Advertisement
Journal of Clinical Oncology  
Search for:
Limit by:
  Browse by Subject or Issue
Home Search or Browse JCO My JCO Subscriptions Customer Service Site Map

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
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 Save to my personal folders
Right arrow Download to citation manager
Right arrowRights & Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Salvesen, H. B.
Right arrow Articles by Akslen, L. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Salvesen, H. B.
Right arrow Articles by Akslen, L. A.
Journal of Clinical Oncology, Vol 17, Issue 5 (May), 1999: 1382
© 1999 American Society for Clinical Oncology

Prognostic Significance of Angiogenesis and Ki-67, p53, and p21 Expression: A Population-Based Endometrial Carcinoma Study

Helga B. Salvesen, Ole Erik Iversen, Lars A. Akslen

From the Department of Pathology, The Gade Institute, and the Department of Gynecology and Obstetrics, Haukeland University Hospital, Bergen, Norway.

Address reprint requests to Helga B Salvesen, MD, Department of Human Genetics, The University of Chicago, J.F. Knapp Medical Center, 924 E 57th St, Chicago, IL 60637; email hsalvese@ genetics.uchicago.edu.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: For endometrial carcinoma patients, there is a need for improved identification of high-risk groups that may benefit from postoperative adjuvant therapy. We therefore studied the prognostic impact of markers for cell proliferation, cell-cycle regulation, and angiogenesis among endometrial carcinoma patients in a population-based setting.

PATIENTS AND METHODS: All patients diagnosed with endometrial carcinoma between 1981 and 1985 in Hordaland County, Norway, were studied. The median follow-up for the survivors was 11.5 years (range, 8 to 15 years), with no patient lost because of insufficient follow-up information. Paraffin-embedded tumor tissue, available in 96% of the cases (n = 142), was studied immunohistochemically for microvessel density (MVD) and expression of Ki-67, p53, and p21 proteins. We used the hot spot method for calculation of MVD, and expression of Ki-67 and p21 protein, because this approach may increase the probability of detecting small aggressive clones of possible prognostic relevance. The importance of these tumor markers was investigated in univariate survival analyses and Cox regression analysis.

RESULTS: The majority of traditional clinicopathologic variables was significantly associated with the tumor biomarkers. Age, International Federation of Gynecology and Obstetrics (FIGO) stage, histologic type, histologic grade, MVD, as well as Ki-67, p53, and p21 protein expression, all significantly influenced survival in univariate analyses (P <= .05). In the Cox regression analysis, age, FIGO stage, MVD, Ki-67 expression, and p53 expression were the only variables with independent prognostic impact (P <= .05), whereas histologic type, histologic grade, and p21 expression had no independent influence. A group of high-risk patients with more than one unfavorable marker was identified.

CONCLUSION: In addition to age and FIGO stage, MVD, Ki-67, and p53 protein expression showed an independent prognostic impact. Thus, information derived from routine histologic specimens identified a subgroup of high-risk endometrial carcinoma patients in this population-based study.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
IN ENDOMETRIAL CARCINOMAS, the prognostic impact of traditional clinicopathologic variables, such as International Federation of Gynecology and Obstetrics (FIGO) stage, histologic type, and histologic grade, is well established.1-5 This information has been used to determine whether hysterectomy alone is likely to be curative. Nevertheless, there is a definite need for more specific prognostic markers to avoid overtreatment of low-risk groups and to ensure that patients with highly aggressive tumors receive adequate postoperative treatment.3,6

Recent advances in molecular biology have shown that malignant tumors may arise from alterations in genes controlling proliferation, apoptosis, angiogenesis, invasion, and metastases.7,8 Immunohistochemical methods have been useful for detecting several biomarkers of possible prognostic importance for a number of cancer types.9-19 Regarding tumor cell proliferation, it is widely accepted that proliferative capacity may influence the clinical course, and hence patient prognosis.11,20 The nuclear Ki-67 antigen, which is expressed in all stages of the cell cycle except G0, may be detected by immunohistochemistry to estimate proliferative activity in tumors.21-23 To detect small aggressive clones of possible prognostic importance, we used the hot spot method according to Weidner et al.11

Angiogenesis is also of crucial importance for tumor growth and development of metastases.24,25 Tumor cells capable of rapid proliferation are thought to be dependent on persistent blood vessels, and the absence of neovascularization would consequently restrict tumor growth.24 Vascularization is likewise considered to permit shedding of cells from a primary tumor to distant body sites, thus facilitating the metastatic process.26 The degree of intratumoral microvessel density (MVD) is thought to reflect the angiogenic activity generated by the neoplastic cells and the supporting stroma,9,27 and because the area of the tumor with the most active neovascularization has been found to be biologically important, we used the hot spot method as previously described.19,28

Over the past few years, it has been shown that p53 has a central role in the regulation of cell-cycle progression by the transcriptional activation of genes such as p21, followed by inhibition of cyclin-dependent kinases.7,8,29 Hence, p21 (also known as WAF1, CIP1, and SDI1) seems to be a critical downstream effector in the p53-specific pathway of growth control, and its inactivation could potentially lead to tumor progression.30,31 It has also been shown that p21 can be downregulated in tumors without p53 alterations, which indicates that additional mechanisms are involved in p21 inactivation.32 Furthermore, p53 has also been implicated in the regulation of angiogenesis.33

The purpose of our study of endometrial carcinoma patients was to evaluate the prognostic significance of certain tumor biomarkers for proliferation (Ki-67), cell-cycle regulation (p53 and p21), and angiogenesis (factor VIII) relative to the information derived from established clinicopathologic variables. To avoid selection bias and to reduce problems with external validity, we examined a 5-year population-based series with complete follow-up.


   PATIENTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
All patients diagnosed with endometrial carcinoma in Hordaland County, Norway, between 1981 and 1985 were studied. Hordaland County has approximately 400,000 inhabitants, representing approximately 10% of the total Norwegian population. The age-adjusted incidence rate of endometrial cancer in Hordaland was similar to that of the entire Norwegian population for the period studied, and was 13.9 per 100,000 per year in 1983.34 Data on patient characteristics, treatment, and follow-up were collected for all 157 patients diagnosed during this period. The patients were retrospectively staged according to the FIGO 1988 criteria,35 and all microscopic slides were reclassified and graded by one pathologist (L.A.A.) according to World Health Organization criteria.36 Depth of myometrial invasion was recorded as less than or equal to 50% or more than 50% of the myometrial thickness. Six cases were excluded because of a changed diagnosis at reclassification (three cases to mixed mesodermal tumors, one case to leiomyosarcoma, one case to primary adenocarcinoma of the cervix, and one case to a benign lesion). In three cases the diagnosis was based on cytologic examination only, with no histologic specimens available. Of the remaining 148 patients, paraffin blocks were available for further investigations in 142 of the cases (96%).

The initial treatment protocol for the period was abdominal hysterectomy and bilateral salpingo-oophorectomy. The pelvic and para-aortic lymph nodes were palpated and biopsied only when considered suspect. Patients who were not eligible for surgical therapy were staged by clinical examinations, curettage results, and chest x-rays. After primary surgery, 116 (82%) of 142 patients had no residual tumor. No patient received preoperative radiation therapy. For patients with myometrial tumor infiltration without distant metastases, postoperative radiation therapy was recommended. However, 27 patients did not receive this adjuvant therapy, mostly because of high age or serious intercurrent disease.

Immunohistochemistry
Hematoxylin-eosin–stained sections were used to select the most cellular part of the tumor and the area with the lowest architectural differentiation in the case of heterogeneity. Thin sections (5µm) of the selected formalin-fixed and paraffin-embedded specimens were then used for immunohistochemistry. The sections were deparaffinized through toluol to alcohol to phosphate-buffered saline (PBS), followed by epitope retrieval. Endogenous peroxidase activity was blocked by incubation with 1% hydrogen peroxide in methanol for 30 minutes. The slides were incubated with primary antibody, stained according to the standard avidin-biotin method recommended by the manufacturer (DAKO, Copenhagen, Denmark), and counterstained with Harris hematoxylin. Blinded for patient characteristics and outcome, the slides were evaluated individually in a standard light microscope for immunohistochemical staining by two of the authors (H.S. and L.A.A.). Positive controls were sections of colon cancers known to express the investigated antibodies. Negative controls were created by omitting the primary antibodies.

Ki-67 Staining
The sections were subjected to microwave epitope retrieval (750W for 7.5 minutes and 500W for four x 5 minutes in citrate buffer, pH 6.0) before incubation for 1 hour (room temperature) with the rabbit anti-human Ki-67 antibody (diluted 1/50 in PBS; polyclonal antibody, code no. A-047; DAKO). The sections were examined at low magnifications (x40 and x100) to identify the area with the most intense staining, similar to the method of Weidner et al.11 Only cells with a distinct nuclear Ki-67 staining were considered positive, and the percentage of immunoreactive nuclei was calculated by counting at least 1,000 tumor cells within the selected area at a magnification of x1,000.

Vessel Staining
After protease-induced epitope retrieval, the vessels were highlighted by staining of the endothelial cells for 1 hour (room temperature) for factor VIII–related antigen (diluted 1/800 in PBS; polyclonal antibody, code no. A-082; DAKO). The tumors were often heterogeneous with respect to the concentration of microvessels, and the sections were examined at low magnifications (x40 and x100) to identify the most vascular area of the tumor (hot spots) according to the method of Weidner et al.26,28 Within these areas, a minimum of 10 fields at x250 magnification (0.4243 mm2/field) were examined. Brown-stained endothelial cells or endothelial-cell clusters that were clearly separate from adjacent vessels, tumor cells, or connective tissue elements were considered single countable microvessels, as previously described by Weidner et al. The average of all fields examined within the hot spot area (mean MVD [MVDmean]) were then recorded and expressed as counts per mm2, as previously described.19

p53 Staining
The sections were subjected to microwave epitope retrieval (750W for two x 5 minutes in citrate buffer, pH 6.0) before incubation for 1 hour (room temperature) with the monoclonal DO-7 antibody (diluted 1/100 in PBS; no. M-7001; DAKO). Nuclear p53 staining was evaluated using a semiquantitative and subjective grading system that considered both the intensity of staining and the proportion of tumor cells in a tissue section showing an unequivocal positive reaction, as described previously.15,19 Staining intensity was graded from zero (no staining) to 3 (strong staining), and percentage of nuclear staining area was graded as zero (no tumor cells positive), 1 (positive staining in < 10% of the tumor cells), 2 (positive staining in 10% to 50% of the tumor cells), or 3 (positive staining in > 50% of the tumor cells). A staining index was calculated as the product of staining intensity and staining area. Because previous studies have shown that the increased expression of p53 protein is most often a dichotomous, and not a continuous, process7,8 indicating whether the gene is altered or not, the tumors were categorized as high-expressing (staining index > 4) or low-expressing (staining index <= 4).

p21 Staining
The sections were subjected to microwave epitope retrieval (750W for 7.5 minutes and 500W for four x 5 minutes in citrate buffer, pH 6.0) before incubation for 3 hours at room temperature and overnight at 4°C with the monoclonal WAF1 antibody EA10 (diluted 1/50 in PBS; Calcibiochem, Cambridge, MA). The sections were examined at low magnifications (x40 and x100) to identify the area with the most intense staining (hot spot). Only cells with distinct nuclear p21 staining were considered positive, and the percentage of immunoreactive nuclei was calculated by counting at least 1,000 tumor cells at x1,000 within the hot spot area. On the basis of previous studies,32 we especially wanted to see whether clearly reduced or lack of expression was of prognostic importance. In the final analysis, we therefore compared the group with the lowest p21 values, defined by the lower quartile, with the rest of the subjects.

In a pilot study of 115 cases (56 belong to the present series), p21 positivity was evaluated by applying the staining index as used for p53 expression. Furthermore, 1,000 tumor cells were randomly counted within the area of the tumor that had the predominant staining pattern. These additional methods were used because p21 has been less studied, and no standard method of registration has been established. However, in terms of identifying the high-risk patients, these two methods were less efficient than the hot spot method, which was used in further studies.

Follow-Up
At the closing date of the study, June 30, 1996, the survivors had been observed for an average of 11.5 years (range, 8 to 15 years), with no patient lost because of insufficient follow-up data. Median observation time for all patients was 9 years. Information about survival was obtained from medical records and correspondence with the primary physician. These data were cross-checked with the population-based Cancer Registry of Norway, which is matched against the Register of Deaths of Statistics Norway.

Statistics
A comparison of groups was performed using the Mann-Whitney test (MW), Kruskal-Wallis test (KW), or the Pearson {chi}2 test. Univariate survival analyses of time to death caused by endometrial carcinoma (cause-specific death) were performed using the product-limit method (Kaplan-Meier), with time of primary operation as entry date. Patients who died of other causes were censored at the date of death. The log-rank test (Mantel-Cox) was used to compare the survival curves for different groups of patients. For the variables age, FIGO stage, histologic grade, microvessel density, and Ki-67 expression, where a linear relation between the categories was assumed, a trend test was used. The variables with an impact on survival in univariate analyses with P <= .10 were further examined by a log-minus-log plot to determine how these variables should be incorporated into the multivariate Cox proportional hazards regression model (backward likelihood ratio elimination procedure).37 In the multivariate study, categories that had a similar prognosis in univariate analysis, or were too small to be analyzed separately, were merged. Because of few events in FIGO stage IA/IB and IC, and a small group of FIGO stage II patients, FIGO stages I and II (FIGO I/II) were merged in the multivariate analysis and compared with stages III and IV (FIGO III/IV). This was considered justified because these two groups would represent different clinical situations: tumor confined to the uterus (FIGO I/II) versus tumor with metastases outside of the uterus (FIGO III/IV). Unadjusted and adjusted hazards ratios were estimated as a measure of effect. Tests for interactions were carried out for the variables with an independent impact on survival in the multivariate analysis. Data were analyzed using the SPSS software package (SPSS, Inc, Chicago, IL).38


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The distribution of traditional clinicopathologic variables, such as patient age, FIGO stage, histologic type, and histologic grade, is shown in Tables 1 and 2. The median age at time of the primary operation was 65 years (range, 37 to 91 years). FIGO stage was significantly correlated with age, histologic type, and histologic grade (Table 1). In univariate survival analysis, age, FIGO stage, depth of myometrial infiltration, histologic type, and histologic grade had a significant impact on survival (Table 2). Postoperative adjuvant treatment did not significantly influence survival rate in univariate analysis, with a 5-year survival rate of 78.2% for patients receiving postoperative radiation therapy (n = 85) compared with 70.7% for the patients who did not get this adjuvant treatment (n = 57; P = .33).


View this table:
[in this window]
[in a new window]
  		Table 1. FIGO Stage Related to Age, Histologic Type, Histologic Grade, Ki-67 Expression, MVDmean, and p53 and p21 Protein Expression
 

View this table:
[in this window]
[in a new window]
  		Table 2. Standard Clinicopathologic Variables, Nuclear Ki-67 Expression MVDmean, and p53 and p21 Expression Related to Survival (Kaplan-Meier Estimates)
 

Ki-67 Expression
Median nuclear Ki-67 positivity was 23% (range, 2% to 86%) within the hot spot area. The Ki-67 expression was significantly higher among patients with FIGO stage III/IV disease (P = .0004), clear-cell or serous papillary histologic types (P = .01 by KW test), and poor histologic grade (P = .0008 by KW test). There was no significant association with patient age (divided into two groups by median value, MW test). On the basis of the counts for nuclear Ki-67 expression, the patients were divided into quartiles. The survival was significantly related to Ki-67 expression (P = .001), with a decreased survival for the category with highest values for Ki-67 expression (> 35%), whereas there was no significant difference between the three other categories (Table 2). In later analyses, the patients were divided into two groups, low Ki-67 (<= 35%) and high Ki-67 (> 35%), by the upper quartile (Fig 1).



View larger version (24K):
[in this window]
[in a new window]
  		Fig 1. Estimated survival among patients with endometrial carcinoma according to nuclear Ki-67 expression, MVD, p53 expression, and p21 expression (Kaplan-Meier method, with death caused by endometrial carcinoma as the end point).

 

MVDmean
Within the most vascular area of the tumors, median MVDmean was 67 microvessels/mm2 (range, 22 to 233 microvessels/mm2). MVDmean was significantly higher for patients with FIGO stage III/IV disease (P = .007) and in tumors with poor histologic grade (P = .03 by KW test). There was no significant association between MVD and patient age (divided into two groups by median value, MW test), histologic type (KW test), or Ki-67 expression (divided into two groups by the upper quartile, MW test). On the basis of the counts for MVD, the patients were divided into quartiles. Survival was significantly lower in the categories with highest values (> 90 microvessels/mm2) and intermediate values (67 to 90 microvessels/mm2), whereas there was no difference between the two lowest categories (Table 2). In later analyses, the patients were divided into three groups, high MVD (> 90 microvessels/mm2), intermediate MVD (67 to 90 microvessels/mm2) and low MVD (<= 66 microvessels/mm2), as illustrated in Fig 1.

p53 Expression
Twenty-six cases (18%) had a high expression of p53 (staining index > 4), whereas 16 cases (11%) were completely negative for p53. The expression of p53 (staining index <= 4 v staining index > 4) was significantly higher for patients with FIGO stage III/IV disease (P = .006), clear-cell or serous papillary histologic types (P = .04 by {chi}2 test), and poor histologic grade (P = .01 by {chi}2 test). With high p53 expression (staining index > 4), there was a tendency toward higher expression of Ki-67, although this association did not reach statistical significance (P = .07 by MW test); no significant association was found with age or MVD (MW tests). High nuclear expression of p53 (staining index > 4) was associated with poor survival in univariate survival analyses (P = .0006; Fig 1 and Table 2).

p21 Expression
Median nuclear p21 expression was 21% (range, 0% to 87%) within the selected area. Only one case was completely negative. Low p21 expression (low v high p21 expression, divided by the lower quartile) was significantly associated with higher age (P = .001 by MW test), clear-cell or serous papillary histologic types (P = .02 by {chi}2 test), poor histologic grade (P = .01 by {chi}2 test), and high p53 expression (P = .05 by {chi}2 test). The association with FIGO stage III/IV was of borderline significance (P = .08), whereas no significant associations were found with Ki-67 or MVD (MW tests). On the basis of the counts for nuclear p21 expression, the patients were divided into quartiles. Survival was significantly reduced in the category with lowest values (<= 9%) for p21 expression compared with the rest (P = .05; Table 2). In later analyses, the patients were divided into two groups, high p21 expression (> 9%) versus low p21 expression (<= 9%), by the lower quartile (Fig 1).

Cox Regression Analysis
Age at primary operation, FIGO stage, histologic type, histologic grade, MVD, and Ki-67, p53, and p21 expression were included in the Cox regression model (backward likelihood ratio elimination procedure). Postoperative radiation therapy did not influence survival significantly in the univariate analysis, and was therefore not included in the multivariate analysis. Age, FIGO stage, MVD, Ki-67 expression, and p53 expression had a significant prognostic influence, whereas histologic type, histologic grade, and p21 expression had no independent prognostic impact (Table 3). No statistically significant interactions were found between the variables with significant impact on survival in the multivariate analysis. The Kaplan-Meier survival curves according to a combination of the three significant tumor markers (MVDmean, Ki-67 expression, and p53 expression) are shown separately for FIGO stage I/II versus FIGO stage III/IV (Fig 2). The tumors with low proliferation (Ki-67 <= 35%), low expression of p53 (index <= 4), and low MVD (MVDmean <= 90 counts/mm2), ie, no positive marker, had a significantly better prognosis than did the tumors with either high proliferation (Ki-67 > 35%), high expression of p53 (index > 4), or high MVD (MVDmean > 90 counts/mm2), ie, one positive marker. The worst prognosis was seen for the patients with more than one positive marker.


View this table:
[in this window]
[in a new window]
  		Table 3. Unadjusted and Adjusted Hazards Ratios Based on the Cox Proportional Hazards Regression Model for Clinicopathologic Variables, Nuclear Ki-67 Expression, MVDmean, and nuclear p53 and p21 Expression
 


View larger version (16K):
[in this window]
[in a new window]
  		Fig 2. Estimated survival for patients with FIGO stage I/II versus FIGO stage III/IV disease, according to combinations of significant tumor markers: Ki-67 expression, MVD, and p53 expression. The survival was significantly better for patients with no positive marker compared with one or more positive markers.

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The prognostic impact of age, FIGO stage, histologic type, and histologic grade is well established1,3,4,39,40 and was clearly confirmed in this population-based study of endometrial carcinoma patients. However, in our study–which included tumor biomarkers related to proliferation, cell-cycle regulation, and angiogenesis–age and FIGO stage, but not histologic type and grade, had independent prognostic significance. The fact that significant associations were present between the traditional variables and most of the tumor biomarkers may explain this finding. To our knowledge, this is the first population-based study on endometrial carcinomas with information about such biomarkers.

Postoperative adjuvant radiation therapy may influence the survival for endometrial carcinoma patients.6 The lack of significant prognostic impact of radiation therapy in our series may be explained by the fact that this is not a randomized study, and the patients were assigned to postoperative radiation therapy according to disease stage. It is therefore difficult to draw conclusions about the potential survival effect of this treatment in our group of patients.

In our study, nuclear Ki-67 expression was an independent prognostic factor. Measures of cell proliferation by S phase estimation have previously been found to have a prognostic significance among endometrial carcinoma patients in a convincing number of studies.14,41-46 The major disadvantage of the flow cytometric method is that S phase estimation is difficult to perform on paraffin-embedded tumor tissue in retrospective studies.45,47,48 Furthermore, tissue homogenates contain variable amounts of nonneoplastic and nonepithelial cells that may dilute the findings in subgroups of all cells examined.49 S phase estimation on fresh tumor tissue was not available for all of the cases in the present study. However, we have previously shown that Ki-67 expression in the hot spot area is superior to S phase fraction in terms of identifying the high-risk patients in a prospective study of 115 endometrial carcinoma patients, comparing S phase measurements on fresh tumor tissue with Ki-67 expression in formalin-fixed and paraffin-embedded tumor tissue.50 In contrast, one study of endometrial carcinomas, by Nordström et al,14 showed a prognostic significance of Ki-67 in univariate survival analysis but not in multivariate analysis, where S phase fraction was identified to be the strongest prognostic factor. That study, however, was hospital-based, and the Ki-67 estimates were performed on paraffin-embedded curettage material collected before preoperative radiation therapy, in contrast to our population-based study with more extensive sampling, and sections selected from the most cellular area of the tumor with lowest architectural differentiation. In contrast to the hot spot method applied in our study, Nordström et al estimated the overall average percentage of positive nuclear Ki-67 staining by a semiquantitative scoring system, making the two studies more difficult to compare. The hot spot method, which is similar to the strategy used by Weidner et al,11 might be more sensitive in detecting small aggressive clones of significant prognostic importance. Our finding, that Ki-67 expression was a significant prognostic indicator for endometrial carcinoma patients, is in line with the recent study of Gassel et al,51 in which the immunohistochemical determination of proliferative activity was found to have an independent prognostic influence. In that study, the proportion of positive cells within the area with the most intense staining was used, and this is equivalent to the hot spot method applied in our present study. Ki-67 has also been found to be a significant prognostic indicator of recurrence in a small study using frozen sections and image analysis.52 In another study using fresh tissue and a semiquantitative technique,53 Ki-67 expression was found to be associated with histologic grade, mitotic count, and previous estrogen therapy but not with the stage of disease, in contrast to our findings, where Ki-67 was highly associated with tumor stage.

We have previously reported an independent prognostic impact of MVD in a smaller study including 60 endometrial carcinoma patients.19 This finding was recently also reported by Kaku et al.54 In our present series, the importance of angiogenesis was confirmed in a larger, population-based setting, including additional molecular markers for cell proliferation (Ki-67) and cell-cycle regulation (p53 and p21). Thus, both tumor cell proliferation (Ki-67) and angiogenesis seem to be important for the development of aggressive tumors and seem to carry an independent significance for the tumor growth potential. The preliminary use of angiogenesis inhibitors25,55 may further motivate the quantitation of MVD for decisions regarding therapeutic strategies.

Mutations of the p53 gene are common in endometrial carcinomas,29,56,57 especially in the serous papillary subtype.58,59 The prognostic influence of increased nuclear p53 protein expression has previously been demonstrated in several studies,12,19,43,57,60-66 some of which could not identify any independent prognostic impact because of strong associations with traditional clinicopathologic factors or DNA ploidy.43,57,61,62 In contrast, other studies have reported an independent prognostic significance of p53 expression,12,60,63,64 even when ploidy results are included in the multivariate analysis.65,66 In the present study, p53 expression influenced prognosis significantly in the Cox regression analysis, adjusted for traditional clinicopathologic variables, Ki-67 expression, and MVD.

Black women have previously been found to have a poorer survival rate compared with white women.67 Serous papillary and clear-cell histologic types, as well as an overexpression of the p53 tumor suppressor gene, have been found to be more frequent among black women, which may contribute to the ethnic disparity.68 These studies also suggest that a larger proportion of black women would likely fall into the high-risk group of endometrial carcinoma patients than in our present cohort of white women.

A decreased expression of p21 in endometrial carcinomas compared with normal endometrial tissue was recently found,69 and a study of 75 endometrial carcinoma patients reported a tendency toward reduced survival for those without p21 expression, although it did not reach statistical significance.17 In our present study, we found a significantly decreased survival among patients with low p21 expression in univariate analysis. This difference might be because of the larger number of cases in our study compared with that of Ito et al.17 However, it may also be because of different methods for antigen retrieval and estimation of p21 staining. The number of completely negative cases with the hot spot method used in our study was only one of 142, compared with 28 (37%) of 75 in the Ito study.

Our finding that nuclear Ki-67 staining, MVD, and p53 expression all had an independent prognostic impact, in addition to age and FIGO stage, in a population-based study of endometrial carcinoma patients implies that these biomarkers could be helpful in identifying high-risk patients who might benefit from postoperative adjuvant treatment. This additional prognostic information is derived from the histologic specimens taken routinely, with no separate sampling or handling techniques required, which makes it easy to apply in a regular clinical setting.


   ACKNOWLEDGMENTS
 
Supported by grants no. D96032 and D94070 from The Norwegian Cancer Society.

We thank Grethe Albrektsen, PhD, at the Division for Medical Statistics, University of Bergen, for statistical advice and critical review of the manuscript; Gerd Lillian Hallseth and Bendik Nordanger for excellent technical assistance; and our colleagues for providing details of patients under their care. We also thank the Cancer Registry of Norway for information.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
1. Abeler VM, Kjørstad KE: Endometrial adenocarcinoma in Norway: A study of a total population. Cancer 67:3093-3103, 1991[Medline]

2. Homsley H, Zaino R: Endometrial cancer: Prognostic factors. Semin Oncol 21:71-78, 1994[Medline]

3. Rose PG: Endometrial carcinoma. N Engl J Med 335:640-649, 1996[Free Full Text]

4. Zaino R, Kurman RJ, Diana KL, et al: Pathologic models to predict outcome for women with endometrial carcinoma. Cancer 77:1115-1121, 1996[Medline]

5. Salvesen HB, Akslen LA, Iversen T, et al: Recurrence of endometrial carcinoma and the value of routine follow up. Br J Obstet Gynaecol 104:1302-1307, 1997[Medline]

6. Rotman M, Aziz H, Halpern J, et al: Endometrial carcinoma: Influence of prognostic factors on radiation management. Cancer 71:1471-1479, 1993[Medline]

7. Jacks T, Weinberg A: Cell-cycle control and its watchman. Nature 381:643-644, 1996[Medline]

8. Kinzler KW, Vogelstein B: Life (and death) in a malignant tumour. Nature 379:19-20, 1996[Medline]

9. Weidner N, Folkman J, Pozza F, et al: Tumor angiogenesis: A new significant and independent prognostic indicator in early-stage breast carcinoma. J Natl Cancer Inst 84:1875-1887, 1992[Abstract/Free Full Text]

10. Gasparini G, Weidner N, Bevilacqua P, et al: Tumor microvessel density, p53 expression, tumor size, and peritumoral lymphatic vessel invasion are relevant prognostic markers in node-negative breast carcinoma. J Clin Oncol 12:454-466, 1994[Abstract]

11. Weidner N, Moore DH, Vartanian R: Correlation of Ki-67 antigen expression with mitotic figure index and tumor grade in breast carcinomas using the novel "paraffin"-reactive MIB1 antibody. Hum Pathol 25:337-342, 1994[Medline]

12. Ito K, Watanabe K, Nasim S, et al: Prognostic significance of p53 overexpression in endometrial cancer. Cancer Res 54:4667-4670, 1994[Abstract/Free Full Text]

13. Fontanini G, Bigini D, Vignati S, et al: Microvessel count predicts metastatic disease and survival in non-small cell lung cancer. J Pathol 177:57-63, 1995[Medline]

14. Nordström B, Strang P, Bergström R, et al: A comparison of proliferation markers and their prognostic value for women with endometrial carcinoma. Cancer 78:1942-1951, 1996[Medline]

15. Aas T, Børresen AL, Geisler S, et al: Specific p53 mutations are associated with de novo resistance to doxorubicin in breast cancer patients. Nat Med 2:811-814, 1996[Medline]

16. Barbareschi M, Caffo O, Doglioni C, et al: p21WAF1 immunohistochemical expression in breast carcinoma: Correlations with clinicopathological data, oestrogen receptor status, MIB1 expression, p53 gene and protein alterations and relapse-free survival. Br J Cancer 74:208-215, 1996[Medline]

17. Ito K, Sasano H, Matsunaga G, et al: Correlations between p21 expression and clinicopathological findings, p53 gene and protein alterations, and survival in patients with endometrial carcinoma. J Pathol 183:318-324, 1997[Medline]

18. Dettmar P, Harbeck N, Thomssen C, et al: Prognostic impact of proliferation-associated factors MIB1 (Ki-67) and S-phase in node-negative breast cancer. Br J Cancer 75:1525-1533, 1997[Medline]

19. Salvesen HB, Iversen OE, Akslen LA: Independent prognostic importance of microvessel density in endometrial carcinoma. Br J Cancer 77:1140-1144, 1998[Medline]

20. Scott RJ, Hall PA, Haldane JS, et al: A comparison of immunohistochemical markers of cell proliferation with experimentally determined growth fraction. J Pathol 165:173-178, 1991[Medline]

21. McCormick D, Hall PA: The complexities of proliferating cell nuclear antigen. Histopathology 21:591-594, 1992[Medline]

22. Cattoretti G, Becker MH, Key G, et al: Monoclonal antibodies against recombinant parts of the Ki-67 antigen (MIB 1 and MIB 3) detect proliferating cells in microwave- processed formalin-fixed paraffin sections. J Pathol 168:357-363, 1992[Medline]

23. McCormick D, Chong H, Hobbs C, et al: Detection of the Ki-67 antigen in fixed and wax-embedded sections with the monoclonal antibody MIB1. Histopathology 22:355-360, 1993[Medline]

24. Folkman J: What is the evidence that tumors are angiogenesis dependent? J Natl Cancer Inst 82:4-6, 1989[Free Full Text]

25. Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenetic switch during tumorigenesis. Cell 86:353-364, 1996[Medline]

26. Weidner N: Intratumor microvessel density as a prognostic factor in cancer. Am J Pathol 147:9-19, 1995[Medline]

27. Holmgren L, O'Reilly MS, Folkman J: Dormancy of micrometastases: Balanced proliferation and apoptosis in the presence of angiogenesis suppression. Nat Med 1:149-153, 1995[Medline]

28. Weidner N, Semple JP, Welch WR, et al: Tumor angiogenesis and metastasis-correlation in invasive breast carcinoma. N Engl J Med 324:1-8, 1991[Abstract]

29. Vile RG: p53: A gene for all tumours? BMJ 307:1226-1227, 1993 (editorial)

30. Gartel AL, Serfas MS, Tyner AL: p21-negative regulator of the cell cycle. Proc Soc Exp Biol Med 213:138-149, 1996[Abstract]

31. El-Deiry WS, Tokino T, Velculescu VE, et al: WAF1, a potential mediator of p53 tumor suppression. Cell 75:817-825, 1993[Medline]

32. Bukholm IK, Nesland JM, Kåresen R, Relationship between abnormal p53 protein and failure to express p21 protein in human breast carcinomas. J Pathol 181:140-145, 1997[Medline]

33. Dameron KM, Volpert OV, Tainsky MA, et al: Control of angiogenesis in fibroblasts by p53 regulation of thrombospondin-1. Science 265:1582-1584, 1994[Abstract/Free Full Text]

34. The Cancer Registry of Norway: Cancer in Norway 1992. Oslo, Norway, Institute for Epidemiological Cancer Research, 1995

35. Mikuta JJ: International Federation of Gynecology and Obstetrics staging of endometrial cancer 1988. Cancer 71:1460-1463, 1993[Medline]

36. Scully RE, Bonfiglio TA, Kurman RJ, et al: Histological typing of female genital tract tumours. International histological classification of tumours. Berlin, Germany, World Health Organization, Springer-Verlag, 1994

37. Cox DR: Regression models and life-tables. J R Stat Soc 34:187-220, 1972

38. Norusis MJ: SPSS Advanced Statistics 6.1. Chicago, IL, SPSS Inc, 1994

39. Hendrickson M, Ross J, Eifel PJ, et al: Adenocarcinoma of the endometrium: Analysis of 256 cases with carcinoma limited to the uterine corpus. Gynecol Oncol 13:373-392, 1982[Medline]

40. Salvesen HB, Akslen LA, Albrektsen G, et al: Poorer survival for nulliparous women with endometrial carcinoma. Cancer 82:1328-1333, 1998[Medline]

41. Wagenius G, Bergström R, Strang P, et al: Prognostic significance of flow cytometric and clinical variables in endometrial adenocarcinoma stages I and II. Anticancer Res 12:725-732, 1992[Medline]

42. Kaleli S, Kosebay D, Bese T, et al: A strong prognostic variable in endometrial carcinoma: Flow cytometric S-phase fraction. Cancer 79:944-951, 1997[Medline]

43. Strang P, Nordström B, Nilsson S, et al: Mutant p53 protein as a predictor of survival in endometrial carcinoma. Eur J Cancer 32A:598-602, 1996

44. Stendahl U, Strang P, Wagenius G, et al: Prognostic significance of proliferation in endometrial adenocarcinomas: A multivariate analysis of clinical and flow cytometric variables. Int J Gynecol Pathol 10:271-284, 1991[Medline]

45. Pfisterer J, Kommoss F, Sauerbrei W, et al: Prognostic value of DNA ploidy and S-phase fraction in stage I endometrial carcinoma. Gynecol Oncol 58:149-156, 1994

46. Podratz KC, Wilson TO, Gaffey TA, et al: Deoxyribonucleic acid analysis facilitates the pretreatment identification of high-risk endometrial cancer patients. Am J Obstet Gynecol 168:1206-1215, 1993[Medline]

47. Sorbe B, Risberg B, Thornthwaite J: Nuclear morphometry and DNA flow cytometry as prognostic methods for endometrial carcinoma. Int J Gynecol Cancer 4:94-100, 1994[Medline]

48. Frierson HF Jr: The need for improvement in flow cytometric analysis of ploidy and S-phase fraction. Am J Clin Pathol 95:439-441, 1991[Medline]

49. Thornton JG, Ali S, O'Donovan P, et al: Flow cytometric studies of ploidy and proliferative indices in the Yorkshire trial of adjuvant progestogen treatment of endometrial cancer. Br J Obstet Gynaecol 100:253-261, 1993[Medline]

50. Salvesen HB, Iversen OE, Akslen LA: Identification of high-risk patients by assessment of nuclear Ki-67 expression in a prospective study of endometrial carcinomas. Clin Cancer Res 4:2779-2785, 1998[Abstract]

51. Gassel AM, Backe J, Krebs S, et al: Endometrial carcinoma: Immunohistochemically detected proliferation index is prognosticator of long term outcome. J Clin Pathol 51:25-29, 1998[Abstract]

52. Geisler JP, Wiemann MC, Zhou Z, et al: Proliferation index determined by MIB-1 and recurrence in endometrial cancer. Gynecol Oncol 61:373-377, 1996[Medline]

53. Nielsen AL, Nyholm HC: Proliferative activity as revealed by Ki-67 in uterine adenocarcinoma of endometrioid type: Comparison of tumours from patients with and without previous oestrogen therapy. J Pathol 171:199-205, 1993[Medline]

54. Kaku T, Kamura T, Kinukawa N, et al: Angiogenesis in endometrial carcinoma. Cancer 80:741-747, 1997[Medline]

55. Baillie CT, Winslet MC, Bradley NJ: Tumour vasculature: A potential therapeutic target. Br J Cancer 72:257-267, 1995[Medline]

56. Sherman ME, Bur ME, Kurman RE: p53 in endometrial cancer and its putative precursors: Evidence of diverse pathways of tumorigenesis. Hum Pathol 26:1269-1274, 1995

57. Reinartz JJ, George E, Lindgren BR, et al: Expression of p53, transforming growth factor alpha, epidermal growth factor receptor, and c-erbB-2 in endometrial carcinoma and correlation with survival and known predictors of survival. Hum Pathol 25:1075-1083, 1994[Medline]

58. Tashoiro H, Isacson C, Levine R, et al: p53 gene mutations are common in uterine serous carcinoma and occur early in their pathogenesis. Am J Pathol 150:177-185, 1997[Abstract]

59. Moll UM, Chalas E, Auguste M, et al: Uterine papillary serous carcinoma evolves via a p53 driven pathway. Hum Pathol 27:1295-1300, 1996[Medline]

60. Kohlberger P, Gitsch G, Loesch A, et al: p53 protein overexpression in early stage endometrial cancer. Gynecol Oncol 62:213-217, 1996[Medline]

61. Lukes AS, Kohler MF, Pieper CF, et al: Multivariable analysis of DNA ploidy, p53, and HER-2/neu as prognostic factors in endometrial carcinoma. Cancer 73:2380-2385, 1994[Medline]

62. Inoue M, Okayama A, Fujita M, et al: Clinicopathological characteristics of p53 overexpression in endometrial cancers. Int J Cancer 58:14-19, 1994[Medline]

63. Geisler JP, Wiemann MC, Zhou Z, et al: P53 as a prognostic indicator in endometrial cancer. Gynecol Oncol 61:245-248, 1996[Medline]

64. Soong R, Knowles S, Williams KE, et al: Overexpression of p53 protein is an independent prognostic indicator in human endometrial carcinoma. Br J Cancer 74:562-567, 1996[Medline]

65. Pisani AL, Barbuto DA, Chen D, et al: HER-2/neu, p53, and DNA analyses as prognosticators for survival in endometrial carcinoma. Obstet Gynecol 85:729-734, 1995[Abstract]

66. Hamel NW, Sebo TJ, Wilson TO, et al: Prognostic value of p53 and proliferating cell nuclear antigen expression in endometrial carcinoma. Gynecol Oncol 62:192-198, 1996[Medline]

67. Matthews RP, Hutchinson-Colas J, Maiman M, et al: Papillary serous and clear cell type lead to poor prognosis of endometrial carcinoma in black women. Gynecol Oncol 65:206-212, 1997[Medline]

68. Clifford SL, Kaminetsky CP, Cirisano FD, et al: Racial disparity in overexpression of the p53 tumor suppressor gene in stage I endometrial cancer. Am J Obstet Gynecol 176:229-232, 1997

69. Palazzo JP, Mercer WE, Kovatich AJ, et al: Immunohistochemical localization of p21(WAF1/CIP1) in normal, hyperplastic, and neoplastic uterine tissues. Hum Pathol 28:60-66, 1997[Medline]

Submitted May 11, 1998; accepted December 28, 1998.




This article has been cited by other articles:


Home page
 J. Clin. Pathol.Home page
N Horree, P J van Diest, P van der Groep, D M D S Sie-Go, and A P M Heintz
Progressive derailment of cell cycle regulators in endometrial carcinogenesis
J. Clin. Pathol., January 1, 2008; 61(1): 36 - 42.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
R. A. Burger
Experience With Bevacizumab in the Management of Epithelial Ovarian Cancer
J. Clin. Oncol., July 10, 2007; 25(20): 2902 - 2908.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
I. M. Stefansson, H. B. Salvesen, and L. A. Akslen
Vascular proliferation is important for clinical progress of endometrial cancer.
Cancer Res., March 15, 2006; 66(6): 3303 - 3309.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
Y.-T. Jeon, S. Kang, D.-H. Kang, K.-Y. Yoo, I.-A. Park, Y.-J. Bang, J. W. Kim, N.-H. Park, S.-B. Kang, H.-P. Lee, et al.
Cyclooxygenase-2 and p53 Expressions in Endometrial Cancer
Cancer Epidemiol. Biomarkers Prev., September 1, 2004; 13(9): 1538 - 1542.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
I. M. Stefansson, H. B. Salvesen, and L. A. Akslen
Prognostic Impact of Alterations in P-Cadherin Expression and Related Cell Adhesion Markers in Endometrial Cancer
J. Clin. Oncol., April 1, 2004; 22(7): 1242 - 1252.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
O. Straume, P. O. Chappuis, H. B. Salvesen, O. J. Halvorsen, S. A. Haukaas, J. R. Goffin, L. R. Begin, W. D. Foulkes, and L. A. Akslen
Prognostic Importance of Glomeruloid Microvascular Proliferation Indicates an Aggressive Angiogenic Phenotype in Human Cancers
Cancer Res., December 1, 2002; 62(23): 6808 - 6811.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Pathol.Home page
O. Straume and L. A. Akslen
Importance of Vascular Phenotype by Basic Fibroblast Growth Factor, and Influence of the Angiogenic Factors Basic Fibroblast Growth Factor/Fibroblast Growth Factor Receptor-1 and Ephrin-A1/EphA2 on Melanoma Progression
Am. J. Pathol., March 1, 2002; 160(3): 1009 - 1019.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
I. Stefansson, L. A. Akslen, N. MacDonald, A. Ryan, S. Das, I. J. Jacobs, and H. B. Salvesen
Loss of hMSH2 and hMSH6 Expression Is Frequent in Sporadic Endometrial Carcinomas with Microsatellite Instability: A Population-based Study
Clin. Cancer Res., January 1, 2002; 8(1): 138 - 143.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Pathol.Home page
O. Straume and L. A. Akslen
Expresson of Vascular Endothelial Growth Factor, Its Receptors (FLT-1, KDR) and TSP-1 Related to Microvessel Density and Patient Outcome in Vertical Growth Phase Melanomas
Am. J. Pathol., July 1, 2001; 159(1): 223 - 235.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
H. B. Salvesen, N. MacDonald, A. Ryan, O. E. Iversen, I. J. Jacobs, L. A. Akslen, and S. Das
Methylation of hMLH1 in a Population-based Series of Endometrial Carcinomas
Clin. Cancer Res., September 1, 2000; 6(9): 3607 - 3613.
[Abstract] [Full Text]

Home page
 Cancer Res.Home page
N. D. MacDonald, H. B. Salvesen, A. Ryan, O.-E. Iversen, L. A. Akslen, and I. J. Jacobs
Frequency and Prognostic Impact of Microsatellite Instability in a Large Population-based Study of Endometrial Carcinomas
Cancer Res., March 1, 2000; 60(6): 1750 - 1752.
[Abstract] [Full Text]

Home page
 Clin. Cancer Res.Home page
H. B. Salvesen, S. Das, and L. A. Akslen
Loss of Nuclear p16 Protein Expression Is Not Associated with Promoter Methylation but Defines a Subgroup of Aggressive Endometrial Carcinomas with Poor Prognosis
Clin. Cancer Res., January 1, 2000; 6(1): 153 - 159.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed