from the remainder of the prostate, which was serially blocked at 4-mm intervals along transversal planes parallel to the initial apical and basal sections (ie, perpendicular to the rectal surface). Sections were thoroughly examined by routine light microscopy following hematoxylin-eosin staining. HGPIN was assigned according to the modified two-grade system as reviewed by Bostwick.17 To determine the volume of HGPIN and cancer of the prostate, the areas of the two entities were encircled on each slice. The area size was then assessed with a digitizing tablet (HDG 1212, Hitachi, Tokyo, Japan) connected to an 80-386 DX personal computer. The scanning software was SigmaScan, version 3.9 (Jandel Scientific, Corte Madera, Calif). The sum of the areas of each slice was multiplied by the thickness of the slice, and the sum of these volumes was multiplied by 1.5 to account for formalin shrinkage. Areas of HGPIN smaller than 0.01 cm2 were not outlined but recorded as a focus with a standardized volume of 0.0025 cm3. As in the study of Qian et al.,18 an area of HGPIN or prostate cancer was considered a separate focus if it was separated by 2 mm or more from the nearest adjacent focus of HGPIN or prostate cancer, respectively. Additional recordings were zonal distribution of prostate cancer, tumor category according to the tumor-node-metastasis (TNM) classification of the International Union against Cancer,19 volume, UROLOGY 57 (3), 2001
and histologic grading according to Gleason.20 Furthermore, multifocality of HGPIN in sextant biopsies was assessed by counting the number of biopsy specimens containing HGPIN. The pathologist who compared the sextant biopsies with the prostatic glands had no knowledge as to source and identity of the specimens. Results were correlated with number of foci of HGPIN and total volume of HGPIN in the corresponding prostatic glands. Significance of differences between groups with nonparametric distribution of data was analyzed with the Mann-Whitney U test. Spearman rank correlations were used to assess the degree of correlation between variables. Twosided P values less than 0.05 were considered significant.
RESULTS The ages of the patients ranged from 38 to 83 years (median 64). Four of the 87 cases were excluded from the study because of infiltration of more than 50% of the prostatic parenchyma by the TCC. Sextant biopsies from all remaining 83 cases were free of TCC, as well as 48 (58%) of the corresponding prostatic glands. Seven prostatic glands (8%) showed TCC in the periprostatic fatty tissue, and 18 (22%) harbored noninvasive TCC in the prostatic part of the urethra and in prostatic ducts. Intraductal TCC with invasion of prostatic stroma and TCC infiltrating the prostatic gland through the capsule were observed in 5 cases each (6%). In these cases, parenchyma infiltration was 1% of the prostate volume in 7 cases and 2%, 3%, and 6% in 1 case, respectively. Incidental prostate cancer was found in 41 (49%) of 83 prostatic glands with a volume range from 0.001 to 6.2 cm3 (median 0.06). Only 4 of these cases contained high-grade areas with Gleason pattern 4, measuring from 0.06 to 3.0 cm3 (median 0.49). Gleason pattern 5 was not found. In 34 cases (82%) prostate carcinoma was located only in the peripheral zone, in 2 cases (5%) it was confined to the transition zone, and in 5 prostates (13%) both peripheral and transition zones were involved. Sextant biopsies of 29 (35%) of 83 cases harbored HGPIN but no prostate cancer, whereas in 6 cases (7%) both HGPIN and prostate cancer were found on sextant biopsy. In the latter cases prostate cancer was found in one and two biopsy specimens in 3 cases each. CORRELATION OF HGPIN IN BIOPSIES AND PROSTATIC GLANDS HGPIN was identified in 82 (99%) of 83 prostatic glands and was multicentric in 79 cases (95%). The one specimen that did not harbor HGPIN was found to be free of prostate cancer as well. Number of foci of HGPIN ranged from 0 to 80 (median 20). Total volume of HGPIN ranged from 0 to 3.45 cm3 (median 0.17). Sextant biopsies of 48 (58%) of 83 cases contained no HGPIN, whereas HGPIN was found in one, two, and three biopsy specimens of sextant 487
TABLE I. Frequency of concurrent prostate cancer according to the number of biopsy specimens involved by HGPIN Cases No. of Biopsy Specimens Involved by HGPIN 0 1 2 3
With Prostate Cancer (%) 17 11 9 4
(35) (58) (82) (80)
Without Prostate Cancer (%) 31 8 2 1
(65) (42) (18) (20)
KEY: HGPIN ⫽ high-grade prostatic intraepithelial neoplasia.
HGPIN on sextant biopsy (35%) to cases with one biopsy specimen involved by HGPIN (58%) and cases with more than one such biopsy specimen (82%), whereas there was no recognizable increase between cases with two and three biopsy specimens containing HGPIN (Table I). With respect to the frequency of concurrent prostate cancer, the difference between the 48 cases without HGPIN on sextant biopsy and the 35 cases with one to three biopsy specimens containing HGPIN was significant (P ⫽ 0.003), whereas the difference between cases with one and more than one biopsy specimen containing HGPIN did not reach significance level (P ⫽ 0.14). COMMENT
FIGURE 1. Number of foci of HGPIN (A) and volume of HGPIN (B) in prostatic glands according to number of biopsy specimens involved by HGPIN. Central stripe indicates median, and central square indicates intraquartile range. Brackets indicate smallest and biggest result within the 1.5-fold of the intraquartile range to each side of the median, respectively, whereas additional results within and outside the threefold of the intraquartile range are indicated by open boxes and asterisks, respectively. R refers to the Spearman rank correlation coefficient.
biopsies in 19 (23%), 11 (13%), and 5 (6%) cases, respectively. There was a positive correlation between the number of biopsy specimens containing HGPIN and the number of foci of HGPIN (R ⫽ 0.58; P ⬍0.001) as well as the total volume of HGPIN (R ⫽ 0.49; P ⬍0.001) in the corresponding prostatic glands (Fig. 1). RELATIONSHIP BETWEEN HGPIN AND PROSTATE CANCER The percentage of prostates involved by prostate cancer increased stepwise from cases with no 488
In this comparative study on HGPIN and concurrent prostate cancer in sextant biopsies and corresponding prostatic glands, we identified HGPIN in 82 (99%) of 83 prostatic glands, whereas prostate cancer was found in 41 cases (49%). Corresponding sextant biopsies harbored both HGPIN and prostate cancer in 6 cases (7%), whereas HGPIN in the absence of cancer was diagnosed in 29 sextant biopsies (35%). The number of biopsy specimens containing HGPIN was positively correlated with the number of foci of HGPIN and the total volume of HGPIN in the corresponding prostatic glands. Prostates with HGPIN on sextant biopsy significantly more often contained prostate cancer when compared to cases with no HGPIN on sextant biopsy. The frequency of concurrent prostate cancer was higher in cases with two or more biopsy specimens containing HGPIN than in cases with only one such biopsy specimen, but the difference did not reach significance level. Although the frequency of incidental prostate cancer (49%) in our study is within the range reported previously for white men with a mean age of 65 years,21,22 the frequency of HGPIN for both sextant biopsies (35%) and prostatic glands (99%) is markedly higher than in most previous studies: Autopsy studies reported a prevalence of HGPIN in UROLOGY 57 (3), 2001
white men with a mean age of 65 years between 50% and 63%.22,23 The reported prevalence of HGPIN in contemporary sextant biopsies of patients with clinical suspicion of prostate cancer because of elevated serum PSA or an abnormal digital rectal examination varies between 0.7% and 25%3,6,11,24 and is considered to depend on the population of men under study.2 In this context the unusually high frequency of HGPIN in sextant biopsies of patients with a normal serum PSA as well as a normal digital rectal examination in our study is remarkable. There are no obvious racial or regional factors that could have led to the increased frequency of HGPIN in our study, whereas methodical differences might well have contributed to the increase. In contrast to other reports in which prostatic glands were sectioned at 5 to 8-mm intervals,14,18 the mean slice thickness in our study was 4 mm. So, in our study more slides were examined per case, thereby increasing the probability of detecting HGPIN. However, a previous study25 on cystoprostatectomy specimens found a markedly increased prevalence (89%) of HGPIN in the prostates of patients with bladder cancer as well. Like the frequency of HGPIN reported in our study, this prevalence is within the range reported for radical prostatectomy specimens of patients with prostate cancer (85% to 100%).18,26,27 These results seem to indicate that the risk of developing HGPIN in patients with bladder cancer is higher than in the general population and that the high frequency of HGPIN in sextant biopsies in our study just reflects the high prevalence of HGPIN in the corresponding prostatic glands. At present, an increased risk of concurrent prostate cancer for patients with TCC of the bladder has not yet been documented, but larger scale studies should further investigate this topic. The predictive value of HGPIN on needle biopsy is apparently in evolution as several recent studies4,5 found the ability of HGPIN to predict carcinoma on repeat biopsy to be much less dramatic than reported earlier. We found concurrent prostate carcinoma significantly more often in prostates with HGPIN on sextant biopsy than in prostates without HGPIN on sextant biopsy, thereby confirming numerous previous reports3,6,7,12 that showed that the presence of HGPIN on needle biopsy significantly increases the likelihood of concurrent prostate cancer. The special feature of the present study is the new study design we have introduced here that enabled us to obtain the true prevalence of prostate cancer in the biopsied population by having the entire prostates available for histologic examination. However, our study is obviously limited in that the risk of developing HGPIN of the population used in the present study is apparently different from that of the general population. Therefore, our results may not prove true if simply UROLOGY 57 (3), 2001
transferred to populations other than the studied one. Another limitation is that the sampling scheme used in the present study differed from the “classical” systematic sextant biopsy in that for ethical reasons, biopsies had to be performed after cystoprostatectomy, that is, on the specimen itself. This sampling scheme precluded an ultrasonographic guidance, but apart from that, efforts had been made to copy the normal biopsy procedure as closely as possible. We found multifocality of HGPIN, assessed by the number of biopsy cores containing HGPIN, to be positively correlated with the number of HGPIN foci and volume of HGPIN in corresponding prostates. Multifocality of HGPIN on sextant biopsy, therefore, seems to be a useful tool for predicting the extent of HGPIN in a given prostate by means of a sextant biopsy. This predictability could be of clinical significance, as Wiley et al.14 reported the risk of concurrent prostate cancer to depend more on the extent of HGPIN than on its absolute presence, and the extent of HGPIN is reported to be positively correlated to the volume of prostate cancer in early stages.18 However, the significance of multifocality and extent of HGPIN in sextant biopsies is poorly investigated. In their retrospective study based on the results of follow-up biopsies, Davidson et al.3 reported the presence but not the amount of HGPIN in sextant biopsies to be a significant predictor of concurrent prostate cancer. In agreement with finding the multifocality of HGPIN on needle biopsy to correlate with the extent of HGPIN in the corresponding prostatic gland and in further support of the hypothesis of Wiley et al.14 that the risk of incidental prostate cancer is related to the extent of HGPIN, we found that the likelihood of concurrent prostate cancer in the specimen is higher in cases with two or more biopsy cores containing HGPIN than in cases with only one involved core. This finding is still of preliminary nature, as numbers of cases with HGPIN on needle biopsy were too small to render the observed differences significant. Because there are no other reports on this topic to date, the predictive value of multifocality of HGPIN in sextant biopsies should be further investigated in larger scale studies. CONCLUSIONS The presence of HGPIN in sextant biopsies is a significant predictor of concurrent prostate cancer. The number of biopsy specimens positive for HGPIN in a sextant biopsy is positively correlated with the extent of HGPIN in the corresponding prostate. The relationship between multifocality of HGPIN in sextant biopsies and the risk of concurrent prostate cancer should be investigated in further studies. 489
REFERENCES 1. Bostwick DG: Prostatic intraepithelial neoplasia (PIN): current concepts. J Cell Biochem 16: 10 –19, 1992. 2. Bostwick DG: Prostatic intraepithelial neoplasia is a risk factor for cancer. Semin Urol Oncol 17: 187–198, 1999. 3. Davidson D, Bostwick DG, Qian J, et al: Prostatic intraepithelial neoplasia is a risk factor for adenocarcinoma: predictive accuracy in needle biopsies. J Urol 154: 1295–1299, 1995. 4. Fowler JE Jr, Bigler SA, Miles D, et al: Predictors of first repeat biopsy cancer detection with suspected local stage prostate cancer. J Urol 163: 813– 818, 2000. 5. O’Dowd GJ, Miller MC, Orozco R, et al: Analysis of repeated biopsy results within 1 year after a noncancer diagnosis. Urology 55: 553–558, 2000. 6. Brawer MK, Bigler SA, Sohlberg OE, et al: Significance of prostatic intraepithelial neoplasia on prostate needle biopsy. Urology 38: 103–107, 1991. 7. Weinstein MH, and Epstein JI: Significance of highgrade prostatic intraepithelial neoplasia on needle biopsy. Hum Pathol 24: 624 – 629, 1993. 8. Chan TY, and Epstein JI: Follow-up of atypical prostate needle biopsies suspicious for cancer. Urology 53: 351–355, 1999. 9. Wills ML, Hamper UM, Partin AW, et al: Incidence of high-grade prostatic intraepithelial neoplasia in sextant biopsy specimens. Urology 49: 367–373, 1997. 10. Hoedemaeker RF, Kranse R, Rietbergen JB, et al: Evaluation of prostate needle biopsies in a population-based screening study: impact of borderline lesions. Cancer 85: 145– 152, 1999. 11. Feneley MR, Green JSA, Young MPA, et al: Prevalence of prostatic intraepithelial neoplasia (PIN) in biopsies from hospital practice and pilot screening: clinical implications. Prostate Cancer Prostatic Diseases 1: 79 – 83, 1997. 12. Ellis WJ, and Brawer MK: Repeat needle biopsy: who needs it? J Urol 153: 1496 –1498, 1995. 13. Deliveliotis C, Louras G, Raptidis G, et al: Evaluation of needle biopsy in the diagnosis of prostatic carcinoma in men with prostatic intraepithelial neoplasia. Scand J Urol Nephrol 32: 107–110, 1998. 14. Wiley EL, Davidson P, McIntire DD, et al: Risk of concurrent prostate cancer in cystoprostatectomy specimens is related to volume of high-grade prostatic intraepithelial neoplasia. Urology 49: 692– 696, 1997.
490
15. Hodge KK, McNeal JE, Terris MK, et al: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol 142: 71–75, 1989. 16. Ayala AG, Ro JY, Babaian R, et al: The prostatic capsule: does it exist? Am J Surg Pathol 13: 21–27, 1989. 17. Bostwick DG: High grade prostatic intraepithelial neoplasia. The most likely precursor of cancer. Cancer 75(suppl): 1823–1836, 1995. 18. Qian J, Wollan P, and Bostwick DG: The extent and multicentricity of high-grade prostatic intraepithelial neoplasia in clinically localized prostatic adenocarcinoma. Hum Pathol 28: 143–148, 1997. 19. Union Internationale contre le Cancer: Prostate, in Sobin LH, and Wittekind CH (Eds): TNM Classification of Malignant Tumors, 5th ed. New York, Wiley, 1997, pp 170 –173. 20. Gleason DF, and the Veterans Administration Cooperative Urological Research Group: Histological grading and clinical staging of prostate carcinoma, in Tannenbaum M (Ed): Urologic Pathology: The Prostate. Philadelphia, Lea & Febinger, 1977, pp 171–197. 21. Montie JE, Wood DP, Pontes JE, et al: Adenocarcinoma of the prostate in cystoprostatectomy specimens removed for bladder cancer. Cancer 63: 381–385, 1989. 22. Sakr WA, Grignon DJ, Crissman JD, et al: High-grade prostatic intraepithelial neoplasia (HGPIN) and prostatic adenocarcinoma between the ages of 20 – 69: an autopsy study of 249 cases. In Vivo 8: 439 – 443, 1994. 23. Sakr WA, Grignon DJ, Haas GP, et al: Epidemiology of high-grade prostatic intraepithelial neoplasia. Pathol Res Pract 191: 838 – 841, 1995. 24. Bostwick DG, Qian J, and Frankel K: The incidence of high-grade prostatic intraepithelial neoplasia in needle biopsies. J Urol 154: 1791–1794, 1995. 25. Troncoso P, Babaian RJ, Ro JY, et al: Prostatic intraepithelial neoplasia and invasive adenocarcinoma in cystoprostatectomy specimens. Urology 24(suppl): 52–56, 1989. 26. Quinn BD, Cho KR, and Epstein JI: Relationship of severe dysplasia to stage B adenocarcinoma of the prostate. Cancer 65: 2328 –2333, 1990. 27. De la Torre M, Haggman M, Brandstedt S, et al: Prostatic intraepithelial neoplasia and invasive carcinoma in total prostatectomy specimens: distribution, volume and DNA ploidy. Br J Urol 72: 207–213, 1993.
UROLOGY 57 (3), 2001