Prediction of pelvic lymph node metastasis by ... - Wiley Online Library

3141

Prediction of Pelvic Lymph Node Metastasis by the Ratio of Cathepsin B to Stefin A in Patients with Prostate Carcinoma Akhouri A. Sinha, Ph.D.1,4,5 Barry J. Quast, B.S.1 Michael J. Wilson, Ph.D.2,3,4,5 Eduardo T. Fernandes, M.D.3,5 Pratap K. Reddy, M.D.5 Stephen L. Ewing, M.D.2,5 Donald F. Gleason, M.D., Ph.D.2

BACKGROUND. Pathologic grade and/or histologic score, extraprostatic extension indicated by invasion of the prostatic capsule, margin, and/or seminal vesicles by prostate cancer cells, serum total prostate-specific antigen (PSA), free PSA, complexed PSA levels and/or their ratios, regional pelvic lymph node metastases, and clinical staging have been used to diagnose and monitor the treatment of prostate carcinoma (PC) patients. The Gleason grading system is also used to grade/score a patient’s stage of disease, with lower to higher scores indicating progression of PC. However, Gleason’s system cannot be used to distinguish biologically aggressive

1

Department of Genetics, Cell Biology and Development, University of Minnesota, Minneapolis, Minnesota.

PCs within a single Gleason score. Our objective was to identify subpopulations (or

2

stroma and metastasis to the lymph nodes.

Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota.

3

clones) of aggressive prostate cancers within an individual Gleason score by utilizing biological molecule(s) that also facilitate cancer cell invasion to prostatic

MATERIALS AND METHODS. Specimens were collected from 97 patients with PC and from 8 patients with benign prostatic hyperplasia. These patients had not been

Department of Urology and Urologic Surgery, University of Minnesota, Minneapolis, Minnesota.

treated with hormonal and/or chemotherapeutic agents before undergoing a pros-

4

Cancer Center, University of Minnesota, Minneapolis, Minnesota.

paraffin or paraplast-embedded prostate tissue sections were stained with hema-

5

for immunohistochemical study. We also collected data on age, race, extraprostatic

Research Service of the Veterans Affairs Medical Center, Minneapolis, Minnesota.

tatectomy at the Minneapolis Veterans Affairs Medical Center. Formalin-fixed, toxylin and eosin for pathologic diagnosis and adjacent sections were stained for extension, margin status, seminal vesicle, and lymph node invasion by cancer cells, clinical stage at prostatectomy, and mortality/survival data, including the available presurgery and postsurgery serum total PSA and prostatic acid phosphatase con-

Supported by USPHS Grant NIDDK-51348 (A.A.S.) and by the Research Service of the Minneapolis Veterans Affairs Medical Center (VAMC), which provided laboratory and office facilities to the first author.

centrations in patients. Immunohistochemical localization of mouse or rabbit

The authors are grateful to Dr. David G. Bostwick (Medical Director, Bostwick Laboratories, Richmond, VA) for his critical review of the manuscript. The authors also gratefully acknowledge the technical assistance of Mr. Trent Kowalchyk and the support of Ross Haller of the Urology Section of the VA, and the staff of the Surgical Pathology, Library, Medical Media, and Research Services of the Minneapolis VAMC.

within an individual Gleason score. Tumors with a Gleason Score of 6 that are

anti-cathepsin B (CB) antibody IgG and mouse antihuman stefin (cystatin) A IgG was quantified using a computer-based image analysis system equipped with Metamorph software.

RESULTS. CB and stefin A identified aggressive and less aggressive clones of PCs similar histologically and morphologically were heterogeneous with respect to the ratios of CB to stefin A (CB ⬎ stefin A, CB ⫽ stefin A, and CB ⬍ stefin A). We also found a significant positive association (P ⫽ 0.0066) between ratios of CB and stefin A (CB ⬎ stefin A) and the incidence of pelvic lymph node metastases, but not with ratios of CB less than stefin A and/or ratios of CB equal to stefin A. Patients with Gleason 7 PCs had a higher incidence of positive lymph nodes than those with Gleason Score 6 tumors. Our data indicated that mortality rates increased in

Address for reprints: Akhouri A. Sinha, Ph.D., Research Service, Veterans Affiars Medical Center (151), 1 Veterans Drive, Minneapolis, MN 55417; Fax: (612) 725-2093; E-mail: [email protected]

patients when the ratios of CB were greater than stefin A.

Received November 29, 2001; revision received December 16, 2001; accepted January 2, 2002.

be identified when the ratio of CB is greater than that of stefin A. Less aggressive

© 2002 American Cancer Society

CONCLUSIONS. PC within an individual Gleason score is a heterogeneous tumor that contains clones or subpopulations of aggressive and less aggressive tumors that can be defined by the ratios of CB to stefin A. PC with an aggressive clone can clones are identified when the ratio of CB is less than that of stefin A or when the

3142

CANCER June 15, 2002 / Volume 94 / Number 12 ratio of CB is equal to that of stefin A. The ratios of CB to stefin A can be used in the differential diagnosis and treatment of patients with PC. This is the first report to identify phenotypes of aggressive and less aggressive PCs within a Gleason score. Cancer 2002;94:3141–9. © 2002 American Cancer Society. DOI 10.1002/cncr.10604

KEYWORDS: prostate carcinoma, molecular markers, cathepsin B, stefin (cystatin) A, ratios of cathepsin B to stefin A, subpopulations of tumors within a Gleason score, aggressive and less aggressive clones, aggressive cancer with lymph node metastasis, less aggressive tumor without lymph node metastasis, differential diagnosis of cancer patients.

G

rade and/or histologic score, pathologic stage, extraprostatic invasion indicated by invasion of the prostatic capsule and margin, levels of serum total prostate-specific antigen (PSA), free PSA, and complexed PSA and/or their ratios, lymph node metastasis, and clinical stages are used to diagnose and monitor the treatment of prostate carcinoma patients.1–7 However, these prognostic factors are unable to predict precisely clinical outcomes in patients because prostate carcinoma is a complex and heterogeneous disease.8 –10 This heterogeneity was well recognized by Gleason in his grading system, which is based primarily on morphologic and histologiccriteria.2– 4 Although the Gleason grading system has the ability to predict progression of cancer using low to high scores/ grades,2,3,8,11–13 it cannot distinguish biologically aggressive and less aggressive clones or subpopulations of prostate carcinomas within a Gleason score/grade. For example, patients with a histologic score of 7–10 have a higher risk of dying of prostate carcinoma than those with lower scores, but there are patients with higher scores who live longer than some with lower scores.8 –10,14 –17 This indicates that there are biologically aggressive and less aggressive clones of prostate carcinomas within any one score/grade, confounding the ability to predict outcome for an individual patient. Cathepsin B (CB) is involved in the degradation of extracellular matrix proteins and the progression of cancer cells in patients with prostate carcinoma18 –21 and with other solid organ carcinomas.22–29 The CB level is elevated in prostate carcinoma patients compared with levels in individuals with a normal prostate and with patients with benign prostatic hyperplasia (BPH).18 –21 Lysosomal and plasma membrane/endosome subcellular fraction-associated CB activities are significantly higher in patients with malignancy than in those with BPH.21 This was further documented by immunogold electron microscopic localization of CB in these fractions.21 The enzymatic function of CB is regulated by endogenous inhibitors, the stefins (cystatins), in patients with malignant and nonmalignant

tumors,22,30,31 including patients with prostate carcinoma.19,32 Therefore, we evaluated CB and stefin A concurrently and in relation to the Gleason histologic score.19,32 We found that within each Gleason score, the ratio of CB to stefin A can be subdivided into three groups: CB greater than stefin A, CB less than stefin A, and CB equals stefin A. Our objective was to determine if a specific ratio of cathepsin B to stefin A in archival prostatectomy tissue samples by quantitative immunohistochemistry could predict aggressiveness of prostate carcinoma as indicated by metastasis of cancer cells to the regional pelvic lymph nodes. Our goal was to relate the phenotypes of aggressive prostate carcinomas with Gleason score, clinical stage, extraprostatic extension of cancer cells to the prostatic capsule, margin, and seminal vesicles, presurgery and postsurgery serum total PSA level, and/or mortality/ survival data of patients.

MATERIALS AND METHODS We collected formalin-fixed, paraffin or paraplast-embedded prostate tissue sections from 105 white veterans who had undergone radical prostatectomy for BPH (8 cases) and prostate carcinoma (97 cases) at the Minneapolis Veteran Affairs Medical Center (VAMC). We randomly selected 47 prostate samples from patients who had lymph node metastases and 50 samples from those without metastases. For pathologic diagnosis, 4 – 6-␮m sections were stained by hematoxylin and eosin (H & E) and adjacent sections were used in the immunohistochemical study. Prostate sections from 97 patients showed tumors with a Gleason Score of 4 –10 and samples from 8 patients showed BPH without any evidence of malignancy. Because only a few prostate carcinoma patients comprise other racial groups at the VAMC, nonwhite patients are the subjects of another study. Benign prostate and BPH tissue samples that were frozen previously in liquid nitrogen and/or on dry ice were used in the immunoblot study. All prostate samples were collected from prostatectomy patients who had not been treated preoperatively with hormonal and/or chemotherapeutic

Identification of Aggressive Prostate Cancer/Sinha et al.

3143

TABLE 1 Distribution of Prostate Carcinoma Patients by Gleason Histologic Scores Score 4–5

Score 6

Score 7

Score 8

Score 9–10

Total no. of samples

10

36

30

11

10

Age (range) Clinical stage TNM range Presurgery PSA (range) Postsurgery PSA Mean ⫾ SEM Median (range) Prostate-confined tumors Invasion of capsule/margin/seminal vesicle Unknown Lymph node positive Lymph node negative Total no. of Deaths (% within each score)

67.2 ⫾ 2.1 (57–70)

66.8 ⫾ 0.8 (54–75)

63.9 ⫾ 0.9 (55–71)

64.4 ⫾ 1.7 (52–69)

70.3 ⫾ 1.3 (64–77)

T1–3 20.0 ⫾ 19.5 (0.5–39)

T1–4 N1–3 M0 6.96 ⫾ 1.4 (1–25)

T1–4 N1–3 M0–1 23.2 ⫾ 7.0 (0.2–79)

T2–4 N1–3 M0–1 59.4 ⫾ 23 (15–146)

T1–4 N1–3 M0–1 10.1 ⫾ 3.9 (1–22)

0.34 ⫾ 0.09 0.16 (0.1–2.7) 5 5 0 2 8 6 (60%)

0.32 ⫾ 0.03 0.2 (0–3) 10 26 0 9 27 12 (33%)

20.1 ⫾ 9.7 0.5 (0.05–1420) 4 23 3 25 5 16 (53%)

220 ⫾ 115 2.8 (0.1–5880) 0 10 1 7 4 9 (82%)

1.85 ⫾ 0.7 0.15 (0.02–14) 1 9 0 5 5 8 (80%)

PSA: prostate-specific antigen.

agents. We also collected data on age, race, prostatic capsule status, margin status, and seminal vesicle and lymph node involvement by cancer cells, clinical stage of each patient at prostatectomy, and mortality/survival data. Many prostate carcinoma patients who had been treated at the VAMC died in other parts of the country and away from the VAMC. Their medical records were incomplete relative to cause of death due to prostate carcinoma and other causes. We also collected the available data on presurgery and postsurgery serum total PSA levels (measurements of serum total PSA began in 1987 at the VAMC). We collected data on presurgery serum PSA levels for 42 of 97 patients and prostatic acid phosphate (PAcP) levels for 49 of 97 patients. There were no PSA or PAcP data for 6 of 97 patients. We also collected postprostatectomy serum total PSA data for 66 of 97 patients whose PSA level was measured at intervals of 3 months or more. The postprostatectomy patients showing 0.2 ng/mL or more serum total PSA concentrations were considered to have recurrent disease as reported by others.7,33–36 There were no recorded measurements of PSA or PAcP levels for 31 postprostatectomy patients. Prostatic tissue samples and clinical data were collected according to the informed consent guidelines of the Minneapolis Veterans Affairs Human Studies Committee and the Institutional Review Board: Human Subjects Committee, University of Minnesota.

Gleason Grading System All tissue samples were graded according to the Gleason grading system by D. F. G and/or S. L. E.2,4 The prostate tissue samples for tumors of Gleason Scores 4 and 5 and Scores 9 and 10 were grouped because they

were limited to 20 specimens (Table 1). We assessed prostate carcinoma aggressiveness on H & E sections by recording extraprostatic invasion of the capsule, margin status, seminal vesicle, and lymph node by cancer cells. It is rare to find patients with Gleason Score 2 and 3 tumors. These patients usually do not receive surgical treatment at the VAMC and are not included in our study. We evaluated 97 prostate carcinoma samples showing Gleason Scores 4 –10 (Table 1).

Antibodies against CB and Stefin A Rabbit antihuman liver CB IgG (PC 41) was obtained from Oncogene Research Products (Calbiochem, Cambridge, MA) and mouse monoclonal antihuman stefin A antibody IgG was purchased from KRKA (Novo Mesto, Slovania). Both antibodies were affinity purified on immobilized Protein A by the manufacturer. Phosphate-buffered saline (PBS) and bovine serum albumin were obtained from Sigma (St. Louis, MO).

Molecular Weights of CB and Stefin A We determined the molecular weights of CB and stefin A in frozen cryostat prostate sections using Western blotting and the chemiluminescence techniques. Briefly, prostate sections were homogenized and extracted in RIPA (100 mL of 0.5 Tris-HCl, pH 7.4, 1.5 M NaCl, 2.5% deoxycholic acid, 10% NP-40, 10 mM EDTA at room temperature) lysis buffer as reported by Ikeda et al.37 Protein concentration was determined using the BioRad (Richmond, CA) protein assay kit. Following electrophoresis, gels were equilibrated in transfer buffer before electroblotting to nitrocellulose mem-

3144

CANCER June 15, 2002 / Volume 94 / Number 12

branes. These were then probed with 100 ng/mL of CB or stefin A antibody followed by sheep anti-mouse horseradish peroxidase (HRP)-labeled secondary antibody. The membranes were developed in chemiluminescent developing medium (Pierce, Rockford, IL). Blots were visualized and images were scanned with a UMAX Scanner.

Immunohistochemical Localization of CB and Stefin A We localized CB and stefin A in prostate tissue sections using immunohistochemical techniques.18,38 Briefly, rabbit anti-CB antibody IgG localized CB and mouse anti-human stefin A antibody IgG localized stefin A in adjacent sections. Negative control sections were incubated with preimmune rabbit or mouse serum in lieu of the primary antibodies. The reaction products were developed with fresh-filtered 3, 3⬘-diaminobenzidine (DAB) solution (0.25 mg/mL; Sigma) in PBS with 0.01% H2O2 as the substrate. Chromogenic development was viewed through a light microscope. Reaction products usually developed in less than 10 minutes.

Quantification of Localization Data by the Metamorph Image Analysis System Immunohistochemical reaction products for CB and stefin A were quantified using a computer-based image analysis system equipped with Metamorph software (Universal Imaging, West Chester, PA).19,20 Briefly, 10 –15 randomly selected images of reaction products for CB and stefin A in each section were acquired at a magnification of ⫻200 directly from the microscope slides to a computer using a digital camera (Photometrics, Tucson, AZ) attached to a Zeiss microscope. On the basis of gray values ranging from 4095 to 0, white to black, respectively, threshold boundaries of immunostaining were created. All immunostained objects included within the designated gray value range were expressed as a percentage of the total field area under view at the magnification of ⫻ 200. Statistical significance was determined using Student t test (P ⬍ 0.05).

RESULTS Profile of Prostate Carcinoma Patients The mean age of prostate carcinoma patients in this study was 65.8 ⫾ 0.51 years (SEM), with a range of 52–77 years at the time of prostatectomy surgery (Table 1). Of 97 patients, 10 (10.3%) had Gleason Score 4 and 5 tumors, 66 (68%) had Gleason Score 6 and 7 tumors, 11 (11.3%) had Score 8 tumors, and 10 (10.3%) had Score 9/10 tumors (Table 1). According to the TNM classification system,39 the clinical stages of tumors ranged from T1-3 to T1-4 N1-3 M0-1 (Table 1).

FIGURE 1.

The graph illustrates the relationship between recurrence of prostate carcinoma, as indicated by the elevation in serum concentration of prostate-specific antigen, in the months after prostatectomy and lymph node status for metastasis. Binomial distribution of the association of prostate carcinoma recurrence and positive lymph nodes was significant at 12 months (P ⫽ 0.0019), at 24 months (P ⫽ 0.0074), at 36 months (P ⫽ 0.0369), and at 48 months (P ⫽ 0.0350). The relationship between recurrence of cancer and lymph node status was not significant at 60 months (P ⫽ 0.6421). Statistical analysis was made using Student t test.

Serum total PSA level before the surgery ranged between 0.2 and 146 ng/mL with a mean of 21.5 ⫾ 5.6 (SEM; Table 1). There was no apparent relationship between serum PSA concentration and Gleason score and clinical stage (Table 1). The serum PSA concentration after prostatectomy varied more in Gleason Score 7 and 8 tumors than in Score 4 – 6 and 9 –10 tumors (Table 1). The likelihood of biochemical failure was greater in patients with lymph node metastases than in those without metastases (Fig. 1). The recurrence of PC was related to positive and negative nodes by binomial distribution. This was also shown by the increased serum PSA concentration up to 48 months after prostatectomy (Fig. 1). Statistical analysis indicated that the association of prostate carcinoma recurrence and positive lymph nodes was significant at 12 months (P ⫽ 0.0019), at 24 months (P ⫽ 0.0074), at 36 months (P ⫽ 0.0369), and at 48 months (P ⫽ 0.0350) when compared with data for negative lymph nodes (Fig. 1). The difference between the recurrence of prostate carcinoma and positive lymph nodes was not significant at 60 months (P ⫽ 0.6421). The recurrence of PC with negative nodes indicated the presence of cancer cells in other organs, as shown by the increase in serum


3145

21 to 31 kilodaltons (kD), which was similar to that reported for other tissue samples22,40 (Fig. 2a). The molecular weight of stefin A was about 11 kD, a molecular size similar to that reported by Kuopio et al.41 (Fig. 2b). No other proteins cross reacted with these antibodies, indicating the specificity of the antibodies used in our immunohistochemical study.

Immunostaining of CB and Stefin A in BPH Glands Immunohistochemical study of CB and stefin A proteins showed localization predominantly in basal cells and in some cuboidal/columnar cells of hyperplastic glands (Fig. 3a,b). The ratio of CB to stefin A was about 1:1 in hyperplastic glands, a ratio similar to that found for mRNAs evaluated by in situ hybridization.19

FIGURE 2. (a) Immunoblot of human prostate carcinoma extract probed with a monospecific polyclonal antibody against human liver cathepsin B (CB) showed that CB had a molecular weight of about 30 kilodaltons (kD). The molecular weight for mouse IgG was 145 kD. The specificity of the antibody used in the immunohistochemical study is indicated by its reactivity with only CB. (b) Immunoblot of human prostate carcinoma extract showed that the molecular weight of stefin A was about 11 kD. The specificity of this antibody is indicated by its reactivity with only stefin A. PSA 12 to 60 months after prostatectomy. Statistical analysis was performed using Student t test. Prostate carcinoma patients who had prostatectomy before the era of measuring serum PSA levels were assessed by presurgery PAcP levels that ranged from 1 to 5 ng/mL, with a mean of 1.99 ⫾ 0.09 and a median of 2. We did not evaluate PAcP level in relation to CB and stefin A levels because PAcP data after prostatectomy are not available in our patients. Our data (Table 1) showed that extraprostatic invasion indicated by the invasion of the prostatic capsule, margin, and seminal vesicle by cancer cells had occurred in 73 of 97 patients (75%), not in 20 of 97 patients (21%). Status was unknown for 4 of 97 patients (4%). These data provided an assessment of aggressive and less aggressive clones of prostate carcinomas. Gleason Score 7 tumors had the highest likelihood (24.7%, 24 of 97) of lymph node metastases than any other scores (Table 1). The highest likelihood of lymph node-negative tumors (27.8%, 27 of 97) was found in Gleason Score 6 tumors (Table 1). The number of deaths among prostate carcinoma patients increased as tumors were scored from lower to higher on the Gleason scale (Table 1).

Immunoblots The molecular weights of CB in prostatic tissue samples as determined by immunoblotting ranged from

Localization of CB and Stefin A in Malignant Prostate Tissue The distribution of reaction products for immunohistochemical localization of CB and stefin A proteins in formalin-fixed, paraffin-embedded prostate sections was similar to that for their proteins and mRNA in frozen prostate tissue sections.19,32 Immunostaining of CB and stefin A showed considerable variation within Gleason Score 6 tumors, which are considered to be similar histologically and morphologically (Fig. 3c–f). Likewise, Gleason Score 4 –5, 7, 8, and 9 –10 tumors showed variation in reaction products for CB and stefin A (Table 2). Our analysis of CB and stefin A distribution showed that the CB level was significantly (P ⫽ 0.0155) increased in primary prostate tumors of patients with positive lymph nodes, compared with those with negative lymph nodes (Fig. 4). Stefin A was not associated with tumor-positive or negative lymph nodes (P ⫽ 0.2439; Fig. 4). Linear regression analysis of the CB or stefin A data did not show a correlation with Gleason score tumors (R2 ⫽ 0.0016, R2 ⫽ 0.0001, respectively), clinical stage (R2 ⫽ 0.0089, R2 ⫽ 0.0031, respectively), and presurgery serum PSA level (R2 ⫽ 0.0001, R2 ⫽ 0.0133, respectively).

Ratios of CB to Stefin A in Neoplastic Prostate Samples Analysis of immunostaining distribution for CB to stefin A showed three distinct ratios: CB greater than stefin A, CB less than stefin A, and CB equals stefin A within Gleason score tumors (Table 2). This finding was similar to that reported previously for mRNA distribution evaluated by in situ hybridization and by immunofluorescence and confocal microscopy.19,32 Gleason Score 6 tumors are similar histologically and morphologically and showed three ratios of CB to stefin A (Table 2, Fig. 3c–f). The mean value of the ratio of CB to stefin A correlated with Gleason scores (Table 2). Regression analysis did not show a linear relation-

3146


FIGURE 3. (a) Micrograph shows that the immunohistochemical localization for cathepsin B (CB) was predominant in the basal cells of glands with benign prostatic hyperplasia (BPH) with some localization in columnar cells. (b) Micrograph illustrates immunostaining of stefin A in basal and some columnar cells of BPH glands. The ratio of CB to stefin A was about 1:1 (CB ⫽ stefin A) as indicated by their reaction products. (c) Micrograph illustrates predominant immunostaining for CB in a Gleason Score 6 tumor. (d) An adjacent section to that in Figure 3c illustrates markedly less immunostaining for stefin A. Comparison of c and d shows that this Gleason Score 6 tumor had a ratio of CB greater than stefin A. (e) Micrograph illustrates that immunostaining for CB was significantly lower in a Gleason Score 6 tumor than that of CB shown in Figure 3c. (f) An adjacent section to that in Figure e illustrates predominant immunostaining for stefin A when compared with CB in Figure 3e. The ratio of CB less than stefin A is shown in Figure 3e and f.

TABLE 2 Immunohistochemical Distribution of Cathepsin B (CB), Stefin A, and Their Ratios in Gleason Score 4 to 10 Prostate Carcinomas Immunostaining Total no. of samples

Score 4–5 10

Score 6 36

Score 7 30

Score 8 11

Score 9–10 10

CB (range) Stefin A (range) CB ⬎ Stefin A CB ⬍ Stefin A CB ⫽ Stefin A BPH CB alone Stefin A alone CB/stefin A ratio

6.66 ⫾ 2.6 (0.7–28) 3.55 ⫾ 0.9 (0.3–10) 7.92 ⫾ 4.7 (n ⫽ 4) 0.49 ⫾ 0.05 (n ⫽ 3) 1.05 ⫾ 0.05 (n ⫽ 3)

5.15 ⫾ 1.1 (0.4–26) 2.26 ⫾ 0.5 (0.6–11) 5.84 ⫾ 1.9 (n ⫽ 20) 0.52 ⫾ 0.08 (n ⫽ 6) 1.26 ⫾ 0.09 (n ⫽ 10)

4.69 ⫾ 1.0 (0.2–23) 3.35 ⫾ 0.6 (0.5–12) 3.93 ⫾ 0.8 (n ⫽ 15) 0.39 ⫾ 0.07 (n ⫽ 10) 1.33 ⫾ 0.09 (n ⫽ 5)

5.61 ⫾ 1.2 (0.8–13) 2.53 ⫾ 0.6 (0.2–6) 6.46 ⫾ 1.9 (n ⫽ 7) 0.63 ⫾ 0.01 (n ⫽ 3) 1.16 (n ⫽ 1)

6.42 ⫾ 1.6 (0.8–15) 2.80 ⫾ 0.6 (0.9–6) 4.54 ⫾ 1.4 (n ⫽ 5) 0.34 ⫾ 0.3 (n ⫽ 2) 1.44 ⫾ 0.17 (n ⫽ 3)

3.55 ⫾ 0.34 4.02 ⫾ 0.8 0.88 ⫾ 0.20

BPH: benign prostatic hyperplasia.

ship of the ratio of CB to stefin A to Gleason score (R2 ⫽ 0.009), clinical stage (R2 ⫽ 0.0014), and presurgery serum PSA level (R2 ⫽ 0.0071). Our data showed a significant association (P ⫽ 0.0066) between the ratio of CB to stefin A (CB ⬎ stefin A) and incidence of

pelvic lymph node metastasis. No relationship with lymph node metastases was found when the ratio of CB was less than stefin A and/or when the ratio of CB was equal to stefin A (Fig. 5). Patients with Gleason 7 prostate carcinomas had a higher incidence of positive


3147

larger sample of patients for a more definitive conclusion.

DISCUSSION

FIGURE 4. Graph illustrates the percent of immunostained areas for cathepsin B (CB) and stefin A in prostate tumors as related to positive and negative lymph nodes. CB was significantly higher (P ⫽ 0.0155) in tumor- positive lymph nodes than in negative lymph nodes. Differences in immunostaining for stefin A were not significant (P ⫽ 0.2439) in prostate tumors with positive and negative lymph nodes. All prostate samples (97 of 97) were examined for this study. Statistical analysis was conducted using Student t test.

FIGURE 5. Graph shows the ratio of cathepsin B (CB) to stefin A- immunostained areas as related to positive and negative lymph nodes. A significantly higher (P ⫽ 0.0066) CB:stefin A ratio was found in prostate carcinoma patients with tumor-positive lymph nodes than with negative lymph nodes. All prostate samples (97 of 97) were examined for this study. Statistical analysis was conducted using Student t test. lymph nodes than patients with Gleason Score 6 or 8 tumors (Tables 1, 2). Mortality rates increased among patients with higher Gleason scores and in patients in whom the ratio of CB was greater than stefin A. However, there was no increase in mortality rates among patients in whom the ratio of CB was less than stefin A or the ratio of CB was equal to stefin A (Table 1). Mortality data, however, remains to be evaluated in a

Increased levels of CB are associated with degradation of basement membrane proteins and invasion of the adjacent stroma in many solid organ tumors.22–29 Our data, which show greater levels of CB than its inhibitor, stefin A, support this conclusion. We have also provided evidence that immunostaining for CB in prostate carcinoma tissue sections, shown by the ratio of CB greater than stefin A, was associated with a significantly higher (P ⬍ 0.0066) level of pelvic lymph node metastases in cancer patients, but not in patients in whom the ratio of CB was less than stefin A (P ⫽ 0.2347). There was a significant (P ⫽ 0.0155) association of CB expression with lymph node metastasis, but the sensitivity of the ratio of CB greater than stefin A ratio was greater than that of CB alone. This association is further supported by earlier studies of CB levels in patients with prostate carcinoma.18,19,32 We have shown that the mature (active) form of CB was more often associated with malignancy than its proenzyme (inactive) form, which was associated with a benign prostate and BPH.20 In addition, CB activities were significantly increased in plasma membrane fractions of prostate carcinoma patients when compared with activities in patients with BPH.21 The ratio of CB equals stefin A suggests a homeostatic balance between the enzyme and its endogenous inhibitor, stefin A, because it is usually found in the same cell types of benign prostate and BPH. This ratio of CB to stefin A in any Gleason score may be indicative of less aggressive subpopulations of prostate cacinomas than in populations where the ratios of CB are greater than stefin A. The ratio of CB less than stefin A suggests predominance of the inhibitor over its enzyme. Therefore, this ratio found in any Gleason score may be indicative of decreased participation of CB in invasive processes in the prostate, much as it was reported for other solid organ tumors. Prostate carcinomas with ratios of CB equals stefin A or CB less than stefin A, regardless of Gleason score, may have low metastatic potentials. Thus, the current study and our previous work on the mature form of CB and subcellular localization of its activities have shown that that the ratios of CB greater than stefin A are linked to increased lymph node metastases and mortality rates. Both increased lymph node metastases and mortality rates are indicative of the aggressiveness of prostate carcinoma in patients. These ratios can be used to identify clones of aggressive and less aggressive prostate carcinomas within a Gleason score tumor. Heterogeneity is well recognized by Gleason in his

3148


grading system2– 4 and by our study of untreated and hormonally treated prostate carcinoma cases.42 This heterogeneity is, however, not identified within Gleason Score6 tumors by histologic and morphologic criteria. We have shown that heterogeneity within a Gleason score tumor can be defined by the ratios of CB to stefin A proteins and their mRNAs.19 Characterization of heterogeneity by CB and stefin A as molecular markers has the advantage of defining the relative aggressiveness of prostate carcinomas within a Gleason score tumor. The ratios of CB to stefin A can distinguish aggressive and less aggressive subpopulations of prostate carcinomas within a Gleason score tumor because the ratios are linked to lymph node metastases and mortality rates in prostate carcinoma patients. In contrast, Gleason had shown that prostatic malignancy increased with the increasing Gleason histologic score or Gleason grade.2– 4 Thus, the presence of heterogeneity within any Gleason histologic score or grade can be defined by the ratios of CB to stefin A. Many criteria (such as increasing prostatic malignancy from lower to higher Gleason scores, invasion of the prostatic capsule, margin, and seminal vesicle by cancer cells, postprostatectomy elevation in serum total PSA level indicative of recurrence of the disease, positive pelvic lymph nodes, and clinical stages) have been used in the diagnosis and prognosis of prostate carcinoma patients. These criteria have also provided estimates of aggressiveness of prostate carcinomas, but these estimates have not predicted reliably the clinical course of malignancy and/or outcome of the disease as related to Gleason scores. For example, some patients with lower Gleason score tumors may die before those with higher scores, whereas others with higher scores may live longer than those with lower scores.8 –10,14 –17 This indicates that the existing criteria for predicting aggressiveness of prostate carcinomas are at the best incomplete. Prediction of aggressiveness of prostate carcinomas may be further improved by using the ratios of CB to stefin A in relation to Gleason scores and lymph node metastases. Criteria of tumor aggressiveness, indicated by invasion of the prostatic capsule/margin/seminal vesicle, and postprostatectomy recurrence of the disease can be supplemented readily with the ratios of CB to stefin A. Our study is limited by a small population (105 prostate tissue samples) compared with large samples evaluated by Gleason.2,3 We envision the ratios of CB to stefin A to have the potential of identifying subpopulations of aggressive and less aggressive prostate carcinomas within a single Gleason score. In conclusion, aggressive and less aggressive clones of prostate carcinomas exist within an individ-

ual Gleason score and they can be defined by the ratios of CB to stefin A. More aggressive clones of prostate carcinomas will be identified by a ratio of CB greater than stefin A, whereas less aggressive clones will be identified by ratios of CB less than stefin A or equal to stefin A. Identification of aggressiveness of prostate carcinomas within a Gleason score may be a refinement to the well established Gleason grading system. Our study needs to be expanded to include a large number of retrospective and prospective prostatectomy samples together with other prognostic factors, such as Gleason grading system, clinical stages, lymph node status, and mortality rates. In the interim, we suggest that patients with aggressive clones of prostate carcinomas should be treated aggressively, whereas patients with less aggressive tumors could be followed by waitful watching (managed conservatively) as suggested by many clinicians.43– 46 Our approach of identifying aggressive and less aggressive clones of prostate carcinomas by the ratios of CB to stefin A can be extended to other solid organ tumors.

REFERENCES 1.

Bostwick DG, Grignon DJ, Hammond EH, et al. Prognostic factors in prostate cancer. Arch Pathol Lab Med. 2000;124: 995–1000. 2. Gleason DF. Classification of prostatic carcinomas. Cancer Chemother Rep. 1966;50:125–128. 3. Gleason DF, VACURG. Histologic grading and clinical staging of prostatic carcinoma. In: Tannenbaum M, edior. Urologic pathology: the prostate. Philadelphia: Lea & Febiger, 1977:171–213. 4. Gleason DF. Histologic grading of prostate cancer. Hum Pathol. 1992;23:273–279. 5. Stamey TA, Sozen TS, Yemoto CM, McNeal JE. Classification of “localized” untreated prostate cancer based on 791 men treated only by radical prostatectomy: common ground for therapeutic trials and TNM subgroups. J Urol. 1998;159: 2009 –2012. 6. Smith CV, Bauer JJ, Connelly RR, et al. Prostate cancer in men age 50 years or younger: a review of the department of defense center for prostate disease research multicenter prostate cancer data base. J Urol. 2000;164:1964 –1967. 7. Pound CR, Partin AW, Eisenberger MA, Chan DW, Pearson JD, Walsh PC. Natural history of progression after PSA elevation following radical prostatectomy. JAMA. 1999; 281:1591–1597. 8. Albertsen PC, Hanley JA, Gleason DF, Barry MJ. Competing risk analysis of men aged 55 to 74 years at diagnosis managed conservatively for clinically localized prostate cancer. JAMA. 1998;280:975–980. 9. Hugosson J, Aus G, Bergdahl C, Berghahl S. Prostate cancer mortality in patients surviving more than 10 years after diagnosis. J Urol. 1995;154:2115–2117. 10. Pisters LL. The challenge of locally advanced prostate cancer. Semin Oncol. 1999;26:202–216. 11. Stamey TA, McNeal JE, Yemoto CM, Sigal BM, Hohnstone IM. Biological determinants of cancer progression in men with prostate cancer. JAMA. 1999;281:1395–1400.

Identification of Aggressive Prostate Cancer/Sinha et al. 12. Veltri RW, Miller MC, Partin AW, Coffey DS, Epstein JI. Ability to predict biochemical progression using Gleason score and a computer-generated quantitative nuclear grade derived from cancer cell nuclei. Urology. 1996;48:685– 691. 13. Epstein JI. Pathology of prostatic intraepithelial neoplasia and adenocarcinoma of the prostate: prognostic influences of stage, tumor volume, grade, and margins of resection. Semin Oncol. 1994;21:527–541. 14. Pisansky TM, Kahn MJ, Bostwick DG. An enhanced prognostic system for clinically localized carcinoma of the prostate. Cancer. 1997;79:2154 –2161. 15. Hegarty NJ, Fitzpatrick JM, Richie JP, et al. Future prospects in prostate cancer. Prostate. 1999;40:261–268. 16. Millikan RE. Chemotherapy of advanced prostatic carcinoma. Semin Oncol. 1999;26:185–191. 17. Kamradt JM, Pienta KJ. Novel molecular targets for prostate cancer therapy. Semin Oncol. 1999;26:234 –243. 18. Sinha AA, Wilson MJ, Gleason DF, Reddy PK, Sameni M, Sloane BF. Immunohistochemical localization of cathepsin B in neoplastic human prostate. Prostate. 1995;26:171–178. 19. Sinha AA, Quast BJ, Korkowski JC, et al. The relationship of cathepsin B and stefin A mRNA localization identifies a potentially aggressive variant of human prostate cancer within a Gleason histologic score. Anticancer Res. 1999;19: 2821–2830. 20. Sinha AA, Quast BJ, Wilson MJ, Reddy PK, Gleason DF, Sloane BF. Co-distribution of pro and mature cathepsin B forms in human prostate tumors detected by confocal and immunofluorescence microscopy. Anat Rec. 1998;252:281– 289. 21. Sinha AA, Jamuar MP, Wilson MJ, Rozhin J, Sloane BF. Plasma membrane association of cathepsin B in human prostate cancer: biochemical and immunogold electron microscopic analysis. Prostate. 2001;49:172–184. 22. Sloane BF, Moin K, Krepela E, Rozhin J. Cathepsin B and its endogenous inhibitors: role in tumor malignancy. Cancer Metastasis Rev. 1990;9:333–352. 23. Frohlich E, Schaumbrug-Lever S, Klessen C. Immunocytochemical and immunoelectron microscopic demonstration of cathepsin B in human malignant melanoma. Br J Dermatol. 1995;132:867– 875. 24. Werle B, Lotterle H, Schanzenbacher U, et al. Immunochemical analysis of cathepsin B in lung tumours: an independent prognostic factor for squamous cell carcinoma patients. Br J Cancer. 1999;81:510 –519. 25. Frosch BA, Berquin I, Emmert-Buck MR, Moin K, Sloane BF. Molecular regulation, membrane association and secretion of tumor cathepsin B. APMIS. 1999;107:28 –37. 26. Chapman HA, Riese RJ, Shi G-P. Emerging roles for cysteine proteases in human biology. Annu Rev Physiol. 1997;59:63– 88. 27. Yan S, Sameni M, Sloane BF. Cathepsin B and human tumor progression. Biol Chem. 1998;379:113–123. 28. Moin K, Cao L, Day NA, Koblinski JE, Sloane BF. Tumor cell membrane cathepsin B. Biol Chem. 1998;379:1093–1099. 29. Berquin IM, Sloane BF. Cathepsin B expression in human tumors. Adv Exp Med Biol. 1996;389:281–294. 30. Koblinski JE, Sloane BF. Is altered localization of cathepsin B

31.

32.

33.

34.

35. 36.

37.

38.

39. 40. 41.

42.

43.

44.

45.

46.

3149

casually related to malignant progression? In: Katunuma H, Kido H, Fritz H, Travis J, editors. Medical aspects of proteases and protease inhibitors. Amsterdam: IOS Press, 1997: 185–194. Sloane BF, Moin K, Sameni M, Tait LR, Rozhin J, Ziegler G. Membrane association of cathepsin B can be induced by transfection of human breast epithelial cells with c-Ha-ras oncogene. J Cell Sci. 1994;107:373–384. Sinha AA, Quast BJ, Wilson MJ, et al. The ratio of cathepsin B to stefin A identifies heterogeneity within Gleason histologic scores for human prostate cancer. Prostate. 2001;48: 274 –284. Dhanasekaran SM, Barrrette TR, Ghosh D, et al. Delineation of prognostic biomarkers in prostate cancer. Nature. 2001; 412:822– 826. Lange PL, Ercole CJ, Lightner DJ, Fraley EE, Vessela R. The value of serum prostate specific antigen determinations before and after radical prostatectomy. J Urol. 1989;141:873– 879. Bunting PS. A guide to the interpretation of serum prostate specific antigen levels. Clin Biochem. 1995;28:221–241. Carter HB, Partin AW, Epstein JI, Chan DW, Walsh PC. The relationship of prostate specific antigen levels and residual tumor volume in stage A prostate cancer. J Urol. 1990;144: 1167–1171. Ikeda K, Monden T, Kanoh T, et al. Extraction and analysis of diagnostically useful proteins from formalin-fixed, paraffin-embedded tissue sections. J Histochem Cytochem. 1998; 46:397– 403. Sinha AA, Gleason DF, Limas C, et al. Localization of cathepsin B in normal and hyperplastic human prostate by immunoperoxidase and protein A gold techniques. Anat Rec. 1989;223:266 –275. Ellis WJ, Lange PH. Prostate cancer. Endocrinol Metab ClinNorth Am 1994;23:809 – 824. Sloane BF. Cathepsin B and cystatins: evidence for a role in cancer progression. Semin Cancer Biol. 1990;1:137–152. Kuopio T, Kankaaranta A, Jalava P, et al. Cysteine prtoteinase inhibitor cystatin A in breast cancer. Cancer Res. 1998; 58:432– 436. Sinha AA, Blackard CE, Seal US. A critical analysis of tumor morphology and hormone treatments in the untreated and estrogen-treated responsive and refractory human prostatic carcinoma. Cancer. 1977;40:2836 –2850. Steinberg GD, Bales GT, Brendler CB. An analysis of watchful waiting for clinically localized prostate cancer. J Urol. 1998;159:1431–1436. Chodak GW, Thisted RA, Gerber GS, et al. Results of conservative management of clinically localized prostate cancer. N Engl J Med. 1994;330:242–248. Albertsen PC, Fryback DG, Storer BE, Kolon TF, Fine J. Long-term survival among men with conservatively treated localized prostate cancer. JAMA. 1995;274:626 – 631. Johansson JE, Holmberg L, Johansson S, Bergstrom R, Adami HO. Fifteen-year survival in prostate cancer. A prospective, population-based study in Sweden. JAMA. 1997; 277:467– 471.