Clinical Cancer Research Grants Advances in Breast Cancer
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
Clinical Cancer Research 13, 6153, October 15, 2007. doi: 10.1158/1078-0432.CCR-07-0671
© 2007 American Association for Cancer Research
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 Kulkarni, A. A.
Right arrow Articles by Stoeber, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kulkarni, A. A.
Right arrow Articles by Stoeber, K.

Imaging, Diagnosis, Prognosis

DNA Replication Licensing Factors and Aurora Kinases are Linked to Aneuploidy and Clinical Outcome in Epithelial Ovarian Carcinoma

Anjana A. Kulkarni1, Marco Loddo1, Elisabetta Leo2, Mohammed Rashid1, Kathryn L. Eward4, Thomas R. Fanshawe5, Jessica Butcher4, Alison Frost4, Jonathan A. Ledermann3, Gareth H. Williams1,2 and Kai Stoeber1,2

Authors' Affiliations: 1 Department of Pathology, Royal Free and University College Medical School, 2 Wolfson Institute for Biomedical Research, 3 Cancer Research UK and UCL Cancer Trials Centre, Department of Oncology, University College London, London, United Kingdom; and 4 Faculty of Science and Technology, Anglia Ruskin University, and 5 Centre for Applied Medical Statistics, Department of Public Health and Primary Care, Forvie Site, Institute of Public Health, University of Cambridge, Cambridge, United Kingdom

Requests for reprints: Gareth H. Williams, Department of Pathology, Royal Free and University College Medical School, University College London, Rockefeller Building, University Street, London, WC1E 6JJ, United Kingdom. Phone: 44-207679-6304; Fax: 44-207388-4408; E-mail: gareth.williams{at}ucl.ac.uk.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose: DNA replication licensing factors and Aurora kinases play critical roles in maintaining genomic integrity. We used multiparameter analyses of these cell cycle regulatory proteins to investigate their role in the progression of epithelial ovarian carcinoma (EOC).

Experimental Design: In a cohort of 143 patients, we linked the protein expression profiles of the proliferation marker Ki67, the replication licensing factors Mcm2 and geminin, and the Aurora A and B kinases to tumor DNA ploidy status and clinical outcome.

Results: Ki67, Mcm2, geminin, and Aurora A and B are significantly associated with tumor grade and ploidy status (P < 0.0001). Aurora A and its substrate H3S10ph are also significantly associated with Federation of International Obstetricians and Gynecologists tumor stage (P = 0.006 and P = 0.002, respectively). Aurora A and tumor ploidy status are predictive of disease-free survival in this cohort [hazard ratio (HR), 1.29; 95% confidence intervals (95% CI), 1.06-1.58, P = 0.01 and HR, 1.80 (1.05-3.08), P = 0.03, respectively], with Aurora A of particular prognostic importance in early stage disease [HR, 1.72 (1.19-2.48), P = 0.004 for disease-free survival and HR, 1.81 (1.14-2.87), P = 0.01 for overall survival].

Conclusions: Our data show that Ki67, Mcm2, geminin and Aurora A and B can be used as an adjunct to conventional prognostic indicators and as an aid to develop a tumor progression model for EOC. Dysregulation of Aurora A seems to be an early event in EOC with a key role in tumor progression. In view of present drug development programs for specific Aurora kinase inhibitors, our findings have important implications for the use of Aurora A as a biomarker and as a potential therapeutic target.

Epithelial ovarian carcinoma (EOC) is the fourth most common cancer in women in the U.S. and the U.K. (1). It often presents with advanced disease, and despite improvements in drug therapy, survival is poor (2). Tumor stage is the most important prognostic factor. Residual disease after surgery, histologic subtype, and tumor grade also predict survival (3, 4), but give little information about the biological variables responsible for stage progression and outcome. The pathogenesis of EOC remains poorly understood and a clear model of tumor progression has not been firmly established. DNA damage in the ovarian surface epithelium due to repetitive ovulation, defective DNA repair mechanisms, and dysregulation of cell proliferation are all postulated to be involved (5). Aggressive tumors are characterized by increased proliferation and chromosomal aberrations (6), with ~45% of stage I and 75% of stage III and IV tumors exhibiting aneuploidy (7). Controversy remains as to whether measurement of DNA ploidy has a proven role in the clinical management of EOC. Several studies, however, have shown that tumor ploidy status is of prognostic significance in both early and advanced stage disease (8).

Genomic instability, a hallmark of malignancy, leads to the acquisition of alterations in specific genes (9). Such alterations may arise from gene mutations and microsatellite instability, or from gross changes in DNA content due to chromosomal instability, which in turn is intricately linked to tumor aneuploidy. Whether aneuploidy drives tumor progression or is a consequence of malignancy has long been a matter of debate (10, 11). However, recent accumulating evidence has shown that specific cell cycle regulatory proteins involved in DNA replication and mitosis play crucial roles in maintaining genomic integrity (12, 13).

Precise duplication of DNA during each cell division cycle is essential for genomic stability and is achieved through tightly regulated initiation events. DNA replication initiation depends on the assembly of prereplicative complexes at replication origins during late mitosis and early G1 phase. Prereplicative complex assembly involves sequential binding of origin recognition complex, Cdc6, Cdt1, and Mcm2-7 to origins and renders chromatin "licensed" for DNA synthesis during S phase. At the G1-S transition, CDKs and the Cdc7/ASK kinase trigger a conformational change in the prereplicative complexes resulting in the recruitment of DNA polymerase {alpha} and elongation factors and initiation of DNA synthesis (14). We and others have shown that the regulation of Mcm2-7 protein levels provides a powerful downstream mechanism for controlling cell proliferation in human tissues (15). Mcm2-7 dysregulation is an early event in multistep tumorigenesis, and these biomarkers are now being widely exploited in cancer screening and diagnosis (16). Inhibition of prereplicative complex reassembly, which ensures that origins are fired once-and-only-once per cell cycle, is critical for maintaining genomic integrity. The licensing repressor geminin is expressed at high levels during S-G2-M phases and blocks Mcm2-7 reloading onto chromatin through its interaction with Cdt1 (17). In human cell populations in vivo, we have shown that geminin expression is also restricted to S-G2-M phases (18, 19). Importantly, depletion of geminin results in profound genomic instability with overreplication of DNA, resulting in the emergence of cells with giant aneuploid nuclei, which are the morphologic/pathologic hallmarks of aggressive cancers (12). Inactivation of geminin also causes centrosome overduplication, which, together with abrogated G2-M checkpoint mechanisms, results in multiple mitotic defects that may promote chromosome missegregation and aneuploidy (20). These findings emphasize the key role that geminin plays in maintaining genomic integrity at multiple stages of the cell cycle.

The Aurora kinases are important regulators of several stages of mitosis, including centrosome maturation and separation, chromosome orientation and segregation, and cytokinesis (13, 21). Dysregulation of Aurora A and Aurora B has been implicated as a cause of malignant transformation in cultured mammalian cells and in the development of aneuploidy (22–24). These kinases are frequently overexpressed in human tumors (25–28), with the gene for Aurora A located at 20q13.2 being commonly amplified in various epithelial malignant tumors including EOC (29). Like geminin, endogenous levels of the Aurora kinases are tightly regulated in a cell cycle–dependent manner, with low levels at G1-S, accumulation during G2-M, and rapid degradation at the end of mitosis (22, 30, 31). Their functional role as essential regulatory kinases has highlighted the Aurora family members as promising targets for anticancer drug development.

In light of the biological, prognostic, and therapeutic implications of these cell cycle regulators in tumorigenesis, we have investigated their role in the progression of EOC. We have used multiparameter analysis of Mcm2, geminin, Aurora A and Aurora B, and their substrate histone H3 (32, 33) to study the cell cycle kinetics of this tumor type in vivo. Furthermore, we have linked the protein expression of these biomarkers to genomic instability, clinicopathologic variables, and clinical outcome in a series of 143 patients. Our data provide new insights into the biological mechanisms underlying tumor progression in EOC, and how these biomarkers might be exploited to predict tumor behavior. Moreover, our findings provide further target validations for novel therapeutic approaches targeting the DNA replication licensing pathway and the mitotic machinery (34, 35).


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Study cohort. One hundred and forty-three patients diagnosed with EOC between January 1, 1999 and December 31, 2004 were identified from the Ovarian Carcinoma Database held in the Department of Oncology (University College London Hospital Gynaecological Cancer Centre, UCL Hospitals, London, UK). Patients were selected on the basis of available histologic material. Histologic specimens had been reviewed by a gynecological oncology pathologist at diagnosis and assessed for histologic subtype and nuclear grade according to WHO criteria. Most patients had been reviewed after completing treatment every 3 to 6 months for 2 years, and annually thereafter. The following clinical information was obtained directly from patients' hospital notes: date of birth, date of diagnosis, operative findings including amount of residual disease, Federation of International Obstetricians and Gynecologists (FIGO) stage based on findings at clinical examination and surgical exploration together with cytology results, CA125 values at diagnosis and relapse, performance status at start of chemotherapy, date of relapse, date of last follow-up, and date and cause of death. Of the 143 patients, 67 (47%) relapsed within the study period. Mean time to relapse among those who relapsed was 16.9 months (SD, 11.0 months; range, 0-47 months). Mean follow-up time among those who had not yet relapsed was 33.2 months (SD, 18.5 months; range 5-75 months). Thirty-four of the patients (24%) died within the study period and 107 were still alive at the last follow-up. Mean survival time among those who had died was 21.9 months (SD, 15.6 months; range 0-60 months). Mean follow-up time among those who had not yet died was 33.3 months (SD, 18.8 months; range, 5-75 months). Two patients were lost to follow-up. Ethics committee approval was obtained from the joint UCL/UCLH Committees on the Ethics of Human Research.

Antibodies. Rabbit polyclonal antibody against human geminin was generated as described (18). Ki67 monoclonal antibody (clone MIB-1) was obtained from DAKO, Mcm2 monoclonal antibody (clone 46) was from BD Transduction Laboratories, Aurora A monoclonal antibody NCL-L-AK2 (clone JLM28) was from Novocastra Laboratories, Aurora B polyclonal antibody Ab2254 was from Abcam PLC, and histone H3 phosphorylated on serine 10 (H3S10ph) polyclonal antibody was from Upstate.

Cell culture and synchronization. HeLa S3 cells (European Collection of Animal Cell Cultures 87110901) were cultured and synchronized as described (36). Cell cycle synchronization was verified by flow cytometry of isolated nuclei as previously described (37).

Preparation of protein extracts and immunoblotting. HeLa S3 cells were harvested by treatment with trypsin, washed in PBS, and resuspended in lysis buffer [50 mmol/L Tris-Cl (pH 7.5), 150 mmol/L NaCl, 20 mmol/L EDTA, 0.5% NP40] at 2 x 107 cells/mL. After incubation on ice for 30 min, the lysate was clarified by centrifugation (13,000 x g, 15 min, 4°C). Lysates were separated by 4% to 20% SDS-PAGE (75 µg protein/well) and immunoblotted as previously described (36). Blocking, antibody incubations, and washing steps were done using the following conditions: PBS/0.1% Tween 20/5% milk for Mcm2 and Aurora A, PBS/1% Tween 20/10% milk for geminin, and PBS/5% milk for Aurora B and H3S10ph.

Immunohistochemistry. Archival formalin-fixed, paraffin-embedded tissue obtained at initial diagnosis was available for all patients, and for each specimen, a block was chosen that contained a representative sample of invasive tumor. Consecutive serial sections cut from each paraffin-embedded tissue block were used for immunohistochemistry. Three-micrometer sections were cut onto Superfrost Plus slides (Vision BioSystems), dewaxed in xylene, and rehydrated through graded alcohol to water. Tissue sections were pressure-cooked in 0.1 mol/L citrate buffer at pH 6.0 for 2 min and immunostained using the Bond Polymer Define Detection kit and Bond-X automated system (Vision BioSystems). Primary antibodies were applied at the following dilutions: Ki-67 (1:100), Mcm2 (1:2,000), geminin (1:600), Aurora A (1:50), Aurora B (1:200), and H3S10ph (1:300). Coverslips were applied with Pertex mounting medium (CellPath Ltd.). Incubation without primary antibody was used as a negative control and colonic epithelial sections as positive controls.

Protein expression profile analysis. Protein expression analysis was done by determining the labeling index of the markers in each tumor, as previously described (38, 39). Slides were evaluated at low-power magnification (x100) to identify regions of tumor with the highest intensity of staining. From these selected areas, three to five fields at x400 magnification were captured with a charged coupled device camera and analysis software (SIS). Images were subsequently printed for quantitative analysis, which was undertaken with the observer unaware of the clinicopathologic variables. Both positive and negative cells within the field were counted and any stromal or inflammatory cells were excluded. Criteria for the identification of positive cells were dependent on the biomarker. For Ki67, Mcm2, geminin, Aurora B, and H3S10ph, cells with any degree of nuclear staining were scored positive. For Aurora A, cells with any degree of nuclear or cytoplasmic staining were scored positive (27). A minimum of 500 cells were counted for each case. The labeling index was calculated using the following formula: labeling index = number of positive cells / total number of cells x 100. Reassessment of 10 randomly selected cases by an independent assessor showed high levels of agreement.

DNA image cytometry. For each case, one 40-µm section of paraffin-embedded tissue obtained from the same block as that assessed by immunohistochemistry was used to prepare nuclei as described (40, 41). The Fairfield DNA Ploidy System (Fairfield Imaging, Ltd.) was used for image processing, analysis, and classification as previously described (40). Lymphocytes and plasma cells were included as internal controls and 40-µm sections of high-grade bladder tumor and normal colonic tissue as external controls for aneuploid and diploid populations, respectively. Histograms were classified according to published criteria (40, 41). Histograms were classified by two independent assessors with a high level of agreement without knowledge of clinicopathologic variables. For statistical analysis, tetraploid and polyploid tumors were grouped together with aneuploid tumors.

Statistical analysis. Spearman's rank correlation coefficient was used to examine associations between biomarkers. Relationships between biomarker expression and tumor grade, stage, and ploidy status were assessed using nonparametric Jonckheere-Terpstra and Mann-Whitney U tests as appropriate. Data were then summarized as the median value and interquartile range of labeling indices observed across the cohort. Analysis of disease-free and overall survival data was carried out using Kaplan-Meier plot (using tertiles for biomarkers), log-rank test, and Cox regression (treating biomarkers as continuous variables unless stated otherwise). For each biomarker, the cohort was divided into tertile groups on the basis of the labeling index. Within each tertile group, the proportion remaining that was either disease-free or alive, for disease-free and overall survival, respectively, was calculated using the Kaplan-Meier method. Hazard ratios (HR) with 95% confidence intervals (95% CI) for biomarkers were first estimated unadjusted, and then adjusted for age, grade and stage. Patients with incomplete data were excluded from the multivariate analysis. Candidate biomarkers are listed in Supplementary Table S1. All tests were two-sided and used a significance level of 0.05 and no allowances were made for multiple hypothesis testing. Analysis was carried out using SPSS 12.0 for Windows (SPSS, Inc.).


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Validation of biomarker multiparameter analysis and its biological implications
The monospecificity of antibodies against Mcm2, geminin, Aurora A, Aurora B, and H3S10ph was confirmed in total cell extracts from asynchronous HeLa S3 cells by detection of a single protein with a molecular mass consistent with the reported electrophoretic mobility of the corresponding human antigen (Fig. 1A ). HeLa S3 cells were selected in the first instance for in vitro studies as this line has well characterized cell cycle phase transit times and established synchronization protocols. Total cell lysates from synchronized cells were immunoblotted with the characterized antibodies (Fig. 1B). Mcm2 levels did not vary significantly during passage through the cell cycle, whereas geminin expression was restricted to S-G2-M as previously reported (18). Aurora A levels increased during S phase and peaked during mitosis, with degradation occurring 2 to 4 h after release from mitotic arrest. Similarly, Aurora B levels were negligible during G1 phase, increased gradually during S phase to reach a peak during G2-M, and decreased after mitosis. The presence of H3S10ph was restricted to mitosis as previously reported (42), consolidating the rationale for its use as a mitotic marker. Identical cell cycle–dependent expression of these biomarkers was observed in synchronized SK-OV3 ovarian cancer cells (data not shown). Because Ki67 is expressed throughout the cell cycle in proliferating cells and geminin expression is restricted to the S-G2-M phase, we have proposed that the geminin/Ki67 ratio may be used as an indicator of the relative length of G1, and therefore, the rate of cell cycle progression (18, 19). The data described above confirms that cell cycle–dependent expression of Aurora A and Aurora B also enables the use of their ratios with Ki67 as indicators of cell cycle progression. Increased geminin expression has been noted in several malignancies and correlates positively with cell proliferation (43). Notably, this increased expression is always restricted to the S-G2-M phase, even in highly aggressive tumors. Our in vitro findings therefore also indicate that an increase in the relative ratio between Aurora A or Aurora B and geminin (ratio > 1) would be indicative of overexpression of the kinase during the cell cycle.


Figure 1
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. A, immunoblots of asynchronous HeLa S3 total cell lysates with antibodies against Mcm2, geminin, Aurora A, Aurora B, and H3S10ph. B, immunoblots of biomarkers and actin (loading control) in total cell lysates from synchronized HeLa S3 cells. Fluorescence-activated cell sorting profiles of synchronized HeLa S3 cells at 2-h intervals. C, photomicrographs of paraffin-embedded tissue sections of EOC immunohistochemically stained with antibodies against Ki67, Mcm2, geminin, Aurora A, Aurora B, and H3S10ph. Original magnification, x400 (insets, x1,000).

 
To assess the prognostic significance of our in vitro findings and their biological implications in EOC, we analyzed the expression of biomarkers in a series of 143 cases (Fig. 1C). Protein expression was also studied in five cases of normal ovarian tissue. Expression of the biomarkers was extremely low (<4%) in normal ovarian surface epithelium, in keeping with its lowered proliferative capacity (data not shown). By contrast, EOC showed high levels of biomarker expression, indicative of cell cycle re-entry and proliferation. Next, we examined the correlations between pairs of biomarkers across the tumor series. The expression levels of Aurora A and Aurora B showed a strong positive correlation with those of their substrate H3S10ph [Spearman correlation, 0.57 (95% CI, 0.45-0.67) and 0.52 (95% CI, 0.39-0.63), respectively]. The expression levels of Mcm2, geminin, and H3S10ph were strongly positively correlated with Ki67 levels [Spearman correlation, 0.73 (95% CI, 0.64-0.8); 0.74 (95% CI, 0.66-0.81); and 0.52 (95% CI, 0.39-0.63), respectively], supporting their role as proliferation markers. Notably, the geminin/Ki67 and Aurora A/Ki67 ratios were positively, but less strongly, correlated with H3S10ph [Spearman correlation, 0.25 (95% CI, 0.09-0.40) and 0.42 (95% CI, 0.27-0.55), respectively], which reflects changes in the relative length of G1 as opposed to prolonged S-G2-M transit times that might arise through the activation of intra-S or G2-M checkpoint pathways. This is consistent with a tumor mass only becoming clinically detectable after the tumor cell has undergone a major proportion of its population doublings, approximately 30 doublings out of a total of 40. This number of population doublings represents the maximum mass compatible with life, a point in somatic clonal evolution in which most cell cycle checkpoint mechanisms have been overridden.

Relationship between biomarkers, tumor DNA ploidy status, and clinicopathologic characteristics
The clinicopathologic characteristics of the study cohort are summarized in Table 1 . To investigate the relationship between the biomarkers and genomic instability, we linked their expression profiles to tumor DNA content. There was a highly significant association between the expression levels of all of the biomarkers and several biomarker/Ki67 ratios and genomic instability (Table 2 ), reflecting an increased proportion of cycling cells and accelerated cell cycle progression in aneuploid tumors as compared with diploid tumors.


View this table:
[in this window]
[in a new window]

  		Table 1. Patient characteristics

 

View this table:
[in this window]
[in a new window]

  		Table 2. Relationship between biomarkers and tumor DNA ploidy status

 
All six biomarkers were also strongly associated with tumor grade (Table 3 ); however, there is some overlap in the distributions of biomarker levels between grades (e.g., Aurora A levels; Supplementary Fig. S1). These data confirm an increasing proportion of cycling cells with increasing tumor anaplasia, but also indicate that the biomarkers do not fully distinguish between grades for all patients within each grade. In keeping with these findings, a highly significant association between tumor grade and ploidy status was found (P < 0.001). The ratios among geminin/Ki67, Aurora A/Ki67, Aurora B/Ki67, and H3S10ph/Ki67 were also significantly associated with tumor differentiation, indicative of an accelerated rate of cell cycle progression in high-grade tumors. By contrast, and consistent with our findings in other tumor types (38, 39), the Mcm2/Ki67 ratio decreased with increasing tumor grade. This reflects a shift in the proportion of DNA replication licensed but nonproliferating cells in well-differentiated tumors to actively cycling cells in poorly differentiated tumors. The positive correlation between geminin expression and increasing tumor anaplasia and genomic instability indicates that this licensing repressor does not behave as a tumor suppressor in EOC. This has also been observed in other tumor types (19), in which the number of geminin-expressing cells is proportional to the cell proliferation index (43). A significant association was found between Aurora A, H3S10ph, Aurora A/Ki67, H3S10ph/Ki67 and tumor stage (Table 4 ). In keeping with a previous study (27), this suggests that Aurora A dysregulation might be a key event in early epithelial ovarian tumorigenesis and progression to advanced stage disease. Furthermore, advanced stage disease was significantly associated with an increase in the Aurora A/geminin ratio (P = 0.04; Table 4), also supporting a link between Aurora A dysregulation and tumor progression.


View this table:
[in this window]
[in a new window]

  		Table 3. Relationship between biomarkers and tumor differentiation

 

View this table:
[in this window]
[in a new window]

  		Table 4. Relationship between biomarkers and tumor stage

 
Relationship between biomarkers, tumor DNA ploidy status, and patient outcome
Univariate analysis. Aurora A (P = 0.01; Fig. 2A ), Aurora A/Ki67, Aurora B/Ki67, and H3S10ph (Supplementary Table S1) were all significantly associated with shorter disease-free survival but not overall survival. Patient age, tumor grade, and stage were also predictive of disease-free survival, with younger patients, well-differentiated tumors, and particularly, early stage disease having a significantly longer time-to-relapse [HR, 1.02 (1.00-1.05), P = 0.05; HR, 1.59 (1.03-2.45), P = 0.04; HR, 2.07 (1.58-2.71), P < 0.0001, respectively]. Patient age and tumor stage also predicted overall survival [HR, 1.05 (1.02-1.09), P = 0.003; HR, 3.21 (1.33-7.79), P = 0.01, respectively] but tumor grade did not (P = 0.70), emphasizing the limitations of the current grading systems (44). Tumor ploidy status also significantly correlated with disease-free survival [HR, 1.80 (1.05-3.08), P = 0.03; Fig. 2B], with a trend towards shorter overall survival in patients with aneuploid tumors, although this did not reach statistical significance [HR, 1.95 (0.88-4.31), P = 0.10].


Figure 2
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Kaplan-Meier curves showing association between Aurora A, tumor ploidy status, and patient survival. A, Aurora A (lower tertile <11.3%, middle tertile 11.3-21.3%, upper tertile >21.3%) and disease-free survival across the whole series; log-rank test, P = 0.01. B, tumor ploidy status and disease-free survival across the whole series; log-rank test, P = 0.03. C, Aurora A (lower tertile <8.7%, middle tertile 8.7-19.6%, upper tertile >19.6%) and disease-free survival in early stage subgroup; log-rank test, P = 0.004. D, tumor ploidy status and disease-free survival in early stage subgroup; log-rank test, P = 0.04. E, Aurora A (lower tertile <8.7%, middle tertile 8.7-19.6%, upper tertile >19.6%) and overall survival in early stage subgroup; log-rank test, P = 0.01. F, tumor ploidy status and overall survival in early stage subgroup; log-rank test, P = 0.08.

 
We subdivided our series into two groups; early stage disease (FIGO stages I and II) and advanced stage disease (FIGO stages III and IV) to more precisely define the specific subgroups for which the biomarkers may have particular prognostic importance. Both Aurora A and the Aurora A/Ki67 ratio were strongly predictive of shorter disease-free survival [HR, 1.72 (1.19-2.48), P = 0.004 (Fig. 2C); HR, 1.59 (1.13-2.24), P = 0.008, respectively] and overall survival [HR, 1.81 (1.14-2.87), P = 0.01 (Fig. 2E); HR, 1.68 (1.11-2.54), P = 0.01, respectively] in the early stage subgroup. This association was not found in the advanced stage subgroup [HR, 1.06 (0.81-1.37), P = 0.67; HR, 1.04 (0.90-1.20), P = 0.58 for disease-free survival; and HR, 0.88 (0.58-1.33), P = 0.88; HR, 0.89 (0.69-1.15), P = 0.36 for overall survival, respectively]. Tumor ploidy status also predicted disease-free survival [HR, 4.58 (1.04-20.19), P = 0.04; Fig. 2D], with a trend towards shorter overall survival in patients with aneuploid tumors [HR, 6.34 (0.82-49.18), P = 0.08; Fig. 2F] in the early stage subgroup. However, it lost its predictive value in the advanced stage subgroup [HR, 1.47 (0.81-2.66), P = 0.21 and HR, 1.36 (0.55-3.33), P = 0.5 for disease-free survival and overall survival, respectively].

Multivariate analysis. Cox regression survival analysis showed that tumor stage was the only significant independent predictor of disease-free survival [HR, 2.06 (1.49-2.85), P < 0.0001]. Patient age and tumor stage were independent predictors of overall survival, with older patients and advanced stage tumors having shorter overall survival times [HR, 1.05 (1.01-1.09), P = 0.007; HR, 3.19 (1.31-7.75), P = 0.01, respectively]. Although several biomarkers were significant prognostic factors in univariate analysis, none was a significant predictor of disease-free or overall survival after adjustment for age, grade, and stage. This is due partly to the highly significant associations between the biomarkers and tumor grade and stage, making it difficult to separate their independent effects.


   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
This study was undertaken to gain further insight into biological markers of EOC that may be of prognostic and predictive value and could lead to a greater understanding of its pathogenesis. Our findings show that the Mcm2 and geminin replication licensing factors, and the Aurora A and B kinases, together with their substrate H3S10ph (33, 34), are of prognostic value in EOC. The association found between tumor differentiation and this set of biomarkers has been noted in several other malignancies (18, 19, 27, 38, 45, 46) and has implications for their use as proliferation markers with potential for further improvements in the current grading system. Our multiparameter analysis shows that it could also be used to provide information about cell cycle progression in patient tumor samples, data that translate into important prognostic information. These findings are in keeping with our more limited analysis of licensing factors in breast and renal cell cancer (38, 39). Furthermore, the highly significant association found between this set of biomarkers and tumor ploidy status suggests that dysregulation of the licensing machinery and mitotic kinases is intricately linked to the development of genomic instability in EOC.

Aurora A plays a regulatory role in several key stages of the G2-M transition (13, 21) and compelling evidence exists for its oncogenic potential (22, 25–29, 47). Here, we report an intriguing link between aberrant regulation of Aurora A and EOC progression. Our data show highly significant associations between Aurora A, H3S10ph and tumor FIGO stage, supporting a previous report that Aurora A dysregulation might be an early event in epithelial ovarian carcinogenesis (27) and suggesting that its dysregulation might play a role in the progression to advanced disease. In line with these findings, the Aurora A/geminin ratio is significantly higher in advanced stage tumors, which also suggests that Aurora A overexpression might play a role in or might be a result of tumor progression. In vivo, it is likely that such overexpression is regulated by not only gene amplification but also other mechanisms such as transcriptional activation and suppression of protein degradation (21).

Our data suggest that multiparameter analysis of Aurora A and H3S10ph allows molecular staging which could be used to complement clinical staging methods. FIGO stage is an important prognostic indicator in EOC, however, surgical and radiological staging methods have their limitations. Randomized trials have shown that adjuvant chemotherapy is of particular benefit in suboptimally staged patients with stage I disease (48). However, in a recent large randomized controlled trial, only 34% of patients were optimally staged according to guidelines (49). In those patients who were inadequately staged, these biomarkers might provide supportive evidence of either true stage I or more advanced disease, assisting a decision about the use of adjuvant chemotherapy. The findings of the ICON 1/ACTION trials suggested a small overall benefit for adjuvant chemotherapy (50), but it remains unclear which adequately staged patients with stage I disease really need chemotherapy. The additional prognostic information from Aurora A and H3S10ph should be tested in a larger data set of patients with stage I cancer in which outcome data and chemotherapy usage are known. However, it should be noted that the prognostic value of these biomarkers does not have independent prognostic importance for more advanced disease.

Aurora A, H3S10ph, and genomic instability are also significant predictors of disease-free survival in this study cohort. Further subgroup analysis showed that Aurora A and tumor ploidy status are predictive of disease-free survival (with Aurora A expression also predicting overall survival) in early disease. However, this association was reduced when prognostic factors such as age and stage were taken into account. By contrast, these variables lost their predictive value in advanced disease [e.g., HR for Aurora A dropped from 1.72 (1.19-2.48) in early stage disease to 1.06 (0.81-1.37) for advanced stage disease], suggesting that other biological factors may take precedence in influencing relapse and outcome in these patients. In addition, the complexity and heterogeneity of treatment regimens might mask the predictive value of these biomarkers in advanced stages. Several studies have shown the prognostic significance of DNA ploidy in EOC (8), however, the underlying mechanisms remain unclear. Taken together, our data are supportive of a biological mechanism by which Aurora A dysregulation at an early point during tumorigenesis might contribute to genetic instability, resulting in aggressive tumors and shorter survival in a subgroup of patients with early stage disease.

DNA replication licensing factors and mitotic kinases are critical regulators of cell cycle progression, and thus, are the focus of current therapeutic drug development programs (32, 35). Here, we have shown that multiparameter expression analysis of core regulators of the G1-S and G2-M transitions allows the assessment of the rate of cell cycle progression in individual patient tumor samples, variables linked to the biological behavior of these tumors. This type of analysis could be used as a predictive test for small molecules targeting the cell cycle machinery or upstream growth signal transduction pathways that accelerate cell cycle progression. Moreover, the observation that Aurora A expression does not fully distinguish between grades shows that traditional clinicopathologic variables do not always allow the prediction of therapeutic response, supporting the concept of coevolution of biomarker and individualized targeted therapy. In view of the recent development of specific Aurora kinase inhibitors, our data have important implications—prognostic, predictive, and therapeutic—for the use of Aurora A as a biomarker and potential therapeutic target. These findings need to be confirmed in a large patient cohort, and it will be of interest to determine whether this form of multiparameter cell cycle analysis is of diagnostic utility in other tumor types.


   Acknowledgments
 
We thank Drs. Julie Crowe (Royal Free Hospital), Su Ramachandra (Whittington Hospital), and Erich Langner (Barnet and Chase Farm Hospitals) for providing access to and assistance with the retrieval of archival tissue specimens.


   Footnotes
 
Grant support: EU Sixth Framework Programme Integrated Project ‘MitoCheck’ and Cancer Research UK Programme grant C428/A6263 (G.H. Williams and K. Stoeber).

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.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

A.A. Kulkarni and M. Loddo contributed equally to this work.

Received 3/21/07; revised 6/28/07; accepted 7/20/07.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. NCI Surveillance, E.a.E.R.S.P.a.t.N.C.f.H.S. National Cancer Institute—a snapshot of ovarian cancer: incidence and mortality rate trends 2006 Sep. Available from: http://planning.cancer.gov/disease/ovarian-snapshot.pdf.
  2. Holschneider CH, Berek JS. Ovarian cancer: epidemiology, biology, and prognostic factors. Semin Surg Oncol 2000;19:3–10.[CrossRef][Medline]
  3. Kosary CL. FIGO stage, histology, histologic grade, age and race as prognostic factors in determining survival for cancers of the female gynecological system: an analysis of 1973–87 SEER cases of cancers of the endometrium, cervix, ovary, vulva, and vagina. Semin Surg Oncol 1994;10:31–46.[Medline]
  4. Clark TG, Stewart ME, Altman DG, Gabra H, Smyth JF. A prognostic model for ovarian cancer. Br J Cancer 2001;85:944–52.[CrossRef][Medline]
  5. Murdoch WJ, McDonnel AC. Roles of the ovarian surface epithelium in ovulation and carcinogenesis. Reproduction 2002;123:743–50.[Abstract]
  6. Berek JS, Martinez-Maza O, Hamilton T, et al. Molecular and biological factors in the pathogenesis of ovarian cancer. Ann Oncol 1993;4 Suppl 4:3–16.[Medline]
  7. van Dam. Ploidy in ovarian cancer and prognosis. In: Leake R, Gore M, Ward RH, editors. The biology of gynaecological cancer. London: RCOG Press; 1995. p. 258–73.
  8. Fox H. Ploidy in gynaecological cancers. Histopathology 2005;46:121–9.[CrossRef][Medline]
  9. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57–70.[CrossRef][Medline]
  10. Duesberg P, Rasnick D, Li R, Winters L, Rausch C, Hehlmann R. How aneuploidy may cause cancer and genetic instability. Anticancer Res 1999;19:4887–906.[Medline]
  11. Duesberg P, Li R. Multistep carcinogenesis: a chain reaction of aneuploidizations. Cell Cycle 2003;2:202–10.[Medline]
  12. Saxena S, Dutta A. Geminin-Cdt1 balance is critical for genetic stability. Mutat Res 2005;569:111–21.[Medline]
  13. Giet R, Petretti C, Prigent C. Aurora kinases, aneuploidy and cancer, a coincidence or a real link? Trends Cell Biol 2005;15:241–50.[CrossRef][Medline]
  14. Takeda DY, Dutta A. DNA replication and progression through S phase. Oncogene 2005;24:2827–43.[CrossRef][Medline]
  15. Tachibana KE, Gonzalez MA, Coleman N. Cell-cycle-dependent regulation of DNA replication and its relevance to cancer pathology. J Pathol 2005;205:123–9.[CrossRef][Medline]
  16. Gonzalez MA, Tachibana KE, Laskey RA, Coleman N. Control of DNA replication and its potential clinical exploitation. Nat Rev Cancer 2005;5:135–41.[CrossRef][Medline]
  17. Machida YJ, Hamlin JL, Dutta A. Right place, right time, and only once: replication initiation in metazoans. Cell 2005;123:13–24.[CrossRef][Medline]
  18. Wharton SB, Hibberd S, Eward KL, et al. DNA replication licensing and cell cycle kinetics of oligodendroglial tumours. Br J Cancer 2004;91:262–9.[Medline]
  19. Obermann EC, Eward KL, Dogan A, et al. DNA replication licensing in peripheral B-cell lymphoma. J Pathol 2005;205:318–28.[CrossRef][Medline]
  20. Tachibana KE, Gonzalez MA, Guarguaglini G, Nigg EA, Laskey RA. Depletion of licensing inhibitor geminin causes centrosome overduplication and mitotic defects. EMBO Rep 2005;6:1052–7.[CrossRef][Medline]
  21. Marumoto T, Zhang D, Saya H. Aurora-A—a guardian of poles. Nat Rev Cancer 2005;5:42–50.[CrossRef][Medline]
  22. Bischoff JR, Anderson L, Zhu Y, et al. A homologue of Drosophila aurora kinase is oncogenic and amplified in human colorectal cancers. EMBO J 1998;17:3052–65.[CrossRef][Medline]
  23. Meraldi P, Honda R, Nigg EA. Aurora-A overexpression reveals tetraploidization as a major route to centrosome amplification in p53–/– cells. EMBO J 2002;21:483–92.[CrossRef][Medline]
  24. Ota T, Suto S, Katayama H, et al. Increased mitotic phosphorylation of histone H3 attributable to AIM-1/Aurora-B overexpression contributes to chromosome number instability. Cancer Res 2002;62:5168–77.[Abstract/Free Full Text]
  25. Zhou H, Kuang J, Zhong L, et al. Tumour amplified kinase STK15/BTAK induces centrosome amplification, aneuploidy and transformation. Nat Genet 1998;20:189–93.[CrossRef][Medline]
  26. Sakakura C, Hagiwara A, Yasuoka R, et al. Tumour-amplified kinase BTAK is amplified and overexpressed in gastric cancers with possible involvement in aneuploid formation. Br J Cancer 2001;84:824–31.[CrossRef][Medline]
  27. Gritsko TM, Coppola D, Paciga JE, et al. Activation and overexpression of centrosome kinase BTAK/Aurora-A in human ovarian cancer. Clin Cancer Res 2003;9:1420–6.[Abstract/Free Full Text]
  28. Jeng YM, Peng SY, Lin CY, Hsu HC. Overexpression and amplification of Aurora-A in hepatocellular carcinoma. Clin Cancer Res 2004;10:2065–71.[Abstract/Free Full Text]
  29. Tanner MM, Grenman S, Koul A, et al. Frequent amplification of chromosomal region 20q12-13 in ovarian cancer. Clin Cancer Res 2000;6:1833–9.[Abstract/Free Full Text]
  30. Kimura M, Kotani S, Hattori T, et al. Cell cycle-dependent expression and spindle pole localization of a novel human protein kinase, Aik, related to Aurora of Drosophila and yeast Ipl1. J Biol Chem 1997;272:13766–71.[Abstract/Free Full Text]
  31. Castro A, Arlot-Bonnemains Y, Vigneron S, Labbe JC, Prigent C, Lorca T. APC/Fizzy-Related targets Aurora-A kinase for proteolysis. EMBO Rep 2002;3:457–62.[CrossRef][Medline]
  32. Crosio C, Fimia GM, Loury R, et al. Mitotic phosphorylation of histone H3: spatio-temporal regulation by mammalian Aurora kinases. Mol Cell Biol 2002;22:874–85.[Abstract/Free Full Text]
  33. Hsu JY, Sun ZW, Li X, et al. Mitotic phosphorylation of histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell 2000;102:279–91.[CrossRef][Medline]
  34. Harrington EA, Bebbington D, Moore J, et al. VX-680, a potent and selective small-molecule inhibitor of the Aurora kinases, suppresses tumor growth in vivo. Nat Med 2004;10:262–7.[CrossRef][Medline]
  35. Shreeram S, Blow JJ. The role of the replication licensing system in cell proliferation and cancer. Prog Cell Cycle Res 2003;5:287–93.[Medline]
  36. Stoeber K, Tlsty TD, Happerfield L, et al. DNA replication licensing and human cell proliferation. J Cell Sci 2001;114:2027–41.[Abstract/Free Full Text]
  37. Krude T, Jackman M, Pines J, Laskey RA. Cyclin/Cdk-dependent initiation of DNA replication in a human cell-free system. Cell 1997;88:109–19.[CrossRef][Medline]
  38. Shetty A, Loddo M, Fanshawe T, et al. DNA replication licensing and cell cycle kinetics of normal and neoplastic breast. Br J Cancer 2005;93:1295–300.[CrossRef][Medline]
  39. Dudderidge TJ, Stoeber K, Loddo M, et al. Mcm2, Geminin, and KI67 define proliferative state and are prognostic markers in renal cell carcinoma. Clin Cancer Res 2005;11:2510–7.[Abstract/Free Full Text]
  40. Sudbo J, Kildal W, Risberg B, Koppang HS, Danielsen HE, Reith A. DNA content as a prognostic marker in patients with oral leukoplakia. N Engl J Med 2001;344:1270–8.[Abstract/Free Full Text]
  41. Haroske G, Giroud F, Reith A, Bocking A. 1997 ESACP consensus report on diagnostic DNA image cytometry. Part I: basic considerations and recommendations for preparation, measurement and interpretation. European Society for Analytical Cellular Pathology. Anal Cell Pathol 1998;17:189–200.[Medline]
  42. Juan G, Traganos F, James WM, et al. Histone H3 phosphorylation and expression of cyclins A and B1 measured in individual cells during their progression through G2 and mitosis. Cytometry 1998;32:71–7.[CrossRef][Medline]
  43. Wohlschlegel JA, Kutok JL, Weng AP, Dutta A. Expression of geminin as a marker of cell proliferation in normal tissues and malignancies. Am J Pathol 2002;161:267–73.[Abstract/Free Full Text]
  44. Silverberg SG. Histopathologic grading of ovarian carcinoma: a review and proposal. Int J Gynecol Pathol 2000;19:7–15.[CrossRef][Medline]
  45. Sen S, Zhou H, Zhang RD, et al. Amplification/overexpression of a mitotic kinase gene in human bladder cancer. J Natl Cancer Inst 2002;94:1320–9.[Abstract/Free Full Text]
  46. Ren B, Yu G, Tseng GC, et al. MCM7 amplification and overexpression are associated with prostate cancer progression. Oncogene 2006;25:1090–8.[CrossRef][Medline]
  47. Wang X, Zhou YX, Qiao W, et al. Overexpression of aurora kinase A in mouse mammary epithelium induces genetic instability preceding mammary tumor formation. Oncogene 2006;25:7148–58.[CrossRef][Medline]
  48. Trimbos JB, Vergote I, Bolis G, et al. Impact of adjuvant chemotherapy and surgical staging in early-stage ovarian carcinoma: European Organisation for Research and Treatment of Cancer-Adjuvant ChemoTherapy in Ovarian Neoplasm trial. J Natl Cancer Inst 2003;95:113–25.[Abstract/Free Full Text]
  49. Vergote I, Trimbos BJ. Treatment of patients with early epithelial ovarian cancer. Curr Opin Oncol 2003;15:452–5.[CrossRef][Medline]
  50. Trimbos JB, Parmar M, Vergote I, et al. International Collaborative Ovarian Neoplasm Trial 1 and Adjuvant ChemoTherapy in Ovarian Neoplasm Trial: two parallel randomized phase III trials of adjuvant chemotherapy in patients with early-stage ovarian carcinoma. J Natl Cancer Inst 2003;95:105–12.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
A. A. Kulkarni, S. R. Kingsbury, S. Tudzarova, H.-K. Hong, M. Loddo, M. Rashid, S. Rodriguez-Acebes, A. T. Prevost, J. A. Ledermann, K. Stoeber, et al.
Cdc7 Kinase Is a Predictor of Survival and a Novel Therapeutic Target in Epithelial Ovarian Carcinoma
Clin. Cancer Res., April 1, 2009; 15(7): 2417 - 2425.
[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 Kulkarni, A. A.
Right arrow Articles by Stoeber, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kulkarni, A. A.
Right arrow Articles by Stoeber, K.

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