Int J Gynecol Cancer 2007, 17, 897–904
Hypoxia contributes to development of recurrent endometrial carcinoma J.M.A. PIJNENBORG*y, M. WIJNAKKERy, J. HAGELSTEINy, B. DELVOUXy & P.G. GROOTHUISy *Department of Obstetrics & Gynecology, Tweesteden Hospital, Tilburg, The Netherlands; and yResearch Institute GROW, Department of Obstetrics & Gynecology, University Hospital Maastricht, Maastricht, The Netherlands
Abstract.
Pijnenborg JMA, Wijnakker M, Hagelstein J, Delvoux B, Groothuis PG. Hypoxia contributes to development of recurrent endometrial carcinoma. Int J Gynecol Cancer 2007;17:897–904.
Tumor hypoxia can trigger the induction of angiogenesis. High microvessel density (MVD) as well as hypoxia-inducible factor-1a (HIF-1a) have been related to recurrent disease and tumor aggressiveness, respectively. In this study, MVD and hypoxic status were investigated in primary and recurrent endometrial carcinomas. A total of 65 primary tumors of patients with recurrent endometrial carcinoma (n ¼ 40), and without recurrent endometrial carcinoma (n ¼ 25) were studied. Immunohistochemical analysis was performed on paraffin-embedded tumor tissue. MVD was determined by quantitative analysis of CD31/FVIII positive vessels. Tumor hypoxia was estimated by evaluating the expression of the hypoxia-regulated gene HIF-1a and its target gene carbonic anhydrase IX (CA-IX). An additional 23 recurrent tumors were available for determination of MVD and HIF-1a expression. Effects of hypoxia on tumor protein p53 (TP53) expression were evaluated in the endometrial cancer cell lines (ECC-1), Ishikawa (derived from adenocarcinomas), and AN3CA (derived from a lymph node metastasis). MVD, CA-IX, and HIF-1a expression were not significantly different in primary tumors of patients with recurrence compared to the control tumors. The MVD was significantly lower, and HIF-1a expression was significantly higher in recurrent tumors when compared with their primary tumors (paired t test, P , 0.05). HIF-1a expression correlated well with TP53 expression levels in primary tumors, but not in recurrences. TP53 protein levels were highest in AN3CA cells. Hypoxic conditions induced TP53 protein in ECC-1 and Ishikawa, but not AN3CA cells. We conclude that MVD, CA-IX, and HIF-1a expression are not independent prognostic markers for recurrent endometrial carcinoma. The low MVD, increased HIF-1a protein levels, dissociation of hypoxia, and TP53 protein induction in the metastatic tumor cells (AN3CA) support a role for hypoxia in the development of recurrent endometrial carcinoma. KEYWORDS:
CA-IX, endometrial carcinoma, HIF-1a, hypoxia, MVD, recurrence.
Endometrial carcinoma has a good prognosis(1), yet still some patients present with recurrent disease shortly after completion of therapy. Currently, the application of radiotherapy is dependent on tumor stage and grade. Molecular markers could be helpful to select risk criteria for the development of recurrent disease and eventually be used for the application of Address correspondence and reprint requests to: Johanna Maria Antoinetta Pijnenborg, MD, PhD, Department of Obstetrics & Gynecology, Tweesteden Hospital, 5000 LA Tilburg, The Netherlands. Email:
[email protected] doi:10.1111/j.1525-1438.2007.00893.x 2007, Copyright the Authors Journal compilation # 2007, IGCS and ESGO
#
adjuvant radiotherapy or even for the prediction of tumor sensitivity to radiotherapy. Angiogenesis is the formation of new blood vessels by proliferation of new capillaries from preexisting vessels and capillary endothelial cells. Angiogenesis appears to play an important role in the malignant progression of several gynecological tumors(2). In the endometrial tissue, a significant increase in the microvessel count was observed from simple to complex hyperplasia(3). In endometrial carcinoma, microvessel density (MVD) was found to be related to myometrial invasion and tumor grade, and overall high MVD has been associated with poor survival, independent of
898 J.M.A. Pijnenborg et al.
tumor stage(4–6). Induction of angiogenesis is a fundamental event in the process of tumor growth and metastatic dissemination. Several mechanisms have been elucidated, which lead to increased angiogenesis in human tumors including the presence of hypoxia. Tumor hypoxia arises due to the rapid growth of the tumor and is a consequential imbalance in oxygen consumption and supply. Hypoxia can also arise due to irregular and poorly developed tumor vasculature. Tumor hypoxia has been shown to be an independent prognostic indicator of treatment outcome and is known to cause direct resistance to both radiotherapy and chemotherapy. Furthermore, hypoxia leads to the upregulation of several classes of genes, including several that stimulate angiogenesis, through stabilization and activation of the hypoxia-inducible factor-1a (HIF-1a) transcription factor(7). HIF-1a has been reported to be related to angiogenesis in early-stage endometrial carcinoma(8). Moreover, in breast cancer HIF-1a expression is associated with poor prognosis and downregulation of the estrogen receptor(9). The tumor-associated transmembrane-linked carbonic anhydrase IX (CA-IX) contains a hypoxiaresponse-element within its promoter and is a known transcriptional target of HIF-1. CA-IX catalyzes the reversible conversion of carbon dioxide to carbonic acid, and hence regulates microenvironmental pH. CA-IX expression is present in normal human upper gastrointestinal mucosa and gastrointestinal associated structures, but has also been identified in solid tumors(10,11). Since CA-IX is known to be a direct target of HIF-1 and is activated at the same oxygen tension at which HIF-1a and its other downstream target genes are induced, it has also been proposed to act as a surrogate marker of hypoxia and thus is used to estimate the hypoxic subpopulation of the tumor(12). So far, the presence of CA-IX has not been reported in endometrial carcinoma. Previously, we demonstrated that progression of endometrioid endometrial carcinomas was associated with significant upregulation of tumor protein p53 (TP53)(13). Particularly in recurrent tumors, extremely high TP expression was observed. However, TP53 overexpression was not correlated with increased MDM2 and p21 protein, indicating that TP53 is dysfunctional in these tumors. We showed that this is not due to mutations in P53(13), suggesting that TP53 protein breakdown is impaired. The relevance of this observation is demonstrated by the fact that hypoxia-induced apoptosis requires a functional TP53(14). Therefore, a dysfunctional TP53 could result in the positive selection of cells resistant to hypoxia-induced apoptosis. #
The objectives of the present study were to assess the MVD and hypoxic status in primary and recurrent tumors as well as their predictive value for the development of recurrent endometrial carcinoma. In addition, we correlated the HIF-1a expression to the TP53 levels reported earlier and studied the effects of hypoxia on TP53 expression in ECC-1 derived from primary tumors and a lymph node metastasis.
Materials and methods Patients and tissue specimens Patients with recurrent endometrioid endometrial carcinoma were selected from the Dutch national pathology database (PALGA). Forty patients diagnosed with recurrent endometrial carcinoma were included. Control patients being free of recurrence for at least 3 years after diagnosis and treatment were selected and served as a control group (n ¼ 25). Both patient groups underwent initially hysterectomy and bilateral salpingo-oophorectomy. Radiotherapy was applied when myometrial invasion appeared to be more than 50%, and/or when the tumor was poorly differentiated. One of the patients was treated by chemotherapy after primary surgery. From 23 patients presented with recurrent disease, recurrent tumor tissue was available for analysis as well. An independent pathologist revised histopathology of primary and recurrent tumor and the diagnosis was confirmed in all cases. The study was approved by the Medical Ethical Committee of the University Hospital of Maastricht (MEC 02-0095). Immunohistochemical analysis Immunohistochemical staining was performed on formalin-fixed, paraffin-embedded specimens. Hematoxylin and eosin–stained sections were used to select the tumor area, and sections of 5 lm were prepared on Starfrost adhesive slides (Klinipath, Duiven, The Netherlands). Sections were deparaffinized in xylol and rehydrated prior to blockage of endogenous peroxidase activity by incubation in 3% hydrogen peroxide/methanol solution for 20 min. Antigen retrieval was performed by heat treatment for 20 min in 10 mM Tris-EDTA (ethylenediaminetetraacetic acid) buffer (pH 9.0). Sections were left to cool down to room temperature and rinsed in phosphate buffer saline (PBS) (pH 7.2). Incubation with polyclonal and monoclonal antibodies directed to CD31/FVIII (1:100/1:1800; Dako, Copenhagen, Denmark), respectively was performed for 2 h at room temperature
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
MVD and hypoxia in recurrent endometrial carcinoma
diluted in PBS with 0.1% Tween and 0.1% bovine serum albumin (BSA). For the CA-IX antibody, antigen retrieval was not required, however, prior to antibody incubation sections were blocked with 5% BSA for 30 min at room temperature. Subsequently, incubation was performed with a monoclonal antibody directed to CA-IX (1:50; M75, kindly given by Dr Silvia Pastorekova, Centre of Molecular Medicine, Institute of Virology, Slovak Academy of Sciences, Bratislava, Slovak Republic), diluted in PBS with 0.1% Tween and 0.1% BSA, for 45 min at room temperature. For HIF-1a immunostaining, antigen retrieval was performed by heat treatment for 20 min in 10 mM Tris-EDTA buffer (pH 9.0). Sections were blocked in blocking buffer (5% BSA/PBS/0.1% Tween-20) for 20 min at room temperature. A monoclonal antibody against HIF-1a (clone 54; BD Biosciences Pharmingen, Franklin Lakes, NJ) was added to the sections (1:120 in blocking buffer) and incubated overnight at 4°C. Antibody binding was visualized with the avidin– biotin complex immunoperoxidase technique (ChemMate detection kit; DAKO) and 3,39-diaminobenzidine. Sections were counterstained with hematoxylin. Normal endometrial tissue served as positive and negative control (Mouse immunoglobin G [IgG], 1:100) for CD31/FVIII staining, and cervix carcinoma served as positive and negative control (Mouse IgG, 1:50) for CA-IX expression.
Evaluation of immunostaining MVD and CA-IX staining were calculated as the percentage positive expression of CD31/FVIII and CA-IX, respectively, of the whole tumor area. The Leica Qwin digital imaging system with software from Leica (Bensheim, Germany) was used for analysis. Briefly, tumor area is determined; subsequently threshold for positive staining is set, followed by excluding nonspecific stained areas, and finally resulting in the percentage of positive staining. The correlation coefficients calculated for the two observations made by the same observer, and for two observations made by two observers were, respectively: 0.95 and 0.92, determined on nine tested slides with the Leica Qwin digital imaging system. For determination of HIF-1a expression, percentage of stained tumor cells (0%, 0–10%, 10–50%, .50%) and intensity of staining (absent, weak, moderate or strong) were determined (0, 1, 2, or 3 for each variable) for the entire tumor area within the section, and a staining index (SI, ranging 0–9) was calculated by multiplying categorized parameters(15). Determination of immunostaining was performed in a blinded #
899
fashion and independently by two observers (P.G.G. and B.D.). Cell cultures and Western blotting ECC-1 cells (derived from an endometrial adenocarcinoma), Ishikawa cells (derived from a well-differentiated adenocarcinoma), and AN3CA cells (derived from a lymph node metastatsis of endometrial carcinoma) were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, 1% Glutamax, 1% penicillin/streptomycin (Gibco, Breda, The Netherlands) at 37°C and 5% CO2. Cells were exposed to hypoxic conditions (0.2% O2) for a period of 24 h, after which they were collected and lysed in RIPA (radioimmunoprecipitation) buffer (4.38 g/L NaCl, 0.25 % DOC [deoxycholic acid sodium salt], 500 mg/L SDS [sodium dodecyl sulphate (2)], 1 mM EDTA, 25 mM Tris pH 8.0, 5 mL/L IGEPAL [octylphenyl-polyethylene glycol]). Cells were also incubated with etoposide (50 lM; Sigma-Aldrich, MO), a topoisomerase inhibitor, to check for TP53 functionality. A protease inhibitor cocktail (Roche, Mannheim, Germany) was added to prevent proteolysis. The mixtures were stored at 280°C until analysis. Fifty microgram of protein was diluted in sample buffer (0.19 M Tris-HCl pH 6.8, 30 mM dithiothreitol [DTT], 2% SDS, 0.1% bromophenol blue, 30% glycerol, 5% b-mercaptoethanol) and denatured for 5 min at 95°C. Proteins were separated on a 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane (Schleicher and Schuell, Dassel, Germany). The membrane was blocked in 5% fat-free milk in PBS/ 0.05% Tween-20 and incubated with an antibody against TP53 (1:500, clone DO-7; Novocastra Laboratories Ltd, Newcastle upon Tyne, UK) for 2 h, followed by a 1 h incubation with an HRP (horseradish peroxidase)-conjugated goat anti-mouse antibody (1:1000; Santa Cruz Biotechnologies, La Jolla, CA). The antibody buffer used was 5% fat-free milk in PBS/0.05% Tween-20. Proteins were visualized by incubation in a chemiluminescent substrate (SuperSignal West Pico Chemiluminescent Substrate, Pierce, Rockford, IL). Statistical analysis The SPSS software (version 11.5) was used for statistical analysis. A Student’s t test was performed to compare differences in expression between patients with and without recurrence. A paired t test was performed to compare expression of CD31/FVIII and HIF-1a in primary and recurrent tumor.
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
900 J.M.A. Pijnenborg et al.
A Pearson’s correlation coefficient was determined to investigate the correlation between CD31/FVIII, CA-IX, and HIF-1a expression. All statistical tests were two-sided, and a P-value of 0.05 was considered significant.
recurrent tumors, MVD was significantly lower compared to their primary tumors, ie, 1.0% ( 0.2) vs 1.7% ( 0.3), respectively (n ¼ 20, paired t test, P , 0.05).
Results
CA-IX expression was determined in 59 primary tumors of patients with endometrial carcinoma. Representative examples of membranous CA-IX expression are shown in Figure 2, and results of the quantitative analysis are summarized in Table 2. Overall, CA-IX expression was not correlated with tumor grade or myometrial invasion. The mean ( Standard Error) percentage of positive CA-IX expression was 8.5% ( 1.8) in primary tumors of patients with recurrence and 12.5% ( 3.3) in primary tumors of controls.
Patients All patients (n ¼ 65) were postmenopausal at the time of diagnosis. Clinicopathologic parameters of patients in both groups are demonstrated in Table 1. The recurrent tumors available for analysis were locally on the vagina vault (n ¼ 20), the pelvis (n ¼ 1), and distant metastasis (n ¼ 2). Recurrences developed after primary well-differentiated tumors (n ¼ 6), moderately differentiated (n ¼ 12), and poorly differentiated tumors (n ¼ 5). Microvessel density MVD could be determined in 63 primary tumors of patients with endometrial carcinoma. Representative examples of high and low CD31/FVIII expression are demonstrated in Figure 1, and results of the quantitative analysis are summarized in (Table 2). Overall, MVD was not correlated with tumor stage, tumor grade, or myometrial invasion. MVD in primary tumors of patients with recurrence was not significantly different compared to MVD in primary tumors of controls: 1.7% ( 0.2) vs 1.6% ( 0.2), respectively. In Table 1.
Patient and tumor characteristics
Characteristics Age at diagnosis (median and range) Figo stage IA IB IC II III Tumor grade 1 2 3 Postoperative radiotherapy Time to recurrence (median and range) Follow-up (median and range) #
Control patients without recurrence (n ¼ 25)
Patients with recurrence (n ¼ 40)
67 (54–93) years
72 (51–83) years
1 14 10 — —
3 16 11 5 5
10 11 4 10
10 18 12 6
—
14 (2–168) months 35 (9–169) months
42 (34–120) months
CA-IX expression
HIF-1a expression HIF-1a expression could be observed in 73% (44/60) of the primary tumors of patients with endometrial carcinoma. Representative examples of high and low HIF-1a expression are demonstrated in Figure 1E-F, and results of the semiquantitative analysis are summarized in Table 2. Overall, HIF-1a expression was not correlated with tumor stage, tumor grade, or myometrial invasion. The staining index of HIF-1a expression in primary tumors of patients with recurrence was not significantly different compared to HIF1a expression in tumors of control patients: 1.4 ( 0.3) vs 1.4 ( 0.3), respectively. In recurrent tumors, HIF1a expression was significantly higher compared to their primary tumors: 3.8 ( 0.4) vs 1.4 ( 0.3), respectively (n ¼ 23, paired t test, P , 0.05). The expression of HIF-1a was inversely correlated with MVD, but only within the group of primary tumors of control patients (Pearson’s correlation coefficient 0.43, P , 0.05). Expression of MVD, HIF-1a, and CA-IX were compared with TP53 expression (previous report(13)). HIF-1a expression correlated well with TP53 expression in primary tumors (Pearson’s correlation coefficient 0.55, P , 0.01), but not in recurrent tumors.
Hypoxia and TP53 expression In all three cell lines TP53 protein was detectable. TP53 protein levels were highest in the metastasisderived carcinoma cells AN3CA (Fig. 2). Exposure of the adenocarcinoma cell lines ECC-1 and Ishikawa to hypoxia-induced TP53 protein expression. Culturing AN3CA cells under hypoxic conditions did
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
MVD and hypoxia in recurrent endometrial carcinoma
901
Figure 1. Representative photographs of immunostained endometrial carcinomas. CD31/ FVIII, CA-IX, HIF-1a (B, C, F) low expression, (A, D, E) high expression. Bars represent 100 lm.
not enhance TP53 levels. Incubation with the topoisomerase inhibitor etoposide enhanced TP53 protein levels in ECC-1 and Ishikawa cells, but not the AN3CA cells (Fig. 2).
Discussion In the current study, MVD was determined in primary tumors of patients with and without recurrence and expressed as a percentage of the total tumor area. In our series, MVD was not predictive for the development of recurrent endometrial carcinoma contrary to other studies(4–6,16,17). MVD was determined by a quantitative evaluation of the whole tumor area. Most of the reports on MVD use the ‘‘hot spot’’ method(4,16). The rationale of counting microvessels in vascular hot spots is that these areas originated from tumor cells with the highest angiogenic potential. Yet, this method is limited by subjective selection of those hot spots(18). The alternative Chalkley count is a relative area estimate rather than a true vessel count, but it is also based on selection of vascular hot spots. #
Determination is performed by measuring the number of grid points that hit stained microvessels. In a series of breast tumors, the Chalkley count appeared a rapid method of quantifying tumor angiogenesis well correlated with computer image analysis(19). Although the quantitative analysis is less subjective, still manually nonspecific staining and nontumor areas should be erased before analyzing. The threshold for positive staining was set once for all tumor sections. The fact that we could not find a difference in MVD in our series could be explained by the other method of evaluation. Moreover, there was a difference between both patient groups regarding tumor stage and tumor differentiation, which could influence MVD expression. Yet, within those patients with advanced stages and poorly differentiated tumors, MVD was equal compared to control patients. HIF-1a expression could also not discriminate between patients with or without a risk for recurrence. As expected, HIF-1a expression was inversely correlated with MVD, but only in primary tumors of control patients. The absence of this correlation in
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
902 J.M.A. Pijnenborg et al.
Table 2.
Results of immunohistochemistry (MVD, CA-IX, HIF-1a) in primary and recurrent tumors
Patients without recurrence
Patients with recurrence
ID
Tumor stage and grade
RT
MVD (%)
CA-IX (%)
HIF-1a (SI)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
IC, 3 IC, 2 IB, I IB, 2 IB, 2 IB, 1 IB, 1 IA, 2 IB, 2 IB, 2 IB, 1 IC, 2 IC, 2 IB, 1 IB, 3 IC, 1 IC, 2 IC, 2 IC, 2 IC, 3 IB, 1 IC, 3 IB, 1 IB, 1 IB, 1
1 1 2 2 1 2 2 2 2 2 2 1 2 2 1 2 1 1 1 1 2 1 2 2 2
3.7 1.6 0.8 3.8 1.1 0.3 1.2 1.9 1.3 1.6 0.7 1.7 0.8 1.5 0.2 0.8 2.2 2.5 2.8 0.7 1.5 1.3 1.5 1.5 2.4
ND 2.9 52.4 5.5 1.6 8.5 4.4 0.0 2.6 18.0 4.8 6.3 36.5 1.7 28.4 4.2 7.6 0.2 42.5 45.6 4.1 1.0 13.5 1.1 5.8
0 0 0 1.5 1 3 1.5 0 0 ND 4 1.5 2 1 4 1 1 1 1.5 4 2 1 0 1.5 2
Recurrences
ID
Tumor stage and grade
RT
MVD (%)
CA-IX (%)
HIF-1a (SI)
MVD (%)
HIF-1a (SI)
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
IC, 3 IC, 2 IB, 1 IB, 2 IB, 3 IB, 1 IB, 1 IB, 1 IB, 2 IB, 2 IA, 3 IC, 2 IB, 2 IB, 1 IC, 2 IC, 1 IB, 3 IC, 2 IC, 2 IC, 3 IB, 1 IC, 3 IB, 1 IB, 1 IB, 2 IA, 2 III, 2 IIIC, 3 IC, 3 IA, 2 II, 3 IB, 1 IB, 2 IA, 2 III, 2 IIA, 3 IIA, 2 IIA, 2 IIIA, 3 IIB, 3
1 1 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 1 2 1 2 1 2 2 2 2 2 1 2 2 1 2 1 1 1 2 1 1 1 1
6.8 4.3 3.2 3.7 0.6 0.7 1.1 2.9 0.7 2.2 4.4 1.4 4.3 2.8 1.2 0.8 1.4 1.2 1.0 1.0 2.1 0.8 0.9 0.6 1.31 3.15 1.74 0.54 1.33 0.8 0.85 ND ND 1.14 1.23 0.47 0.9 0.6 0.46 0.54
5.9 3.7 39.1 25.4 3.6 1.0 1.7 15.0 15.0 3.2 10.7 26.2 0.0 9.9 38.6 3.8 2.3 3.4 1.6 3.2 0.3 3.5 0.4 6.4 0.0 2.0 ND ND ND 2.1 17.7 ND ND 24.2 4.1 3.7 11.5 3.4 0.6 6.2
ND ND ND 0 0 ND ND ND ND ND 0 1.5 0 1.5 1.5 1 1 1 5 6 1.5 2 ND 1.5 6.25 1.5 1 5 1.5 1 0 0 0 1 2 1 6 1 1 0
ND ND 0.3 ND ND ND ND ND ND ND 0.21 0.9 1.1 0.85 2.7 ND 0.4 0.31 ND ND 1.8 1.3 ND 0.2 0.2 1.5 2.7 ND 1.09 1.9 ND 0.4 0.26 1.3 ND ND 0.7 0.17 ND 0.76
3 ND 5 3 ND ND ND ND ND ND 1.5 7.5 1 3 0 2 ND 2.3 ND ND 3 1.5 ND 4 4 5 4 ND 4 6 ND 6 6 ND ND ND 9 4 ND 3
SI, staining index; RT, radiotherapy; ND, not determined.
Figure 2. Effects of hypoxia and etoposide on TP53 protein expression in endometrial carcinoma cell lines. E, etoposide; H, hypoxia; C, control. #
primary tumors of patients with recurrence and in the recurrent tumors suggests that the physiologic response to hypoxia in these tumors is disturbed. In several tumor cell lines, CA-IX can be induced by hypoxia, a response mediated by HIF-1(7). CA-IX expression has been related to prognosis and survival of various cancer types including head and neck cancer, renal carcinoma, and sarcomas(20–22). The current study is the first to evaluate expression of CA-IX in endometrial carcinomas and derived relapsed tumors. In contrast to head and neck cancer, where high
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
MVD and hypoxia in recurrent endometrial carcinoma
expression of either CA-IX or MVD was associated with increased risk of developing recurrences(23), CAIX expression in primary tumors of patients who developed a recurrence and primary tumors of control patients was not different. Erythropoietin is also a hypoxia-inducible gene, which controls erythropoiesis and protects neurons from hypoxic damage, and stimulates proliferation and migration of endothelial cells and promotes angiogenesis(24). The effects are mediated through the erythropoietin receptor (EpoR), which is overexpressed in endometrial cancers. The pattern of expression was similar to the expression of HIF-1a and overexpression was highly associated with advanced stage disease, loss of estrogen receptor, lymph node metastasis, and adverse clinical outcome(24). The aim of this study was to test whether MVD and hypoxia in stage I tumors are prognosticators of a poor prognosis. The conclusions with regard to the relation between MVD and hypoxia and the stage of tumor development must, therefore, be considered with caution. However, the absence of necrotic areas and the clear reduction in MVD and HIF-1a overexpression in the recurrent tumors, compared to their primary tumors, strongly suggests that these tumors consist of cells selected for hypoxia resistance, which grow so rapidly that the revascularization cannot keep up. In a previous study, we showed that TP53 expression was dramatically increased in recurrent tumors of endometrial carcinomas(13). Overexpression of TP53 is usually an indication of dysfunctional TP53 proteins, and hypoxia-induced apoptosis requires a functional TP53(14). When correlating the HIF-1a expression with the TP53 expression in both primary and recurrent tumors assessed in the same series of samples (Pijnenborg et al.), we observed that TP53 correlated well with HIF1a expression in primary tumors, but not recurrent tumors. The absence of this correlation in recurrent tumors suggests dysfunctional TP53, and explains the aggressive tumor behavior with the failure to induce TP53 mediated apoptosis(13). This confirms that hypoxia may act as a selection pressure for cells with diminished apoptotic potential, which in turn explains the more aggressive character of hypoxic tumors(25,26). The clinical observations could be reproduced in vitro. We confirmed that endogenous TP53 expression was highest in the aggressive AN3CA cell line, which is derived from a lymph node metastasis of endometrial carcinoma. Moreover, hypoxia-induced TP53 expression in endometrial carcinoma cell lines derived from primary tumors (ECC-1, Ishikawa), but not in the AN3CA cell line. This supports the contention that cells which escape the tumor environment have dys#
903
functional TP53, loose cell cycle control, and gain resistance to hypoxia-induced apoptosis. Both hypoxia and angiogenesis have been related to resistance of tumors to radiotherapy and/or chemotherapy(11). The major causes for this are(1) genomic changes leading to genomic instability and heterogeneity and clonal selection of cells with loss of differentiation and apoptosis(2), changes in gene expression leading to changes in cellular metabolism and enzyme activities, or(3) deprivation of oxygen required for the therapeutics to generated their cytotoxicity (Harrison and Blackwell, Oncologist 2006). Our data did not support a relation between the application of radiotherapy and the hypoxic status. This may be due to the limited number of tumors tested. In conclusion, the results show that MVD, CA-IX, and HIF-1a expression are not independent prognostic markers for recurrent endometrial carcinoma. The low MVD, increased HIF-1a protein levels, and lack of association between hypoxia and TP53 support a role for hypoxia in the development of recurrent endometrial carcinoma.
References 1 Rose PG. Endometrial carcinoma. N Engl J Med 1996;335:640–9. 2 Abulafia O, Triest WE, Sherer DM. Angiogenesis in malignancies of the female genital tract. Gynecol Oncol 1999;72:220–31. 3 Abulafia O, Sherer DM. Angiogenesis of the endometrium. Gynecol Oncol 1999;94:148–53. 4 Salvesen HB, Iversen OE, Akslen LA. Prognostic significance of angiogenesis and ki-67, p53, and p21 expression: a population based endometrial carcinoma study. J Clin Oncol 1999;17:1382–90. 5 Ozalp S, Yalcin OT, Acikalin M, Tanir HM, Oner U, Akkoyunlu A. Microvessel density (MVD) as a prognosticator in endometrial carcinoma. Eur J Gynaecol Oncol 2003;24:305–8. 6 Giatromanolaki A, Sivridis E, Koukourakis MI, Georgoulias V, Gatter KC, Harris AL. Intratumoral angiogenesis: a new prognostic indicator for stage I endometrial adenocarcinomas? Oncol Res, 1999; 11:205–12. 7 Wykoff CC, Beasley NJ, Watson PH et al. Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 2000;60: 7075–83. 8 Sivridis E, Giatromanolaki A, Gatter KC, Harris AL, Koukourakis MI. Association of hypoxia-inducible factors 1alpha and 2alpha with activated angiogenic pathways and prognosis in patients with endometrial carcinoma. Cancer 2002;95:1055–63. 9 Kronblad A, Jirstrom K, Ryden L, Nordenskjold B, Landberg G. Hypoxia inducible factor-1alpha is a prognostic marker in premenopausal patients with intermediate to highly differentiated breast cancer but not a predictive marker for tamoxifen response. Int J Cancer 2006;118:2609–16. 10 Pastorekova S, Parkkila S, Parkkila AK et al. Carbonic anhydrase IX, MN/CA IX: analysis of stomach complementary DNA sequence and expression in human and rat alimentary tracts. Gastroenterol 1997;112:398–408. 11 Loncaster JA, Harris AL, Davidson SE et al. Carbonic anhydrase (CA IX) expression, a potential new intrinsic marker of hypoxia: correlations with tumor oxygen measurements and prognosis in locally advanced carcinoma of the cervix. Cancer Res 2001;61:6394–9. 12 Jiang BH, Semenza GL, Bauer C, Marti HH. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am J Physiol 1996;271:C1172–80. 13 Pijnenborg JM, van de Broek L, Dam de Veen GC et al. TP53 overexpression in recurrent endometrial carcinoma. Gynecol Oncol 2006; 100:397–404.
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
904 J.M.A. Pijnenborg et al.
14 Graeber TG, Osmanian C, Jacks T et al. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature 1996;379:88–91. 15 Salvesen HB, MacDonald N, Ryan A et al. Methylation of hMLH1 in a population-based series of endometrial carcinomas. Clin Cancer Res 2000;6:3607–13. 16 Ozuysal S, Bilgin T, Ozan H, Kara HF, Ozturk H, Ercan I. Angiogenesis in endometrial carcinoma: correlation with survival and clinicopathologic risk factors. Gynecol Obstet Invest 2003;55:173–7. 17 Kaku T, Kamura T, Kinukawa N et al. Angiogenesis in endometrial carcinoma. Cancer 1997;80:741–7. 18 Vermeulen PB, Gasparini G, Fox SB et al. Second international consensus on the methodology and criteria of evaluation of angiogenesis quantification in solid human tumours. Eur J Cancer 2002;38: 1564–79. 19 Fox SB, Leek RD, Weekes MP, Whitehouse RM, Gatter KC, Harris AL. Quantitation and prognostic value of breast cancer angiogenesis: comparison of microvessel density, Chalkley count, and computer image analysis. J Pathol 1995;177:275–83. 20 Bui MH, Visapaa H, Seligson D et al. Prognostic value of carbonic anhydrase IX and KI67 as predictors of survival for renal clear cell carcinoma. J Urol 2004;171:2461–6.
#
21 Maseide K, Kandel RA, Bell RS et al. Carbonic anhydrase IX as a marker for poor prognosis in soft tissue sarcoma. Clin Cancer Res 2004;10:4464–71. 22 Beasley NJ, Wykoff CC, Watson PH et al. Carbonic anhydrase IX, an endogenous hypoxia marker, expression in head and neck squamous cell carcinoma and its relationship to hypoxia, necrosis, and microvessel density. Cancer Res 2001;61:5262–7. 23 Koukourakis MI, Giatromanolaki A, Sivridis E et al. Hypoxia-regulated carbonic anhydrase-9 (CA9) relates to poor vascularization and resistance of squamous cell head and neck cancer to chemoradiotherapy. Clin Cancer Res 2001;7:3399–403. 24 Acs G, Xu X, Chu C, Acs P, Verma A. Prognostic significance of erythropoietin expression in human endometrial carcinoma. Cancer 2004;100:2376–86. 25 Weinmann M, Belka C, Guner D et al. Array-based comparative gene expression analysis of tumor cells with increased apoptosis resistance after hypoxic selection. Oncogene 2005;24:5914–22. 26 Hammond EM, Giaccia AJ. The role of p53 in hypoxia-induced apoptosis. Biochem Biophys Res Commun 2005;331:718–25.
Accepted for publication December 5, 2006
2007 IGCS and ESGO, International Journal of Gynecological Cancer 17, 897–904
