Clinical Cancer Research Bridging the Lab and the Clinic in Cancer Medicine Infection and Cancer: Biology, Therapeutics, and Prevention
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kuniyasu, H.
Right arrow Articles by Pettaway, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kuniyasu, H.
Right arrow Articles by Pettaway, C. A.
Clinical Cancer Research Vol. 9, 2185-2194, June 2003
© 2003 American Association for Cancer Research

Molecular Oncology, Markers, Clinical Correlates

The Relative mRNA Expression Levels of Matrix Metalloproteinase to E-Cadherin in Prostate Biopsy Specimens Distinguishes Organ-confined from Advanced Prostate Cancer at Radical Prostatectomy1

Hiroki Kuniyasu, Rinzo Ukai, Dennis Johnston, Patricia Troncoso, Isaiah J. Fidler and Curtis A. Pettaway2

Department of Oncological Pathology, Cancer Center, Nara Medical University, Kashihara, Japan [H. K.]; JR Hiroshima General Hospital of West Japan Railway Company, Hiroshima, Japan [R. U.]; and the Departments of Biomathematics [D. J.], Pathology [P. T.], Cancer Biology and [I. J. F., C. A. P.], and Urology [C. A. P.], The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Purpose: To determine whether the expression ratio of matrix metalloproteinase (MMP) to E-cadherin mRNA (MMP:E-cadherin) in biopsy (BX) samples of prostate cancer correlate with that of radical prostatectomy (RP) specimens and assists in predicting pathologic stage.

Experimental Design: The mRNA expression levels for MMP-2 and -9 and of E-cadherin were determined by a colorimetric in situ hybridization assay in 44 paired BX and RP specimens. Clinical stage, BX Gleason score (GS), total length of cancer in BX cores, and serum prostate-specific antigen levels were also assessed.

Results: Clinical stage was confined in 39 of 44 (89%) patients. Subsequent to RP, however, only 17 of 44 (39%) patients had proven organ-confined disease (pT2). BX GSs agreed with the RP GS in 77% of RP specimens. We found a strong correlation between BX and RP MMP:E-cadherin ratios (correlation coefficient = 0.755). The ratio increased as the GS increased and pathologic stage advanced (pT2 versus ≥ pT3). Increasing clinical stage, GS, and serum prostate-specific antigen were significantly associated with advanced cancer at RP (P = 0.004–0.0001). The BX MMP:E-cadherin ratio, however, exhibited the strongest association with pathologic stage and independently predicted the status of 89% of the RP cases based on a BX ratio of <6 (predicted stage pT2 cancer at RP) versus ≥6 (predicted stage ≥ pT3 at RP).

Conclusion: The BX MMP:E-cadherin ratio represents a novel prognostic assay for predicting stage of cancer at RP. These data provide proof of principle for directly assessing the biological potential of prostate cancer using molecular strategies in patient’s specimens.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
To date, surrogate assessments predicting the extent of disease and the biology of prostate cancer are provided by digital palpation, radiological imaging studies, serum PSA,3 the BX GS, and other additional tumor-related information gained from sextant prostate BX strategies. Although imprecise, these assessments provide clinicians with the ability to discern "ballpark estimates" of the likelihood of organ-confined versus advanced cancer, as well as freedom from PSA progression at 5 years when treated with RP or radiotherapy (1, 2, 3, 4, 5) .

The process of metastasis involves a series of sequential steps regulated by a variety of different genes (6) . Considering the growing body of evidence available with respect to the molecular characteristics of prostate cells that are capable of metastasis, relatively little has been done to define the use of molecular markers in clinical practice as adjuncts to clinical staging (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17) . Some examples of specific genes and pathways studied thus far in patient specimens include Ki-67 expression (cell proliferation), Bcl-2 and its family members (cell survival), mutant p53 protein, E-cadherin (cell cohesion), collagenases (degradation of basement membranes and invasion), tumor angiogenesis (angiogenesis factor expression and microvessel density counts), and apoptosis (via tunnel assay; Refs. 8 , 11 , and 15, 16, 17 and reviewed in Ref. 18 ). The use of DNA ploidy and nuclear morphometry has also been described (19 , 20) . The relevant question that remains, however, is can such markers of progression and metastasis provide independent information with respect to staging and prognosis in prostate cancer for the individual patient?

Previously, we characterized several genes associated with early events in prostate cancer progression by describing the invasive and angiogenic capacities of human prostate cancer first in animal models and subsequently in patient specimens (7 , 8 , 12 , 14 , 21) . Subsequent to in vivo studies in nude mice, we evaluated the expression of several genes in human specimens to validate our animal model and determine their significance in our own data set. The mRNA expression levels of MMPs, E-cadherin, and vascular endothelial growth factors in RP specimens were assessed using a novel ISH technique in archival specimens (21) . This technique allowed the identification of focal intense areas of gene expression within individual tumors (intratumoral heterogeneity) that would not necessarily be demonstrated by assessing average tissue levels (i.e., Northern blot analysis). Of importance, the technique also determined differences in gene expression among tumors of the same histology that were either organ confined or advanced (largely the GS 7 category). Thus, it appeared to enhance the value of histology.

Increasing expression of MMPs and vascular endothelial growth factor with decreasing E-cadherin expression in prostatectomy specimens characterized the histological progression of the disease in the prostate with respect to grade and pathologic stage (21) . The relative expression of MMP-2/-9 to E-cadherin detected by ISH MMP:E-cadherin ratio) was an excellent marker for the metastatic potential of prostate cancer and the strongest factor associated with organ-confined versus advanced cancer (21) .

We have also shown previously that the MMP:E-cadherin ratio detected by ISH closely correlates with disease progression of pancreas (22) , colon (23) , gastric (24) , and lung cancers (25) . In these studies, the inverse correlation between low E-cadherin and high type IV collagenase (MMP-2/-9) expression and tumor progression appeared to quantify the relationship between factors tending to inhibit cell migration and invasion to those promoting the same steps in metastasis. In fact, an expression ration of ≥6 in prostatectomy specimens was always associated with extraprostatic extension of cancer or with metastasis (21) .

In the present study, we extend our previous observations by assessing the expression of MMPs (2 , 9) and E-cadherin in pretherapy prostate BXs and their corresponding prostatectomy specimens. This relationship is relevant in that prostate cancer examined in prostatectomy specimens showed considerable heterogeneity of gene expression within the same tumor, especially in advanced cases (21) . Data from the present study revealed a strong correlation for the MMP:E-cadherin expression ratio between BX and prostatectomy specimens. Furthermore, the BX MMP:E-cadherin ratio was highly predictive of organ-confined versus advanced prostate cancer subsequent to RP.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patient Characteristics and Histopathology.
Formalin-fixed, paraffin-embedded paired RP and BX specimens from 44 randomly selected patients untreated previously of The University of Texas M. D. Anderson Cancer Center (n = 13, years 1994–1997) and JR Hiroshima General Hospital of West Japan Railway Company (n = 31, years 1996–1999) were examined. The presence, extent, and grade of cancer were determined for both BX and RP specimens by examining 4-µm-thick tissue sections stained for H&E light microscopy. For the Japanese cohort (n = 31), the total length of cancer in all BX cores (millimeter) was measured, subjected to gene expression analysis, and included in the statistical evaluation. In the M. D. Anderson Cancer Center cohort, gene expression analysis was limited to one of the biopsies with the highest GS, and therefore, total tumor length in the BX was not included in the statistical evaluation. For RP specimens, the number of tumors present and the location, GS, and pathologic stage of the individual tumor foci were assessed. The highest GS of an individual BX or tumor focus in the RP specimen was the value assigned to the BX and RP specimen, respectively. Pathologic stage was assigned using the 1997 Tumor-Node-Metastasis staging system (26) as follows: (a) pT2 = organ-confined cancer; (b) pT3a = extraprostatic extension; and (c) pT3b = seminal vesicle invasion. As all patients exhibiting pelvic lymph node metastasis at surgery exhibited extraprostatic extension in the primary tumor, stage was assigned as pT3a or b and N+. Clinical stage was assigned using the 1992 Tumor-Node-Metastasis staging system (27) . Serum PSA levels were assessed using the Bayer Assay (Bayer Diagnostics, Redmond, WA) for the cohort from Japan and the Tosoh AIA Assay (Tosoh Medics, Foster City, CA) for the M. D. Anderson Cancer Center cohort. The use of radiological imaging studies varied by center and physician; however, a bone scan was usually obtained in patients with a PSA > 10 ng/ml and computed tomography of the abdomen and pelvis for serum PSA levels > 20 ng/ml.

Oligonucleotide Probes.
Specific antisense oligonucleotide DNA probes were designed and synthesized as reported previously (21) . The sequences and working dilution of the probes were as follows: (a) MMP-9, 5'-CCG GTC CAC CTC GCT GGC GCT CCG GA-3' (1:200); (b) MMP-2, 5'-GGC CAC ATC TGG GTT GCG GC-3' (1:200); and (c) E-cadherin, mixture of 5'-TGG AGC GGG CTG GAG TCT GAA CTG-3' (1:200) and 5'-GAC GCC GGC GGC CCC TTC ACA GTC-3' (1:200). Ad (T) 20 oligonucleotide was used to verify the integrity of mRNA in each sample (28 , 29) . The lyophilized probes were reconstituted to a 1 µg/µl stock solution in 10 mM Tris HCl (pH 7.6) and 1 mM EDTA. The stock solution was diluted with Probe Diluent (Research Genetics) immediately before use.

ISH.
ISH was performed as described previously (21) and carried out according to the Microprobe manual staining system (Fisher Scientific, Pittsburgh, PA). Formalin-fixed, paraffin-embedded specimens (4-µm thick) were mounted on silane-coated ProbeOn slides (Fisher Scientific; Ref. 30 ). The slides were placed in the Microprobe slide holder, dewaxed, and dehydrated, followed by enzymatic digestion with pepsin (29) . Hybridization of the probe was carried out for 60 min at 45°C, and the samples were then washed three times with 1 x SSC for 2 min at 45°C. The samples were incubated in alkaline phosphatase-labeled avidin for 30 min at 45°C, briefly rinsed in 50 mM Tris buffer (pH 7.6), rinsed with alkaline phosphatase enhancer (Biomeda Corp., Foster City, CA) for 1 min, and finally incubated with chromogen substrate Fast Red (Research Genetics) for 30 min at 45°C. A positive reaction in this assay stained red. To check the specificity of the hybridization signal, the following controls were used: (a) RNase pretreatment of tissue sections; (b) a biotin-labeled sense probe; and (c) competition assay with unlabeled antisense probe. A markedly decreased or absent signal was obtained under the above conditions.

Image Analysis.
Stained sections were examined in a Zeiss photomicroscope (Carl Zeiss, Inc., Thornwood, NY) equipped with a three chip-charged coupled device color camera (model DXC-960 MD; Sony Corp., Tokyo, Japan). The images were analyzed using the Optimas image analysis software, version 5.2 (Bothell, WA). The slides were prescreened by one of the investigators to determine the range in staining intensity of the slides to be analyzed. Images covering the range of staining intensities were captured electronically; a color bar (montage) was created; and a threshold value was set in the red, green, and blue mode of the color camera. All subsequent images were quantified based on this threshold. The integrated absorbance of the selected fields was determined based on its equivalence to the mean log inverse gray scale value multiplied by the area of the field. The samples were not counterstained, so the absorbance was solely the product of the ISH reaction.

The methodology used to determine gene expression via ISH was similar to that used in a previous study (21) . For RP specimens, measured fields were 1-mm2-wide areas at the center (four areas) and the edge (at least five fields, approximately one field every 2 mm) of a tumor. In BX specimens, 1-mm fields were examined along the whole length of the tissue core (Fig. 1)Citation . For the Japanese cohort, this analysis was performed on all of the BX cores, irregardless of GS. For the M. D. Anderson Cancer Center cohort, the analysis was performed only on the BX core that represented the highest GS. Gene expression was assessed in normal tissues (at least two fields for built-in control) for both RP and BX specimens. Within each field, ≥10 cells (range, 10–20) with adequate cytoplasm, permitting measurement of staining intensity, were analyzed. Areas of nuclear staining and necrotic cells were avoided. Staining of the cells was next quantified to derive an average value of the field. All specimens used in the present study exhibited strong staining for the poly d(T)20 probe, indicating that the mRNA was well preserved. We adjusted specific gene expression by the poly d(T)20 in each slide to compare the expression levels of different slides. This is analogous to a loading control used in mRNA Northern blot analysis. In addition, on each slide, the gene expression levels in tumor epithelium were further normalized to the expression in histologically normal epithelium on the same slide to control for samples that were run on different days.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1. Analysis of core-needle BX specimens. The BX was separated into 1-mm fields containing tumor and normal glands. Within each field, we analyzed ≥10 cells (range, 10–20) and calculated the MMP:E-cadherin ratio of each field. The largest MMP:E-cadherin ratio among the examined fields was that assigned for the BX.

 
The intensity of staining for each gene was evaluated by measuring the integrated absorbance of the poly d(T)20 probe for each specimen, as well as the specific mRNA probe of interest in both histologically normal and tumor epithelium. The absorbance for a specific gene of interest was divided by the poly d(T)20 expression level in the same area to adjust for mRNA integrity. The resulting normalized value obtained in tumor epithelium was divided by value obtained from measuring expression levels in non-neoplastic epithelium. This ratio was multiplied by 100, and the result represented relative corrected gene expression in tumor versus histologically normal epithelium. The MMP:E-cadherin ratio was calculated using the corrected mRNA expression levels using the following formula: (MMP-2 + MMP-9)/2 ÷ E-cadherin expression level (Fig. 2)Citation . The field providing the highest MMP:E-cadherin ratio was the one assigned to either the BX or RP specimen.



View larger version (100K):
[in this window]
[in a new window]
  		Fig. 2. ISH analysis of E-cadherin and type IV collagenase mRNA expression in a BX specimen and the corresponding RP. H&E histology reveals a Gleason 7 adenocarcinoma in both the BX and RP (stage = pT3a,N0). Hybridization with hyperbiotinylated poly d(T)20 probe confirmed mRNA integrity. The numbers for E-cadherin, MMP-9, and MMP-2 indicate expression intensities as compared with the epithelium of normal glands, which was assigned the value of 100. The MMP:E-cadherin ratios [(MMP-2 + MMP-9)/2 ÷ E-cadherin] for the prostatectomy and BX were 8.8 and 6.8, respectively. Bar, 50 µm.

 
Statistics.
One-way ANOVA (with the Tukey honestly significant difference multiple comparison post-hoc tests) was used to compare the mean BX and RP MMP:E-cadherin ratios (31) . The Spearman correlation for the continuous ratio variables among BX and RP specimens was assessed, as well as the categorical ratio variables of <6 versus ≥6 (binary correlation described by the {kappa} statistic). The statistical tests were also compared between the Japanese and M. D. Anderson cohorts. The cutoff value of 6 was based on a previous study in which organ-confined prostate cancer (pT2) exhibited MMP:E-cadherin ratios of <6 compared with those that were advanced (pT3a, pT3b, and pT3a-bN1, ratio = ≥6; Ref. 21 ). {chi}2 tests of association were used to relate categorical data [e.g., PSA, GS, clinical stage, total length of cancer (Japanese cohort only), and BX MMP:E-cadherin ratio] to pathologic stage of RP specimens. Logistic regression was used to determine which variables independently predicted final stage (organ confined versus not organ confined) based on all clinical variables assessed. SPSS software (SPSS, Inc., Chicago, IL) was used for the statistical analysis (31) .

Furthermore, we combined the clinical variables serum PSA (<10 ng/ml, ≥10–20 ng/ml, >20 ng/ml), GS (<7, 7, >7), and clinical stage (T1–T2a, T2b–c, T3) into a single variable and compared the ability of this new variable (Model 1) versus BX MMP:E-cadherin ratio (Model 3) to predict for organ-confined versus advanced cancer using binary logistic regression analysis. The same comparison was also performed using individual patient data for serum PSA, clinical stage, and BX GS to determine the median probability for the prediction of organ-confined disease according to Partin et al. (Ref. 1 ; Model 2). The Minus 2 Log Likelihood was used to calculate the difference between the actual observed outcome and that predicted by the model. The percentage of cases actually classified correctly for each model was compared using {chi}2 analysis.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Clinical Parameters.
The clinical, serological, and pathologic characteristics of the cohort of 44 patients in whom matched RP and BX specimens were obtained are shown in Table 1Citation . Overall, clinical stage was confined (T2) in 39 of 44 (89%) patients as determined on the basis of rectal examination; however, subsequent to RP only, 17 of 44 (39%) patients were pathologically confirmed (i.e., pT2). Serum PSA values among the cohort ranged from 3.2 to 35.2 ng/ml (mean ± SD, median, 15.7 ± 8.8, 14.7). BX GSs were ≤6 in 9 patients, 7 in 19 patients, and ≥8 in 16 patients. In 10 cases (23%), the BX GS was discordant with the RP GS (lower in 9 patients, higher in 1). The total length of cancer in BX cores from the Japanese cohort ranged from 1 to 35 mm (median, 10 mm).


View this table:
[in this window]
[in a new window]
  		Table 1 Clinical characteristics of 44 patients with matched BX and RP specimens

 
ISH Examination of Prostate Specimens.
The BX MMP:E-cadherin ratios were stratified by GS and compared with those from RP specimens (Table 2)Citation . In both BX and RP specimens, the MMP:E-cadherin ratio increased with GS category. There was no difference between the MMP:E-cadherin values of the BX and RP specimens for the GS ≤ 6 category. However, for those of the GS 7 and GS ≥ 8 cancers, the MMP:E-cadherin ratios were significantly higher among RP specimens compared with those from BX (Table 2)Citation . Overall, there was a strong correlation between the BX and RP MMP:E-cadherin ratios measured along a continuous spectrum. (Spearman coefficient = 0.755; P < 0.0001; Fig. 3Citation ). Likewise, the correlation between BX and RP MMP:E-cadherin ratios for the categories of <6 versus ≥6 was also quite strong ({kappa} coefficient = 0.773, P < 0.0001). We also assessed the correlation between BX and RP MMP:E-cadherin ratios for the Japanese and M. D. Anderson cohorts separately. In both patient cohorts, the correlations were strong [0.831 P = 0.002 (M. D. Anderson) and 0.723 P = 0.0001 (Japanese cohort)]. The scattergram in Fig. 3Citation shows that the MMP:E-cadherin ratio between RP and BX agreed with respect to the cutoff value of 6 in all but 5 of 44 (11%) cases (Table 3)Citation . In 3 of the 5 cases (cases 1–3), only 4 BX cores were obtained. In 2 of the cases (cases 1 and 3), only 1–2 mm of cancer were present in the BX material. The RP specimens from case 4 demonstrated a large cancer with remarkable intratumoral heterogeneity for the MMP:E-cadherin ratio. Analysis of the prostatectomy specimen for the case revealed that the ratio varied widely, from 1.3 to 7.2, and positive fields (≥6) were noted in only 4 (17%) of 23 examined. In 2 cases, both BX GS and MMP:E-cadherin ratio were different between the BX and RP specimen. In 1 case (case 5), the BX ratio was reflective of one of the tumor foci present in the RP specimens but not of the tumor focus of the highest GS and pathologic stage. Thus, intratumoral and intertumoral heterogeneity of gene expression, as well as "under sampling" of the prostate, contributed to falsely lower readings than those present in the RP specimen.


View this table:
[in this window]
[in a new window]
  		Table 2 Comparison of ISH-MMP:E-cadherin ratio between core needle BX and prostatectomy specimens

 


View larger version (10K):
[in this window]
[in a new window]
  		Fig. 3. Case-by-case correlation of MMP:E-cadherin ratios among 44 matched RP and prostate BX specimens. Coefficients for continuous (Spearman = 0.755) and binary ({kappa} = 0.773, for cutoff <6 versus ≥6) correlations were strong. In five cases, the BX MMP:E-cadherin ratio was less than in the corresponding prostatectomy specimen.

 

View this table:
[in this window]
[in a new window]
  		Table 3 Clinical parameters in cases where BX MMP:E-cadherin ratio falsely predicted organ-confined cancer at RP

 
Correlation of ISH Findings with Clinical Parameters.
Clinical stage, serum PSA levels, BX GS, total cancer length, and BX MMP:E-cadherin ratio as related to the pathologic stage of the RP specimens are shown in Table 4Citation . Increasing values for all of the measured variables with the exception of total cancer length were significantly associated with advancing pathologic stage. The mean total cancer length in prostate biopsies measured among the Japanese cohort trended toward significance comparing organ-confined versus advanced prostate cancer cases (pT2 = 8 ± 5 mm versus ≥ pT3 = 13 ± 10 mm, P = 0.07); however, a statistically significant cutoff for total BX cancer length could not be determined from the data set. In contrast, the BX MMP:E-cadherin ratio exhibited the strongest association with RP pathologic stage. Among 17 cases of proven organ-confined cancer, the BX MMP:E-cadherin ratio correctly identified all 17 with a cutoff of <6. Thus, the sensitivity for organ-confined cancer was 100%. However, 5 cases exhibited falsely positive results, providing a specificity of 17 of 22 (77%). If the BX ratio was ≥6, all of the RP specimens exhibited advanced cancer. When we compared the ability of the BX MMP:E-cadherin ratio to predict for organ-confined versus advanced cancer among the cohorts from Japan and M. D. Anderson Cancer Center, we found both to be highly significant (Table 4)Citation . Subsequently, forward stepwise logistic regression analysis was performed to determine which of the variables independently predicted organ-confined (pT2) versus advanced cancer (≥pT3). The BX MMP:E-cadherin ratio (cutoff of <6 versus ≥6) correctly classified 89% (P < 0.0001) of the cases when used alone. In the second step of the regression model, the serum PSA level was the only other variable that could be added to enhance the predictive value of the model (91% of cases were correctly classified, P = 0.003).


View this table:
[in this window]
[in a new window]
  		Table 4 Clinical parameters and biopsy MMP:E-cadherin ratio related to pathologic stage

 
Among this patient cohort, clinical stage (≤T2a versus T3), serum PSA level (<10 ng/ml versus >20 ng/ml), and GS (≤6 versus ≥8) were predictive of stage pT2 versus ≥ pT3 cancer (Table 4)Citation . However, the intermediate categories for each variable (clinical stage T2b and c, serum PSA 10–20 ng/ml, and GS 7) accounted for 43–75% of the patients. Among these latter groups, the incidence of pathologic-confined cancer ranged from 32 to 57% and that of advanced cancer from 58 to 68%. When evaluating the ability of the BX MMP:E-cadherin ratio to predict the final pathologic stage among the patients with intermediate clinical parameters (Table 5)Citation , we found that the cutoff value of <6 versus ≥6 correctly identified 79–85% of cases as being either pT2 or ≥pT3.


View this table:
[in this window]
[in a new window]
  		Table 5 Biopsy MMP:E-cadherin ratio in patients of intermediate clinical stage, GS, and serum PSA

 
To further characterize the unique ability of the BX ratio to predict pathologic stage when compared with contemporary strategies, we performed two different comparisons. First, we combined the clinical variables serum PSA (<10 ng/ml, ≥10–20 ng/ml, >20 ng/ml), GS (<7, 7, >7), and clinical stage (T1–T2a, T2b–c, T3) into a single variable and compared the ability of this new variable versus BX MMP:E-cadherin ratio to predict for organ-confined versus advanced cancer using binary logistic regression analysis (Table 6)Citation . Using this new variable, (model 1) ~82% of cases were correctly classified (8 cases incorrectly assigned). Using individual patient data to assign a "Partin" table prediction provided a similar prognostic capability (80% of cases correctly assigned, 9% incorrect; model 2).


View this table:
[in this window]
[in a new window]
  		Table 6 Comparison of strategies using pretherapy variables to predict final pathologic stagea

 
However, the BX MMP:E-cadherin ratio (model 3) provided a significantly enhanced ability to correctly predict pathologic stage (89% correctly classified compared with models 1 and 2, P = 0.0094–0.0029). Combining models 1 and 3 allowed for the assignment of all but 4 cases (91% correctly assigned) and was significantly better than using the BX ratio alone (P = 0.018). Thus, the BX MMP:E-cadherin ratio significantly enhanced the value of standard staging modalities.


   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The MMP:E-cadherin ratio was found previously to be highly associated with pathologic stage in RP specimens, and it provided the potential for individualized prognosis for patients (21) . However, given the intratumoral heterogeneity of gene expression found in some prostate cancers as noted in the previous study (21) , the feasibility and efficacy of performing the assay on pretherapy BXs needed to be determined. Furthermore, it was necessary to assess whether determining specific gene expression in prostate BX specimens provided independent prognostic information compared with standard staging modalities.

The present data reveal that an increasing MMP:E-cadherin ratio in pretherapy prostate BXs correlates with increasing GS, RP pathologic stage, and the MMP:E-cadherin ratio in RP specimens. Furthermore, the BX MMP:E-cadherin ratio was the best single test for determining whether prostate cancer, which was thought to be clinically localized previous to RP surgery, was truly confined to the prostate. These findings support the concept that measuring the molecular events associated with invasion and metastasis in prostate cancer BXs should provide clinically relevant information. Tumor GS and serum PSA level provide surrogate assessments of tumor biology, whereas assays that directly evaluate the expression of MMPs and E-cadherin in tumors may be more sensitive than currently available morphological, serological, or anatomical parameters.

Recent data from two other studies support the present data. Still et al. (32) and De Marzo et al. (33) described increased expression of MMP-2 in high-grade, advanced stage prostate cancer and decreased E-cadherin expression among the same types of prostate cancer specimens. Still et al. (32) also evaluated MMP-2 expression in prostate tissue as a function of its inhibitor. MMP-2 levels were expressed as a ratio of tissue inhibitor of MMP expression, and this ratio was enhanced 2.8–3.3-fold in high-grade, advanced stage cases compared with normal or low-grade, low-stage prostate cancer (32) . Our data agree that the "invasive profile" of tumors is highly associated with the MMP:E-cadherin ratio.

We have reported previously that an MMP:E-cadherin cutoff ratio of <6 among RP specimens defined all organ-confined cancers, whereas those specimens exhibiting a ratio ≥6 exhibited advanced cancer (pT3a–b and/or lymph node metastasis). This was true in 39–44 (89%) of the cases in the current study. Several factors related to tumor sampling could have contributed to the 5 cases in which the BX ratio was markedly lower than the RP ratio. These factors include intertumoral heterogeneity of gene expression, which is the presence of multiple tumors of different biological potential (Table 3Citation , case 5), and intratumoral heterogeneity of gene expression, which is the presence of heterogeneous areas within the same tumor (Table 3Citation , case 1) of gene expression. The exact number of BX cores that are required to adequately sample the prostate for both diagnostic and prognostic purposes remains controversial; however, among the 5 patients with false negative findings, 3 had only 4 BX cores taken. Recent findings strongly suggest that detecting of prostate cancer and defining its extent are enhanced by taking >6 BX cores (34, 35, 36, 37) . Thus, it is conceivable that false negative findings may be minimized in future studies via enhanced tumor sampling. Of interest, although only the BX with the highest GS was used to perform the ISH analysis in the M. D. Anderson cohort, neither the correlation with the RP MMP:E-cadherin ratio or the ability to predict pathologic stage appeared to be adversely affected. This is likely related to the fact that we preselected the BX with the highest GS, which often reflected the RP dominant tumor. Thus, the specimen providing the highest MMP:E-cadherin ratio was effectively sampled using this strategy. Whether this abbreviated BX analysis technique will reproducibly reflect final pathologic stage requires prospective analysis.

Although the mRNA ISH technique we describe is labor intensive compared with that of traditional staging modalities, it provides unique information in some cases. Among patients exhibiting features that place them at significantly high risk (clinical stage T3, GS ≥ 8, serum PSA > 20 ng/ml) or low risk (clinical stage ≤ T2a, GS < 7, and serum PSA < 10 ng/ml) for extraprostatic extension of cancer and biochemical failure, performing additional molecular studies may not be necessary (Table 4Citation ; Refs. 1, 2, 3, 4, 5 ). However, among patients with intermediate features consisting of GS 7 cancer, serum PSA levels of 10–20 ng/ml, and clinical stage T2b–c, only 32–42% of cancers were pathologically confined. At a cutoff of <6 versus ≥6, the MMP:E-cadherin ratio correctly classified 79–85% of the cases within these intermediate categories. This finding is clinically relevant, as 30–40% of patients in contemporary series may exhibit such "intermediate prognostic features" (5) . Considering the latter scenario, combining GS, clinical stage, and serum PSA together as predictors of pathologic stage has been shown to be superior to using any of the same variables alone. Prediction tables published by Partin et al. (1 , 38) are a clinically validated, widely used resource for counseling patients regarding the probability of finding organ-confined cancer at RP using pretherapy variables. The data in the present study reveal that analyzing the BX MMP:E-cadherin ratio significantly enhanced our ability to predict pathologic stage over that of the "Partin" table prediction or by combining our own clinical variables into a similar prediction model. The implication is that such a molecular test may move clinicians closer to the goal of individualized patient prognosis (89% of patient specimens correctly classified; Table 6Citation ). Indeed, combining the BX ratio with clinical variables allowed the prediction of 91% of cases.

Several caveats exist in interpreting the data in the present study. The patients in the present series exhibited a higher incidence of advanced stage, higher GSs, and higher serum PSA levels than those in contemporary RP series (reviewed in Ref. 39 ). Although the patient cohort represents patients treated within the "PSA era," 70% of the patients were from the Japanese cohort where all of the patients were diagnosed after presenting with symptoms related to the prostate compared with ~50% of patients today who are diagnosed solely on the basis of an elevated PSA (40) . Whether the assay provides relevant data in the latter setting is unknown. In addition, mRNA ISH is time consuming, because it requires expert pathologic correlation in identifying specific areas within a prostate BX to be analyzed. Therefore, we believe that the current data are very promising and show proof of principle for the concept of molecular staging of prostate cancer. However, a large prospective analysis determining MMP:E-cadherin expression ratios among patients undergoing contemporary BX strategies driven by serum PSA elevation is warranted to validate the feasibility and effectiveness of this promising assay.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by NIH Specialized Programs of Research Excellence Grant CA90270-01. Back

2 To whom requests for reprints should be addressed, at The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Department of Urology, Box 446, Houston, TX 77030. Phone: (713) 792-3250; Fax: (713) 794-4824; E-mail: cpettawa{at}mdanderson.org Back

3 The abbreviations used are: PSA, prostate-specific antigen; BX, biopsy; RP, radical prostatectomy; ISH, in situ hybridization; MMP, matrix metalloproteinase; GS, Gleason score. Back

Received 9/12/02; revised 1/30/03; accepted 2/19/03.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Partin A. W., Kattan M. W., Subong E. N. P., Walsh P. C., Wojno K. J., Oesterling J. E., Scardino P. T., Pearson J. D. The use of prostate-specific antigen, clinical stage, and GS to predict pathological stage in men with localized prostate cancer: a multi-institutional update. JAMA, 277: 1445-1451, 1997.[Abstract]
  2. Kattan M. W., Eastham J. A., Stapleton A. M. F., Wheeler T. M., Scardino P. T. A preoperative monogram for disease recurrence following radical prostatectomy for prostate cancer. J. Natl. Cancer Inst. (Bethesda), 90: 766-771, 1998.[Abstract/Free Full Text]
  3. D’Amico A. V., Whittington R., Malkowicz S. B., Fondurulia J., Chen M., Tomaszewski J. E., Wein A. The combination of preoperative prostate-specific antigen and postoperative pathological findings to predict prostate specific antigen outcome in clinically localized prostate cancer. J. Urol., 160: 2096-2101, 1998.[CrossRef][Medline]
  4. Kattan M. W., Zelefsky M. J., Kupelian P. A., Scardino P. T., Fuks Z., Leibel S. A. Pretreatment nomogram for predicting the outcome of three-dimensional conformal radiotherapy in prostate cancer. J. Clin. Oncol., 18: 3252-3259, 2000.
  5. D’Amico A. V. Combined-modality staging for localized adenocarcinoma of the prostate. Oncology, 15: 1049-1058, 2001.[Medline]
  6. Fidler I. J. Critical factors in the biology of human cancer metastasis: twenty-eighth G. H. A. Clowes Memorial Award lecture. Cancer Res., 50: 6130-6138, 1990.[Abstract/Free Full Text]
  7. Pettaway C. A., Pathak S., Greene G., Ramirez E., Wilson M. R., Killion J. J., Fidler I. J. Selection of highly metastatic variants of different human prostatic carcinomas utilizing orthotopic implantation in nude mice. Clin. Cancer Res., 2: 1627-1636, 1996.[Abstract]
  8. Greene G. F., Kitadai Y., Pettaway C. A., von Eschenbach A. C., Fidler I. J. In situ mRNA hybridization technique for analysis of metastasis-related genes in human prostate carcinoma cells. Am. J. Pathol., 150: 1571-1582, 1997.[Abstract]
  9. Stearns M., Stearns M. E. Evidence for increased activated metalloproteinase-2 (MMP-2): expression associated with human prostate cancer progression. Oncol. Res., 8: 69-75, 1996.[Medline]
  10. Nakamoto T., Chang C., Li A., Chodak G. W. Basic fibroblast growth factor in human prostate cancer cells. Cancer Res., 52: 571-577, 1992.[Abstract/Free Full Text]
  11. Ferrer F. A., Miller L. J., Andrawis R. I., Kurtzman S. H., Albertsen P. C., Laudone V. P., Kreutzer D. L. Human prostate cancer: in situ and in vitro expression of VEGF by human prostate cancer cells. J. Urol., 157: 2329-2333, 1997.[CrossRef][Medline]
  12. Balbay M. D., Pettaway C. A., Kuniyasu H., Inoue K., Ramirez E., Li E., Fidler I. J., Dinney C. P. N. Highly metastatic human prostate cancer growing within the prostate of athymic mice overexpresses vascular endothelial growth factor. Clin. Cancer Res., 5: 783-789, 1999.[Abstract/Free Full Text]
  13. Borgstrom P., Bourdon M. A., Hillan K. J., Sriramarao P., Ferrara N. Neutralizing antivascular endothelial growth factor antibody completely inhibits angiogenesis and growth of human prostate carcinoma micro tumors in vivo. Prostate, 35: 1-10, 1998.[CrossRef][Medline]
  14. Inoue K., Slaton J. W., Eve B. Y., Kim S. J., Perrotte P., Balbay M. D., Yano S., Bar-Eli M., Radinsky R., Pettaway C. A., Dinney C. P. N. Interleukin 8 expression regulates tumorigenicity and metastases in androgen-independent prostate cancer. Clin. Cancer Res., 6: 2104-2119, 2000.[Abstract/Free Full Text]
  15. Umbas R., Schalken J. A., Alders T. W., Carter B. S., Karthas H. F. M., Schaafsma H. E., Debruyne F. M. J., Isaac W. B. Expression of the cellular adhesion molecule E-cadherin is reduced or absent in high-grade prostate cancer. Cancer Res., 52: 5104-5109, 1992.[Abstract/Free Full Text]
  16. Umbas R., Isaacs W. B., Bringuier P. P., Schraafsma H. E., Karthaus H. F. M., Oosterhof G. O. N., Debruyne F. M. J., Schalken J. A. Decreased E-cadherin expression is associated with poor prognosis in patients with prostate cancer. Cancer Res., 54: 3929-3933, 1994.[Abstract/Free Full Text]
  17. Richmond P. J. M., Karayiannakis A. J., Nagafuchi A., Kaisary A. V., Pignatelli M. Aberrant E-cadherin and E-catenin expression in prostate cancer: correlation with patient survival. Cancer Res., 57: 3189-3193, 1997.[Abstract/Free Full Text]
  18. Pettaway C. A. Prognostic markers in clinically localized prostate cancer. Tech. Urol., 4: 35-42, 1998.[Medline]
  19. Borre M., Hoyer M., Nerstrom B., Overgaard J. DNA ploidy and survival of patients with clinically localized prostate cancer treated without intent to cure. Prostate, 36: 244-249, 1998.[CrossRef][Medline]
  20. Veltri R. W., Miller M. C., Partin A. W., Coffey D. S., Epstein J. I. Ability to predict biochemical progression using GS and a computer-generated quantitative nuclear grade derived from cancer cell nuclei. Urology, 48: 685-691, 1996.[CrossRef][Medline]
  21. Kuniyasu H., Troncoso P., Johnston D., Bucana C. D., Tahara E., Fidler I. J., Pettaway C. A. Relative expression of type IV collagenase, E-cadherin, and vascular endothelial growth factor/vascular permeability factor in prostatectomy specimens distinguishes organ-confined from pathologically advanced prostate cancers. Clin. Cancer Res., 6: 2295-2308, 2000.[Abstract/Free Full Text]
  22. Kuniyasu H., Ellis L. M., Evans D. B., Abbruzzese J. L., Fenoglio C. J., Bucana C. D., Cleary K. R., Tahara E., Fidler I. J. Relative expression of E-cadherin and type IV collagenase genes predicts disease outcome in patients with resectable pancreatic carcinoma. Clin. Cancer Res., 5: 25-33, 1999.[Abstract/Free Full Text]
  23. Kitadai Y., Ellis L. M., Tucker S. L., Green G. F., Bucana C. D., Cleary K. R., Takahashi Y., Tahara E., Fidler I. J. Multiparametric in situ mRNA hybridization analysis to predict disease recurrence in patients with colon carcinoma. Am. J. Pathol., 149: 1541-1551, 1996.[Abstract]
  24. Anzai H., Kitadai Y., Bucana C. D., Sanchez R., Omoto R., Fidler I. J. Intratumoral heterogeneity and inverse correlation between expression of E-cadherin and collagenase type IV in human gastric carcinomas. Differentiation, 60: 119-127, 1996.[CrossRef][Medline]
  25. Herbst R. S., Yano S., Kuniyasu H., Khuri F. R., Bucana C. D., Guo F., Liu D., Kemp B., Lee J. J., Hong W. K., Fidler I. J. Differential expression of E-cadherin and type IV collagenase genes predicts outcome in patients with stage I non-small cell lung carcinoma. Clin. Cancer Res., 6: 790-797, 2000.[Abstract/Free Full Text]
  26. Fleming I. D. Cooper J. S. Henson D. E. Hutter R. V. P. Kennedy B. J. Murphy G. P. O’Sullivan B. Sobin L. H. Yarbro J. W. eds. . AJCC Cancer Staging Manual, Ed. 5 121-126, Lippincott-Raven Philadelphia 1997.
  27. Schroder F. H., Hermanek P., Denis L., Fair W. R., Gospodarowicz M. K., Pavone-Macaluso M. The TNM classification of prostate cancer. Prostate Suppl., 4: 129-138, 1992.[Medline]
  28. Bucana C. D., Radinsky R., Dong Z., Sanchez R., Brigati D. J., Fidler I. J. A rapid colorimetric in situ mRNA hybridization technique using hyperbiotinylated oligonucleotide probes for analysis of mdr-1 in mouse colon carcinoma cells. J. Histochem. Cytochem., 41: 499-506, 1993.[Abstract]
  29. Park C-S., Manahan L. J., Brigati D. J. Automated molecular pathology: one hour in situ DNA hybridization. J. Histotechnol., 14: 219-229, 1991.
  30. Reed J. A., Manahan L. J., Park C. S., Brigati D. J. Complete one-hour immunocytochemistry based on capillary action. Biotechniques, 13: 434-443, 1992.[Medline]
  31. Zar J. H. eds. . Biostatistics Analysis, Ed 4 Prentice-Hall Upper Saddle River, NJ 1999.
  32. Still K., Robson C. N., Autzen, Robinson M. C., Hamdy F. C. Localization and quantification of mRNA for matrix metalloproteinase-2 (MMP-2) and tissue inhibitor of matrix metalloproteinase-2 (TIMP-2) in benign and malignant prostatic tissue. Prostate, 42: 8-25, 2000.[CrossRef][Medline]
  33. DeMarzo A., Knudsen B., Chan-Tack K., Epstein J. I. E-cadherin expression as a marker of tumor aggressiveness in routinely processed radical prostatectomy specimens. Urology, 53: 707-713, 1999.[CrossRef][Medline]
  34. Epstein J. I., Walsh P. C., Sauvageot J., Carter H. B. Use of repeat sextant and transition zone biopsies for assessing extent of prostate cancer. J. Urol., 158: 1886-1890, 1997.[CrossRef][Medline]
  35. Cookson M. S. Update on transrectal ultrasound-guided needle biopsy of the prostate. Mol. Urol., 4: 93-98, 2000.[Medline]
  36. Djavan B., Zlotta A. R., Ekane S., Remzi M., Kramer G., Roumeguere T., Etemad M., Wolfram R., Schulman C. C., Marberger M. Is one set of sextant biopsies enough to rule out prostate cancer? Influence of transition and total prostate volumes on prostate cancer yield. Eur. Urol., 38: 218-224, 2000.[CrossRef][Medline]
  37. Babaian R. J., Toi A., Kamoi K., Troncoso P., Sweet J., Evans R., Johnston D., Chen M. A. comparative analysis of sextant and an extended 11-core multisite directed biopsy strategy. J. Urol., 163: 152-157, 2000.[CrossRef][Medline]
  38. Blute M. L., Bergstralh E. J., Partin A. W., Walsh P. C., Kattan M., Scardino P. T., Montie J. E., Pearson J. D., Slezak J. M., Zincke H. Validation of the Partin tables for predicting pathologic stage of clinically confined prostate cancer. J. Urol., 164: 1591-1595, 2000.[CrossRef][Medline]
  39. Han M., Partin A. W., Pound C. R., Epstein J. I., Walsh P. C. Long-term biochemical disease-free survival following anatomic radical retropubic prostatectomy. Urol. Clin. North Am., 28: 555-565, 2002.
  40. Catalona W. J., Richie J. P., Ahmann F. R., Hudson M. A., Scardino P. T., Flanigan R. C., deKernion J. B., Ratliff T. L., Kavoussi L., Dalkin B. L., Waters W. B., MacFarlane M. T., Southwick P. C. Comparison of digital rectal examination and serum prostate specific antigen in the early detection of prostate cancer: results of a multicenter clinical trial of 6,630 men. J. Urol., 151: 1283-1290, 1994.[Medline]



This article has been cited by other articles:


Home page
 Cancer Res.Home page
T. Hara, H. Miyazaki, A. Lee, C. P. Tran, and R. E. Reiter
Androgen Receptor and Invasion in Prostate Cancer
Cancer Res., February 15, 2008; 68(4): 1128 - 1135.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
M. Myoishi, H. Hao, T. Minamino, K. Watanabe, K. Nishihira, K. Hatakeyama, Y. Asada, K.-i. Okada, H. Ishibashi-Ueda, G. Gabbiani, et al.
Increased Endoplasmic Reticulum Stress in Atherosclerotic Plaques Associated With Acute Coronary Syndrome
Circulation, September 11, 2007; 116(11): 1226 - 1233.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
E. Efstathiou, P. Troncoso, S. Wen, K.-A. Do, C. A. Pettaway, L. L. Pisters, T. J. McDonnell, and C. J. Logothetis
Initial Modulation of the Tumor Microenvironment Accounts for Thalidomide Activity in Prostate Cancer
Clin. Cancer Res., February 15, 2007; 13(4): 1224 - 1231.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
X. Zi, Y. Guo, A. R. Simoneau, C. Hope, J. Xie, R. F. Holcombe, and B. H. Hoang
Expression of Frzb/Secreted Frizzled-Related Protein 3, a Secreted Wnt Antagonist, in Human Androgen-Independent Prostate Cancer PC-3 Cells Suppresses Tumor Growth and Cellular Invasiveness
Cancer Res., November 1, 2005; 65(21): 9762 - 9770.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
V. J. Assikis, K.-A. Do, S. Wen, X. Wang, J. H. Cho-Vega, S. Brisbay, R. Lopez, C. J. Logothetis, P. Troncoso, C. N. Papandreou, et al.
Clinical and Biomarker Correlates of Androgen-Independent, Locally Aggressive Prostate Cancer with Limited Metastatic Potential
Clin. Cancer Res., October 15, 2004; 10(20): 6770 - 6778.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kuniyasu, H.
Right arrow Articles by Pettaway, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kuniyasu, H.
Right arrow Articles by Pettaway, C. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online