Path. Res. Pract. 191, 850-855 (1995)
Prostatic Intraepithelial Neoplasia and Well Differentiated Adenocarcinoma Maintain an Intact Basement Membrane* D. G. Bostwick1, D. A. Leske1, Junqi Qian 1 and A. A. Sinha2 1Mayo Clinic, Rochester, MN, and 2Department of Veterans Affairs, Minneapolis, MN, USA
SUMMARY The status of the basement membrane in prostatic intraepithelial neoplasia (PIN) and adenocarcinoma is unsettled. Previous studies using antibodies directed against Type IV collagen have been hindered by intense staining around the stromal smooth muscle fibers, making interpretation of acinar staining difficult. We employed a monoclonal antibody to heparan sulfate proteoglycan (HSPG) to overcome this problem, recognizing that HSPG is present in the basement membrane of epithelial and endothelial cells, but not stromal smooth muscle cells. In 22 totally-embedded whole mount radical prostatectomies for adenocarcinoma which contained PIN, intense HSPG immunoreactivity was observed in the basement membrane of all normal and hyperplastic acini, 98% of acini with high grade PIN, and 100% of acini of well differentiated (Gleason score 5) adenocarcinoma; vessels served as the internal positive control, with consistent staining throughout each specimen. The extent of HSPG immunoreactivity in cancer decreased with increasing Gleason grade (measured as percent of acini staining, in 10% increments; p = 0.002). These findings indicate that HSPG is a consistent component of the basement membrane of benign, hyperplastic, and early neoplastic prostatic acini, and, unlike other extracellular matrix proteins such as type IV collagen, is not hindered by background staining around stromal smooth muscle cells. High grade PIN and well differentiated adenocarcinoma usually maintain an intact basement membrane, and loss of the basement membrane occurs with histologic dedifferentiation.
Introduction Adenocarcinoma arising in many organs disrupts the basement membrane, but this is not a constant finding l . There have been conflicting reports regarding the status of the basement membrane in prostate cancer, with some claiming that it is disrupted and others refuting this l - 1o • Many of these studies have been hindered by .. Supported in part by National Institutes of Health Grant CA58225-01C. 0344-0338/95/0191-0850$3.50/0
the use of antibodies directed against type IV collagen5,6,9,11; this extracellular matrix protein is abundant in the prostate, encircling stromal smooth muscle fibers and confounding interpretation of staining in benign and neoplastic acini which are often intimately associated with the stroma. To overcome the difficulty of stromal smooth muscle staining, we employed antibodies directed against heparan sulfate proteoglycan (HSPG) to evaluate the extent and intensity of basement membrane immunoreactivity in normal and hyperplastic prostatic acini, high grade prostatic intraepithelial neoplasia, © 1995 by Gustav Fischer Verlag, Stuttgart
Basement Membrane in PIN . 851
and carcinoma 12,13. We found that HSPG is a useful marker of the prostatic basement membrane which is not hindered by the abundant staining of stromal smooth muscle which occurs with type IV collagen.
Fig. 1. Intact basement membrane in high grade PIN. A: x200; B: x400 (heparan sulfate proteoglycan immunostaining).
Our results indicate that the HSPG-immunoreactive basement membrane is maintained in prostatic intraepithelial neoplasia and well differentiated cancer, but is progressively lost with increasing cancer grade.
852 . D. G. Bostwick et al.
Results Case Description
Patients ranged in age from 52 to 76 years, with a mean of 65 years (standard deviation, 6.9 years). Pre-
operative serum PSA concentration was 1.6 to 68 ng/ mL (monoclonal), with a mean of 18 ng/mL (standard deviation, 16.5 ng/mL); 4 of 26 patients had PSA concentration below 4.0 ng/mL. Clinical stages were T2a+b NO MO (10 patients) and T2c NO MO (12
Fig. 2. Loss of basement membrane in prostatic adenocarcinoma (heparan sulfate proteoglycan immunostaining). In A, most of the malignant acini (Gleason pattern 3) retain an immunoreactive basement membrane. Compare with B, showing a focus containing only a few malignant acini (Gleason pattern 4) with an intact base-
ment membrane. A, B: x200 (heparan sulfate proteoglycan immunostaining) .
Basement Membrane in PIN· 853
patients). All patients underwent radical retropubic prostatectomy with open bilateral pelvic lymphadenectomy, as described elsewhere l4, 15. The fresh prostates weighed from 25 to 90 grams, with a mean of 51 grams (standard deviation, 15.5 grams), and measured from 18 to 72 cc, with a mean of 36 cc (standard deviation, 12.1 cc). Pathologic stages included T2a+b NO MO(l patient), T2c NO MO(12 patients), T3a+b NO MO (5 patients), T3c NO MO (3 patients) and T3c N2 MO (1 patient). The prostate cancer volume ranged from 0.56 cc to 29.01 cc, with a mean of 8.32 cc (standard deviation, 7.25 cc). Gleason grade was score 5 (6 patients), 6 (6 patients), 7 (7 patients), 8 (2 patients) and 9 (1 patient). All tumors involved the peripheral zone, and 12 also involved the transition zone. DNA ploidy analysis showed that 20 were diploid and 2 were non-diploid. The number of high power fields of PIN chosen for immunohistochemical studies ranged from 1 to 49 fields per section, with a mean of 17.2 (standard deviation, 11.7) and a median of 15.5 fields. The number of fields of cancer ranged from 5 to 111 fields, with a mean of 35.8 (standard deviation, 27.5) and a median of 27 fields. Benign prostatic hyperplasia was present in 12 of 22 cases.
Basement Membrane Staining in Normal Epithelium, Nodular Hyperplasia and High Grade PIN
All normal and hyperplastic acini (100%) displayed intact HSPG-immunoreactive basement membrane staining except in areas disrupted by inflammation. Similarly, high grade PIN showed intact basement membranes in all but 5 cases which had 90% acini staining (overall mean percent staining, 98 % ) (Fig. 1). The mean intensity of HSPG-immunoreactivity on a 0-4 scale was 3.4 in normal epithelium (range,
U as CI>
Cancer grade (Gleason score)
Fig. 3. Heparan sulfate proteoglycan Immunoreactivity prostate cancer according to Gleason score.
2-4),3.6 in hyperplastic epithelium (range, 2-4) and 2.5 in PIN (range, 1-4).
Basement Membrane Staining in Prostate Cancer The mean percent of acinar staining in Gleason score 5 cancer was 100%, Gleason score 6 was 98% (range, 90-100%; standard deviation, 5.5%), Gleason score 7 was 82.9% (range, 70-90%; standard deviation, 7.6%), Gleason score 8 was 60% (range, 50-70%; standard deviation, 10.0%) and Gleason score 9 was 40% (Figs. 2, 3). The mean percent of cancer acini with HSPG-immunoreactive basement membranes was inversely proportional to Gleason score (p = 0.002). The intensity of basement membrane immunoreactivity on a 0-4 scale was similar in all Gleason scores (p = 0.77), with mean intensity of 3.3 (range, 2 to 4; standard deviation, 0.7). Discussion In this study, we evaluated the continuity of the acinar basement membrane in a series of radical prostatectomy specimens with high grade prostatic intraepithelial neoplasia (PIN) and adenocarcinoma. Our results indicate that PIN and well differentiated cancer maintain an intact basement membrane. Intense immunoreactivity for heparan sulfate proteoglycan (HSPG) was observed in the basement membrane of 100% of acini of normal and hyperplastic epithelium, 98% of acini with PIN, and 100% of acini with low grade (Gleason score 5) adenocarcinoma; only foci disrupted by inflammation failed to be stained. Using type IV collagen as a marker, Shultz et al. 6 found that the acinar basement membrane was patchy in 60% of cases of PIN and absent in 10%. We also found that there was a progressive loss of the basement membrane with dedifferentiation of prostatic adenocarcinoma, similar to the results of Fuchs et al. 5, Bonkhoff et al. 3 and Sinha et al. 9 • Well-differentiated adenocarcinoma displayed HSPG immunoreactivity in most acini, including all cases of Gleason score 5 and 6 cancer. The mean percent of malignant acini with immunoreactive basement membranes was inversely proportional to Gleason score. These findings are in agreement with others who employed type IV collagen immunohistochemistry and other basement membrane markers2-6, 8, 9. The basement membrane which surrounds benign prostatic acini has a mean thickness of 85.9 nanometers, well below the normal resolution of routine light microscopy5. Immunohistochemical stains utilizing antibodies directed against Type IV collagen and laminin demonstrate this thin basement membrane, but there is also staining around blood vessels and the abundant stromal smooth muscle fibers in the prostate, making interpretation of subtle changes in acini difficult6, 8, 11. To avoid the problem of smooth muscle fiber interference staining, we used an antibody direc-
854 . D. G. Bostwick et al.
ted against heparan sulfate proteoglycan, another extracellular matrix component which is found exclusively in acinar and blood vessel basement membranes. Heparan sulfate proteoglycan has been used successfully in other organs for basement membrane evaluation 13 , and does not decorate the basement membrane surrounding smooth muscle cells. The clinical and prognostic significance of the basement membrane in prostate cancer is uncertain. Destruction of the basement membrane in moderately and poorly differentiated adenocarcinoma may be essential for metastasis with some cancers, and may be prognostically valuable, but this has been refuted 16 • Bonkoff et aJ.2 identified type IV collagen-immunoreactive basement membranes around primary and metastatic prostate cancer, indicating that loss of this membrane was not essential for extraprostatic vascular spread. Growth and invasion of prostate cancer probably depends on interaction with and remodelling of the basement membrane and extracellular matrix. Cancer growth and invasion in vitro correlates with expression of urokinase, urokinase receptors 17, fibroblast growth factor 18 , epidermal growth factor 19 and alpha-3 Beta-l and alpha-6 Beta-4 integrins 2o • Invasive potential is inhibited by antagonists of urokinase 17 and the laminin cell adhesion peptide 18 • There is also downregulation of adhesion molecules such as E-cadherin in prostate cancer with increasing grade 21 , 22. Maintenance of an intact basement membrane by PIN and well-differentiated prostatic adenocarcinoma is further evidence linking these two histologic findings, and supports the hypothesis that PIN is a precursor to prostatic adenocarcinoma.
Material and Methods Case Selection Twenty-two consecutive cases of prostate cancer with high grade prostatic intraepithelial neoplasia (PIN) were chosen from a large series of systematically-sectioned wholemounted radical retropubic prostatectomies. Only cases with the following were selected for study: (1 ) no clinical history of androgen deprivation therapy (e.g., leuprolide, flutamide, or finasteride) or radiation therapy prior to radical prostatectomy; (2) a single whole mount tissue section was available which contained benign prostatic tissue, high grade PIN and abundant carcinoma; and (3) there was immunoreactivity for prostate specific antigen (PSA) in the benign epithelium, indicative of antigenic preservation. The radical prostatectomy specimens were handled by protocol. Briefly, the unfixed prostate and attached seminal vesicles were measured and weighed, inked and fixed for 24-72 hours in 10% neutral buffered formalin. The apex and base were amputated at 4-5 mm thickness, sectioned perpendicular to the inked surgical margins, and submitted in their entirety at 2 mm intervals. The prostate and seminal vesicles were sectioned with a sharp knife at 5 mm intervals perpendicular to the long axis from the apical end to the base.
Dehydration, clearing and paraffin embedding were performed routinely on a 24 hour cycle, and tissue sections were cut at 5 microns.
Immunohistochemical Studies The labeled streptavidin-biotin peroxidase method was used in combination with a monoclonal antibody directed against heparan sulfate proteoglycan (Chemicon Int, Inc., Temecula, CA; 1 : 100 dilution with trypsin predigestion)13. Cases were stained in a single batch to allow direct comparison and avoid inter-run variation. Known positive and negative controls were run in parallel and gave appropriate results; also, substitution of antibody by non immune serum served as a control. Vessels within each specimen served as internal positive control. Slides were evaluated by two observers (DGB, JQ) without knowledge of clinical features or other pathologic findings. Immunoreactivity was measured in normal acini, hyperplastic acini, high grade PIN and carcinoma in each case according to the percent of positive acini (in 10% increments) and the intensity of staining based upon mean immunoreactivity (0-4 scale) .
Statistics Kruskall Wallis nonparametric tests compared the percent of acini with HSPG immunoreactivity and the intensity of staining with Gleason score. Significance level was 5%.
References 1 Barsky SH, Siegal GP, Jannotta F, et al. (1983) Loss of basement membrane components by invasive tumors but not by their benign counterparts. Laboratory Investigation 49: 140-147 2 Bonkhoff H, Wernert N, Dhom G, Remberger K (1992) Distribution of basement membranes in primary and metastatic carcinomas of the prostate. Hum Pathol 23: 934-939 3 Bonkhoff H, Wernert N, Dhom G Remberger K (1991) Basement membranes in fetal, adult normal, hyperplastic and neoplastic human prostate. Virch Archiv [A] Pathol Anat Histopathol418: 375-381 4 BonkoffH, Stein U, Remberger K (1993) Differential expression of alph6 and alpha 2 very late antigen integrins in the normal, hyperplastic, and neoplastic prostate: Simultaneous demonstration of cell surface receptors and their extracellular ligands. Hum Pathol 24: 243-248 5 Fuchs ME, Brawer MK, Rennels MA, Nagle RB (1989) The relationship of basement membrane to histologic grade of human prostatic carcinoma. Mod Pathol 2: 105 -111 6 Schultz OS, Amin MB, Zarbo RJ (1993) Basement membrane Type IV collagen immunohistochemical staining in prostatic neoplasia. Appl Immunohistochemistry 1: 123 -126 7 Schamhart DHJ, Kurth KH (1994) Proteoglycans and glycosaminoglycans in tumor growth and migration: first experience with tumors of bladder and prostate origin. World J Uro112: 55-61 8 Sinha AA, Gleason DF, Wilson MJ et al. (1989) Immunohistochemical localization of laminin in the basement membranes of normal, hyperplastic, and neoplastic human prostate. Prostate 15: 299-313 9 Sinha AA, Gleason OF, DeLeon OF, Wilson M], Limas C, Reddy PK, Furcht LT (1991) Localization of type IV collagen in the basement membranes of human prostate and lymph
Basement Membrane in PIN· 855 nodes by immunoperoxidase and immunoalkaline phosphatase. Prostate 18: 93-104 10 Nagle RB, Knox]D, Wolf C, Bowden GT, Cress AE. Adhesion molecules, extracellular matrix, and proteases in prostate carcinoma (submitted) 11 Bostwick DG, Srigley ], Grignon D, Maksem], Humphrey P, van der Kwast T, Bose D, Harrison ], Young R (1993) Atypical adenomatous hyperplasia of the prostate: Morphologic criteria for its distinction from well-differentiated carcinoma. Hum Pathol. 24: 819-832 12 Kosir MA, Quinn CC (1995) Sorting of heparan sulfate proteoglycan into matrix compartments of prostate adenocarcinoma cells. ] Surg Res 58: 46-52 13 Kemeny E, Fillit HM, Damle S, et al. (1988) Monoclonal antibodies to hepar an sulfate proteoglycan: Development and application to the study of normal tissue and pathologic human kidney biopsies. Conn Tiss Res 18: 9-25 14 Bostwick DG, Myers RP, Oesterling]E (1994) The staging of prostate cancer. Seminars in Surgical Oncology 10: 6073 15 Blackwell KL, Bostwick DG, Myers RP, Zincke H, Oesterline ]E (1994) Combining prostate specific antigen with cancer and gland volume to predict more reliably pathological stage: The influence of prostate specific antigen cancer density.] Urol. 151: 1565-1570 16 Hidda ]-1, Matsuda T, Kitaoka M, Machidera N, Kubo R, Yasutomi M (1994) The role of basement membrane in colorectal cancer invasion and liver metastasis. Cancer 74: 592-598
17 Hoosein NM, Boyd DD, Hollas W], Mazar A, Henkin], Chung LW (1991) Involvement of urokinase and its receptor in the invasiveness of human prostatic carcinoma cell lines. Cancer Comm 3: 255-264 18 Passaniti A, Isaacs ]T, Haney]A, Adler SW, Cujdik T], Long PV, Kleinman HK (1992) Stimulation of human prostatic carcinoma tumor growth in athymic mice and control of migration in culture by extracellular matrix. Int ] Cancer 51: 318-324 19 Jarrard DF, Blitz BF, Smith RC, Patai BL, Rukstalis DB (1994) Effect of epidermal growth factor on prostate cancer cell line PC3 growth and invasion. Prostate 24: 46-53 20 Dedhar S, Saulnier R, Nagle R, Overall CM (1993) Specific alterations in the expression of alpha 3 beta 1 and alpha 6 beta 4 integrins in highly invasive and metastatic variants of human prostate carcinoma cells selected by in vitro invasion through reconstituted basement membrane. Clin Exp Metastasis 11: 391-400 21 Otto T Rembrink K, Goepel M, Meyer-Schwickerath M, Rubben H (1993) E-cadherin: a marker for differentiation and invasiveness in prostatic carcinoma. Urol Research 21: 359-362 22 Umbas R, Schalken ]A, Aalders TW, Carter BS, Karthaus HFM, Schaafsma HE, Debruyne FM], Isaacs WB (1992) Expression of the cellular adhesion molecule, E-cadherin, is reduced or absent in high grade prostate cancer. Cancer Research 52: 5104-5109
Received June 10, 1995 . Accepted July 4, 1995
Key words: Prostate cancer - Prostatic intraepithelial neoplasia - Benign prostatic hyperplasia - Basement membrane - Heparan sulfate proteoglycan - Type IV collagen - Smooth muscle - Stroma - Invasion David G. Bostwick M.D. Department of Pathology, Mayo Clinic, 200 First St. S.W., Rochester, MN 55905, USA, Phone: (507) 284-1870; FAX: (507) 284-1599