Annals of Oncology Advance Access originally published online on September 13, 2006
Annals of Oncology 2006 17(12):1797-1802; doi:10.1093/annonc/mdl310
© 2006 European Society for Medical Oncology
gynecologic tumors |
Automated quantitative analysis of DCC tumor suppressor protein in ovarian cancer tissue microarray shows association with ß-catenin levels and outcome in patients with epithelial ovarian cancer
1 Department of Clinical Therapeutics, University of Athens, School of Medicine, Athens, Greece; Departments of
2 Medical Oncology and
3 Pathology, Yale University School of Medicine, New Haven, Connecticut 06520, USA
4 Histopathology Department, Alexandra Hospital, Athens, Greece
* Correspondence to: Dr A. Bamias, 31 Komninon Street, Haidari, 124 62 Athens, Greece. Tel: 030 210 3381546; Fax: 030 210 3381511; E-mail: abamias{at}med.uoa.gr
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: The deleted in colorectal cancer (DCC) protein, the product of DCC tumor suppressor gene, is frequently altered in cancer. Preclinical data demonstrate that DCC regulates ß-catenin levels. Here, we sought to determine the association of DCC with ß-catenin protein levels, clinicopathological parameters and patient outcome in ovarian cancer using a method of in situ compartmentalized protein analysis.
Methods: A tissue array composed of 150 advanced-stage ovarian cancers, treated with surgical debulking and platinum–paclitaxel (Taxol) combination chemotherapy, was constructed. For evaluation of protein expression, we used an immunofluorescence-based method of automated in situ quantitative measurement of protein analysis (AQUA).
Results: One hundred and twelve patients (74%) had sufficient tissue for AQUA. The median follow-up time for the entire cohort was 33 months. Patients with low nuclear DCC expression had a 3-year progression-free survival (PFS) rate of 0% compared with 33% of those with high DCC expression (P = 0.0067). In multivariate analysis, low nuclear DCC expression level retained its prognostic significance for PFS. Between DCC and ß-catenin, a significant relationship was found, where tumors with low DCC had low ß-catenin and vice versa (P = 0.003).
Conclusions: Low nuclear DCC levels predict for poor patient outcome in epithelial ovarian cancer. DCC may exert its antitumor function, in part, through regulation of ß-catenin levels.
Key words: ß-catenin, DCC, ovarian cancer, prognosis
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Ovarian cancer is ranked fifth in incidence of cancers among women in the United States [1]. When the disease is confined to the ovary, the 5-year survival rate is 95%, however, fewer than one-third of the cases are detected at this stage. The current management of patients with advanced disease [International Federation of Gynecology and Obstetrics (FIGO) stages III and IV] involves surgical debulking followed by chemotherapy. The current standard chemotherapeutic approach for ovarian cancer patients includes platinum-based (plus or minus taxanes) chemotherapy. While the majority of patients respond initially, 60%–80% of them still die of the disease. Within these advanced stages, traditional clinicopathological factors do not accurately predict patients' prognosis. Therefore, considerable interest lies in identifying molecular prognostic indicators in order to predict prognosis and guide treatment decisions.
Deleted in colorectal cancer (DCC) gene is a putative tumor suppressor gene located on chromosome 18q [2]. The prognostic significance of 18q loss of heterozygosity (LOH) and loss of expression of DCC mRNA have been studied in solid tumors [3–7], predominantly in colorectal cancer, where LOH or low protein expression seems to be an adverse prognostic factor [8–11]. Data in ovarian cancer are limited. DCC loss has been reported in only 10% of adenomas and 6% of borderline tumors compared with 60% in carcinomas, suggesting an association of DCC inactivation with carcinogenesis in the ovary [12]. Genetic alterations and altered mRNA and DCC protein expression have been reported in 30%–60% of the cases [13–17], while an association with serous subtype has been suggested [16, 17]. The function of DCC has not been clarified. DCC protein has significant homology to neural cell adhesion molecules [2]. In addition, DCC protein may play a role in cell–cell or cell–substrate interactions through a functional link between DCC and N-cadherin/catenin-dependent cell adhesion complex. This link has been supported by in vitro data showing that overexpression of a truncated DCC protein resulted in decreased ß-catenin expression in neuroblastoma cells [18].
Tissue microarray is a useful tool for molecular marker analysis. Positive features of this technology include studying specimens from hundreds of patients simultaneously as well as uniform handling of all specimens. In addition, a fully quantitative method of analysis for tissue microarrays that allows calculation of expression ratios has been developed [19]. This novel technology uses molecular methods to define subcellular compartments. It then quantifies the amount of protein expressed within the compartment by co-localization. Therefore, this technology permits preservation of tissue morphology while quantifying protein expression in paraffin-embedded tissue.
Here, we sought to determine the association of DCC protein levels with survival in a cohort ovarian cancer tissue microarray using a novel in situ quantitative method of protein expression and correlated those with ß-catenin levels, clinical and pathological data.
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
patient population
Included patients should have primary epithelial ovarian cancer patients, FIGO stages III and IV. They all underwent surgical debulking in the Department of Gynecology of Alexandra University Hospital in Athens from 1996 to 2003 and treated postoperatively with paclitaxel (Taxol BMS, UK)–carboplatin chemotherapy. In all cases, an effort was made for optimal surgical cytoreduction and adequate staging, which included at least: total abdominal hysterectomy with bilateral salpingo oophorectomy, inspection and palpation of all peritoneal surfaces and retroperitoneal area, biopsies of suspect lesions for metastases, infracolic omentectomy and peritoneal washings. Gynecological examination, CA-125 assay and radiological investigations were carried out before chemotherapy initiation. During chemotherapy, CA-125 was carried out monthly, while clinical assessment of response in patients with bidimensional disease was recorded at the middle and at the end of treatment according to World Health Organization criteria [20]. Follow-up examinations were carried out every 3 months.
tissue microarray construction
A tissue microarray consisting of tumors from each patient in the cohort was constructed at the Yale University Tissue Microarray Facility. Following institutional review board approval, the tissue microarray was constructed as previously described [21], including 150 cases. Briefly, tissue cores 0.6 mm in size were obtained from paraffin-embedded formalin-fixed tissue blocks from the Alexandra University Hospital, Department of Pathology archives. Representative areas of invasive tumor to be cored were chosen by a pathologist (SM). The cores were placed on the recipient microarray block using a tissue microarrayer (Beecher Instrument, Silver Spring, MD). All tumors were represented with two-fold redundancy. The tissue microarray was then cut to yield 5-µm sections and placed on glass slides using an adhesive tape transfer system (Instrumedics, Inc., Hackensack, NJ) with UV cross-linking.
quantitative immunohistochemistry
Slides were deparafinized, rehydrated and antigen retrieval was accomplished by application of proteinase K for 30 min. Endogenous peroxidase activity was blocked by incubating in 0.3% hydrogen peroxide in methanol for 30 min. Nonspecific antibody (Ab) binding was then blocked with 0.3% bovine serum albumin (BSA) for 30 min at room temperature. Slides were then incubated with primary Ab to DCC (Mouse Monoclonal Ab, Clone G97-449; BD Biosciences Pharmingen, CA, USA) at 4°C overnight, at 1 : 500 dilution in 0.3% BSA/Tris buffered saline (TBS). Primary Ab to ß-catenin (clone 14; BD Transduction Laboratories, CA, USA) was used at 1 : 450 dilution in 0.3% BSA/TBS. Subsequently, slides were incubated with goat anti-mouse secondary Ab conjugated to a horseradish peroxidase-decorated dextran polymer backbone (Envision; DAKO Corp., Carpinteria, CA) for 1 h at room temperature. Tumor cells were identified by the use of anticytokeratin Ab cocktail (rabbit anti-pancytokeratin Ab z0622; DAKO Corp.) with subsequent goat anti-rabbit Ab conjugated to Alexa546 fluourophore (A11035; Molecular Probes, Eugene, OR). We added 4',6-diamidino-2-phenylindole (DAPI) to visualize nuclei. Target (DCC and ß-catenin) molecules were visualized with a fluorescent chromogen (Cy-5-tyramide; Perkin Elmer Corp., Wellesley, MA). Cy-5 (red) was used because its emission peak is well outside the green–orange spectrum of tissue autofluorescence.
automated image acquisition and analysis
Automated image acquisition and analysis using automated in situ quantitative measurement of protein analysis (AQUA) has been described previously [21]. In brief, monochromatic, high-resolution (1024 x 1024 pixel; 0.5-µm) images were obtained of each histo spot. We distinguished areas of tumor from stromal elements by creating a mask from the cytokeratin signal. DAPI signal was used to identify nuclei, and the cytokeratin signal was used to define cytoplasm. Overlapping pixels [to a 99% confidence interval (CI)] were excluded from both compartments. The DCC and ß-catenin signals were scored on a normalized scale of 1–255 expressed as pixel intensity divided by the target area. AQUA scores for each subcellular compartment (nuclear and cytoplasmic signal) were recorded. AQUA scores for duplicate tissue cores were averaged to obtain a mean AQUA score for each tumor.
statistical analysis
Histo spots containing <10% tumor as assessed by mask area (automated) were excluded from further analysis. AQUA scores represent expression of a target protein on a continuous scale from 1 to 255. It is often useful to categorize continuous variable in order to stratify patients into high versus low categories. Several methods exist to determine a cut point, including biological determination, splitting at the median and determination of the cut point which maximizes effect difference between groups. If the latter method (the so-called ‘optimal P-value’ approach) is used, a dramatic inflation of type I error rates can result [22]. A recently developed program, X-Tile, allows determination of an optimal cut point while correcting for the use of minimum P-value statistics [22]. As the AQUA technology is new, there are no established cut points available for quantitative protein expression. Therefore, for categorization of ß-catenin and DCC expression levels, the X-Tile program was used to generate an optimal cut point. This approach has been successfully applied to AQUA data analysis. Two methods of statistical correction for the use of minimal P-value approach were utilized. First, the X-Tile program output includes calculation of a Monte Carlo P value for the optimal cut point generated. Cut points that yield Monte Carlo P values <0.05 are considered robust and unlikely to represent type I error. Secondly, the Miller–Siegmund minimal P-value correction referenced by Altman et al. [23] was utilized. This approach is accepted in the statistical literature, but relatively unknown in the medical/biological research community. Briefly, when making multiple comparisons to find the minimum P value using the log-rank test, the false-positive rate (i.e. the percentage of times a marker that has no true prognostic value will be found to have a P < 0.05) can approach 40%.
Altman's statistical adjustment generates a minimum P value corrected to yield a true false-positive rate of 5%. The corrected P value (Pcor) is calculated as follows:
 |
Our calculations were carried out using an epsilon of 0.10. Disease-free survival (DFS) and overall survival (OS) were subsequently assessed by Kaplan–Meier analysis with log-rank for determining statistical significance, and only P corrected was reported. This approach has been successfully applied to AQUA data analysis. All survival analysis was carried out at 3-year cut-offs. CIs were assessed by univariate and multivariate Cox proportional hazards model.
OS was defined as time from first day of chemotherapy to death from any cause. DFS was defined as time from first day of chemotherapy to the first of either death from any cause or disease progression (assessed by CA 125 increase and/or imaging studies). ECOG performance status (PS) was dichotomized into ‘0’ versus all others, and histologic type into serous versus all others. Patient population was divided into two groups according to the extent of residual disease at first surgery; 
2 cm and >2 cm. Comparisons of DCC expression with FIGO stage and grade was made by Mantel–Haenszel chi-square test. The association between DCC and ß-catenin expression (as determined by AQUA) was analyzed by Spearman's correlation. Comparisons of DCC expression with PS, histology, clinical response and residual disease were made by Fisher's exact test. Comparison of DCC expression status with age was made using Pearson correlation. All calculations and analyses were carried out with SPSS 12.0 for Windows (SPSS Inc., Chicago, IL).
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
patients
One hundred and fifty patients were included in the study. One hundred and twelve patients (74%) had sufficient tissue for AQUA. Tissues deemed insufficient had <10% tumor mask within the histo spot, as represented on the tissue microarrays. Median follow-up time for the entire cohort was 33 months (range 1–91.7). There were 85 (75%) stage III and 27 (25%) stage IV patients. Eighty-one (72%) patients had tumors of serous histology. Initial histological grade was 13 well-differentiated (11%), 35 moderately differentiated (31%) and 63 poorly differentiated (52%) and one not recorded. Residual disease after surgical debulking was distributed as follows: 25 (22%) with <2 cm and 87 (78%) with >2 cm. For clinical response to initial therapy, complete response (CR) or partial response (PR) was recorded in 66 (59%) patients and stable disease/no response in 46 (41%) patients. Demographic and clinicopathological variables for the cohort analyzed by AQUA are summarized in Table 1.
|
generation of optimal cut point by X-Tile analysis
As visualized by fluorescent immunohistochemistry (IHC), DCC displayed predominantly cytoplasmic and nuclear staining (Figure 1). Normalized AQUA scores were represented on a 1–255 scale. DCC expression followed a skewed distribution as expected for a cancer tissue biomarker (Figure 2).
|
|
Using the X-Tile program, an optimal cut point for DCC nuclear expression was determined at 9.96 AQUA units, with a Monte Carlo P value of 0.04 as determined by X-Tile. Monte Carlo P values <0.05 indicate robust and valid cut-point selection; patients with DCC nuclear expression
9.96 were classified as low expressers (n = 14) and patients with nuclear DCC expression >9.96 were classified as high expressers (n = 98). There was no association between nuclear DCC expression and any of the clinicopathological variables.
The association between DCC and ß-catenin expression (as determined by AQUA) was analyzed by Spearman's correlation. A significant relationship was found, where tumors with high DCC had high ß-catenin and vice versa (Spearman's 
= 0.278; P = 0.003).
univariate survival analysis
progression-free survival.
The expression status of nuclear DCC expression was evaluated for association with progression-free survival (PFS) using Kaplan–Meier survival analysis with log-rank statistic for determining significance. This analysis demonstrates that low nuclear DCC expression is associated with inferior 3-year PFS. Patients with high nuclear DCC expression had a PFS of 33% compared with 0% for patients with low DCC tumors (P = 0.0067) (Figure 3).
|
overall survival.
The expression status of DCC was also evaluated for association with OS. Kaplan–Meier analysis demonstrated that there was no significant 3-year OS rate between patients with low and high nuclear DCC expressing tumors (P = 0.63) (Figure 3). Results of the univariate survival analysis are summarized in Table 2.
|
No significant association of ß-catenin expression with PFS or OS was observed (Figure 4).
|
multivariable analysis
Using the Cox proportional hazards model, we carried out multivariable analysis to assess the predictive value of nuclear DCC expression by AQUA for PFS. We also included the following known prognostic variables in the regression model: FIGO stage, PS, response to chemotherapy, initial histology, residual disease and tumor grade. DCC level (95% CI 1.12–4.17; P = 0.02) along with FIGO stage (95% CI 0.63–0.87; P = 0.02) and residual disease (95% CI 1.34–1.54; P = 0.0001) were significant predictor variables of PFS. Results of multivariable analysis are summarized in Table 3.
|
|
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
In the present study, using a method of in situ proteomic analysis on a tissue microarray, we demonstrate that decreased levels of DCC tumor suppressor protein are associated with shorter 3-year PFS in a cohort of chemotherapy-treated advanced stage ovarian cancer patients. Strikingly, all patients with low DCC protein expression relapsed within 3 years. In addition, we show that there is a direct correlation between DCC and ß-catenin levels consistent with preclinical observations that DCC loss is implicated in the aberrant cellular migration through down-regulation of N-cadherin and ß-catenin levels.
DCC loss is a very common genetic alteration in colorectal carcinomas. In ovarian cancer, DCC mRNA down-regulation and/or protein loss by IHC have been reported in 40% of ovarian cancers. Down-regulation of DCC mRNA was reported in 50% of serous cancers, while it was rare in other histological types. Saegusa et al. [15] analyzed immunohistochemically 124 ovarian carcinomas, 55 cystadenomas and 41 low malignant potential (LMP) tumors for DCC protein expression and compared the results with p53 protein expression, clinicopathological factors and outcome. The authors also analyzed 26 malignant, five LMP, eight benign and seven normal ovarian samples for DCC mRNA levels using RT-PCR and Southern blot hybridization. DCC mRNA and protein levels were significantly decreased in carcinomas compared with benign and LMP lesions. DCC loss was significantly associated with serous subtype, high-grade and advanced FIGO stage, whereas there was no correlation with survival or p53 protein expression.
In the present cohort, no association between DCC protein expression levels and clinical–pathological parameters was found. However, low DCC protein expression had a significant adverse impact on prognosis in this cohort of chemotherapy-treated patients with advanced ovarian cancer. Despite the high response rates of ovarian cancer patients to platinum–paclitaxel combination chemotherapy, the majority of them will relapse and die of disease. Molecular prognostic markers are needed to identify patients in need of more aggressive treatment and spare those with favorable profile, the sequelae of aggressive treatment. DCC expression status, if validated in a prospective study, may be used for patient selection for aggressive treatment protocols. DCC expression was not associated with OS, indicating that the impact on prognosis is mainly related to first-line chemotherapy. It should be mentioned that OS is determined by various factors (residual disease, PS, tumor grade, etc.), while the median follow-up was relatively short. In addition, response to subsequent lines of treatment at relapse may affect prognosis. This factor could not be studied in our cohort of patients.
The mechanism via DCC loss confers poor prognosis is not completely clear. As previously mentioned, DCC loss may be involved in the regulation of cell–cell, cell–stroma interactions. DCC may, therefore, control cell growth, differentiation and metastatic potential. Consistent with these observations, a direct association between ß-catenin and DCC levels was found in our study indicating that DCC down-regulation may confer increased metastatic potential through ß-catenin loss. However, cell biology experiments are required to confirm this association.
In conclusion, we demonstrate that DCC tumor protein levels, evaluated by AQUA, determine patient outcome in ovarian cancer possibly through regulation of ß-catenin levels. These results, if validated in a second cohort, may identify patients who will fail conventional chemotherapy and should receive additional therapy on experimental protocols.
|
Acknowledgements |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
This study was funded by Yale School of Medicine Institutional startup funds (AP).
|
Footnotes |
|---|
Where 
indicates the probability density function, Pmin is the minimum P value generated by evaluating multiple cut points, 
is the (1 – Pmin/2)-quantile of the standard normal distribution and 
denotes the proportion of values excluded from consideration as an optimal cut point. 
Received for publication April 16, 2006. Revision received July 18, 2006. Accepted for publication July 18, 2006.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. Jemal A, Tiwari RC, Murray T, et al. (2004) Cancer statistics, 2004. CA Cancer J Clin 54:8–29.
2. Cho KR and Fearon ER. (1995) DCC: linking tumor suppressor genes and altered cell surface interactions in cancer? Eur J Cancer 31A:1055–1060.
3. Uchino S, Tsuda H, Noguchi M, et al. (1992) Frequent loss of heterozygosity at the DCC locus in gastric cancer. Cancer Res 52:3099–3102.
4. Fang DC, Jass JR, Wang DX. (1998) Loss of heterozygosity and loss of expression of the DCC gene in gastric cancer. J Clin Pathol 51:593–596.[Abstract]
5. Cropp CS, Lidereau R, Campbell G, et al. (1990) Loss of heterozygosity on chromosome 17 and 18 in breast carcinoma: two additional regions identified. Proc Natl Acad Sci U S A 87:7737–7741.
6. Gao X, Honn KV, Grignon D, et al. (1993) Frequent loss of expression and loss of heterozygosity of putative tumor suppressor gene DCC in prostatic carcinomas. Cancer Res 353:2723–2727.
7. Peng HQ, Bailey D, Bronson D, et al. (1995) Loss of heterozygosity of tumor suppressor genes in testis cancer. Cancer Res 55:2871–2875.
8. Jen J, Kim H, Piantadosi S, et al. (1994) Allelic loss of chromosome 18q and prognosis in colorectal cancer. N Engl J Med 331:213–221.
9. Watanabe T, Tsung-The W, Catalano PJ, et al. (2001) Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 344:1196–1206.
10. Shibata D, Reale MA, Lavin P, et al. (1996) The DCC protein and prognosis in colorectal cancer. N Engl J Med 335:1727–1732.
11. Sun XF, Rutten S, Zhang H, Nenskjold B. (1999) Expression of the deleted in colorectal cancer gene is related to prognosis in DNA diploid and low proliferative colorectal adenocarcinoma. J Clin Oncol 17:1745–1750.
12. Fujita M, Enomoto T, Murata Y. (2003) Genetic alterations in ovarian carcinoma: with specific reference to histological subtypes. Mol Cell Endocrinol 202:97–99.[CrossRef][ISI][Medline]
13. Takakura S, Okamoto A, Saito M, et al. (1999) Allelic imbalance in chromosome band 18q21 and SMAD4 mutations in ovaraina cancers. Genes Chromosomes Cancer 24:264–271.[CrossRef][ISI][Medline]
14. Chenevix-Trench G, Leary J, Kerr J, et al. (1992) Frequent loss of heterozygosity on chromosome 18 in ovarian adenocarcinoma, which does not always include the DCC locus. Oncogene 7:1059–1065.[ISI][Medline]
15. Saegusa M, Machida D, Okayasu I. (2000) Loss of DCC gene expression during ovarian tumorigenesis: relation to tumor differentiation and progression. Br J Cancer 82:571–578.[CrossRef][ISI][Medline]
16. Lassus H, Salovaara R, Aaltonen LA, Butzow R. (2001) Allelic analysis of serous ovarian carcinoma reveals two putative tumor suppressor loci at 18q22-q23 distal to SMAD4, SMAD2, and DCC. Am J Pathol 159:35–42.
17. Schwartz DR, Kardia SLR, Shedden KA, et al. (2002) Gene expression in ovarian cancer reflects both morphology and biological behavior, distinguishing clear cell from other poor-ptognosis ovarian carcinomas. Cancer Res 62:4722–4729.
18. Reyes-Mugica M, Meyerhardt JA, Rzasa J, et al. (2001) Truncated DCC reduces N-cadherin/catenin expression and calcium-dependent cell adhesion in neuroblastoma cells. Lab Invest 81:201–210.[ISI]
19. Camp RL, Chung GG, Rimm DL. (2002) Automated subcellular localization and quantification of protein expression in tissue microarrays. Nat Med 8:1323–1327.[CrossRef][ISI][Medline]
20. Miller AB, Hoogstraten B, Staquet M, et al. (1981) Reporting results of cancer treatment. Cancer 47:207–214.[CrossRef][ISI][Medline]
21. Psyrri A, Bamias A, Yu Z, et al. (2005) Subcellular localization and protein levels of cyclin-dependent kinase inhibitor p27 independently predict for survival in epithelial ovarian cancer. Clin Cancer Res 11:8384–8390.
22. Camp RL, Dolled-Filhart M, Rimm DL. (2004) X-tile: a new bio-informatics tool for biomarker assessment and outcome-based cut-point optimization. Clin Cancer Res 10:7252–7259.
23. Altman DG, Lausen B, Sauerbrei W, Schumacher M. (1994) Dangers of using ‘optimal’ cutpoints in the evaluation of prognostic factors. J Natl Cancer Inst 86:829–835.
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in Ann Oncol:
- in this issue
Ann Oncol 2006 17: 1725.[Extract] [Full Text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||