Molecular Human Reproduction vol.6 no.10 pp. 867–872, 2000
Expression of the adrenomedullin gene in epithelial ovarian cancer
Kohkichi Hata1,4, Yuji Takebayashi2, Suminori Akiba3, Ritsuto Fujiwaki1, Kohji Iida1, Kentaro Nakayama1,2, Satoru Nakayama1, Manabu Fukumoto2, and Kohji Miyazaki1 1Department
of Obstetrics and Gynecology, Shimane Medical University, Izumo 693–8501, 2Department of Pathology, Institute of Aging and Cancer, Tohoku University, Sendai 980-8575, and 3Department of Public Health, Faculty of Medicine, Kagoshima University, Kagoshima 890-8520, Japan
whom correspondence should be addressed at: Department of Obstetrics and Gynecology, Shimane Medical University, Izumo 693-8501, Japan. E-mail: [email protected]
Adrenomedullin (AM) gene expression was analysed in 60 cases of epithelial ovarian cancer, (29 serous, 14 mucinous, 13 endometrioid, three clear cell, and one undifferentiated) using reverse transcription–polymerase chain reaction (RT–PCR); 10 of the cases were of low malignant potential; 25 were stage I; three were stage II; 27 were stage III; and five were stage IV. The level of AM gene expression was described in terms of the relative yield of the AM gene to the β2-microglobulin gene. AM gene expression ranged from 0.04 to 1.57 (median 0.36). The association between histological grade and AM gene expression was significant (P ⍧ 0.027), however, the association with other clinico–pathological features, i.e. patients’ age at diagnosis, stage of disease, residual tumour mass after initial surgery, and histological subtype were not significant. Survival data were available for all patients and univariate Cox regression analysis showed that the AM gene expression was significantly associated with a poor prognosis (P ⍧ 0.019). Immunohistochemical studies showed that AM was localized in the outer cell membrane or the cytoplasm of the carcinoma cells and in the endothelial cells of the tumour stroma. The AM gene expression level may play a key role in the biology of epithelial ovarian cancer and may define a more aggressive tumour phenotype. Key words: adrenomedullin/epithelial ovarian cancer/gene expression/RT–PCR
Introduction In spite of significant advances in surgery and the use of new, more effective chemotherapeutic regimens, the overall 5-year survival of patients with ovarian cancer is only ~30% (Kosary, 1994). The high mortality rate of ovarian cancer is due predominantly to occult progression of the tumour within the peritoneal cavity with the initial diagnosis usually only being made at an advanced stage (Richardson et al., 1985a,b). Modifications in chemotherapy and/or surgery are unlikely in the near future to improve the dismal prognosis that is associated with this disease (Kacinski, 1992). An improved understanding of the mechanisms regulating the growth of ovarian cancer cells may eventually lead to techniques which facilitate early diagnosis, establish the prognosis, or determine the response to therapy. Eventually, it may even be possible to design effective target therapies which will work by interfering with the biochemical processes which govern the growth of ovarian cancer cells. Adrenomedullin (AM) is a hypotensive peptide initially isolated from pheochromocytoma (Kitamura et al., 1993). AM has been shown to elicit a potent and long-lasting hypotensive effect when injected i.v. in anaesthetized rats (Ishiyama et al., 1993). Endothelial cells produce AM, and patients with hypertension show higher concentrations of plasma AM than normotensive controls, suggesting that AM participates in the regulation of blood pressure and vascular homeostasis © European Society of Human Reproduction and Embryology
(Kitamura et al., 1994). It has been reported that AM is expressed in a number of human tumours of both pulmonary and neural lineage, including small cell lung carcinoma, adenocarcinoma, bronchoalveolar carcinoma, lung carcinoids ganglioblastoma and neuroblastoma (Martinez et al., 1995; Satoh et al., 1995). AM has been shown to induce cAMP production in the target cells (Ishizaka et al., 1994). It has been previously observed that most of the peptide hormones able to induce cAMP synthesis act as modulators of cell growth (Moody et al., 1993; Ishizuka et al., 1994); thus, AM may also function as a growth regulator in tumour proliferation. In the present study, we used reverse transcriptase–polymerase chain reaction (RT–PCR) to examine AM mRNA expression in 60 cases of epithelial ovarian cancer. In addition, AM localization was studied by immunochemistry. The gene expression of this enzyme has been related to clinical and pathological parameters to evaluate further the role of AM in epithelial ovarian cancer.
Materials and methods Patients A total of 60 patients (aged 19–81 years, median 53 years) with histologically confirmed primary epithelial ovarian cancer were studied (Table I). No patient received any therapy before surgery. The patients were staged according to criteria recommended by the
I I I II III III III III III III III III III III IV IV III III III IV III IV III III I III III I III I
Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Serous Mucinous
Histology 1 1 1 2 3 3 4 3 3 3 2 4 4 4 3 4 3 3 2 3 3 4 4 3 3 4 3 3 3 1
Histological gradea 0.11 0.11 0.67 0.12 0.95 0.32 0.30 0.07 0.33 0.35 0.06 0.12 1.18 0.58 0.14 1.49 0.22 0.61 0.10 0.09 0.53 0.42 0.84 0.48 0.04 0.20 0.54 0.49 0.47 0.45
AM gene expression 26 48 48 9 24 14 98 108 21 12 13 2 12 38 60 12 42 43 60 96 8 21 19 14 15 14 110 10 7 108
Follow-up (months) a a a a d a d a d d a a a d a d d a a a a a a a a a d a a a
Dead (d) or alive (a) 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
⫽ low malignant potential; 2 ⫽ well differentiated; 3 ⫽ moderately differentiated; 4 ⫽ poorly differentiated.
28 56 42 54 58 48 58 58 40 66 63 46 57 67 30 62 48 53 49 19 55 74 66 49 43 54 58 73 47 55
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 26 27 28 29 30
Table I. Adrenomedullin (AM) gene expression in each ovarian cancer
61 26 53 46 42 47 58 47 38 74 63 23 81 60 45 51 60 43 70 46 49 37 45 54 38 45 52 40 60 76
Age (years) I I I III III IV I I I I I I I I I I I I II III III III III II I III I I I III
Stage Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Mucinous Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Endometrioid Clear cell Clear cell Clear cell Undifferentiated
Histology 1 1 2 3 2 4 1 4 1 2 1 1 2 3 3 4 2 3 3 4 4 3 3 3 4 3 4 2 3 4
Histological gradea 0.14 0.85 0.09 1.47 0.54 0.59 0.09 0.05 0.23 0.23 0.17 0.25 0.18 0.96 0.63 1.53 0.07 1.44 0.31 0.44 0.75 1.54 0.12 0.11 0.83 0.75 0.38 0.74 0.27 1.57
AM gene expression
18 36 120 17 48 32 60 17 24 12 14 17 60 50 20 43 72 48 4 10 72 60 6 60 25 12 54 36 8 2
a a a d d d a a a a a a a a a d a a a d a a d a a d a a a d
Dead (d) or alive (a)
K.Hata et al.
Adrenomedullin in epithelial ovarian cancer International Federation of Obstetricians and Gynecologists (FIGO, 1987). The staging system defined by FIGO assumes that an adequate staging operation has been performed (Cannistra, 1993). The staging operation included collection of ascites or peritoneal washing from the pelvis, gutters and diaphragms for cytological studies; total abdominal hysterectomy with bilateral salpingoophorectomy; infracolic omentectomy and appendectomy; selective pelvic and paraaortic lymphadenectomy; and debulking of all gross diseases. If an obvious macroscopic tumour was not present, the following were performed: biopsy of any lesion suspected of being a tumour metastasis or any adhesion adjacent to the primary tumour; blind biopsy of bladder peritoneum and cul-de-sac, right and left paracolic gutter and pelvic side walls; biopsy or smear of right hemidiaphragm. Survival data were available for all patients (median 24 months, range 2–120 months). Of these, 55 patients received cisplatin-containing regimens. Four stage I tumours of low malignant potential and one stage I tumour of mucinous cystadenocarcinoma received no further treatment after surgery. Tissue specimen and RNA preparation Fresh surgical specimens from all patients were obtained, and the tissues for investigation were prepared carefully under a dissecting microscope to eliminate inappropriate components. The tissue samples were stored at –80°C for subsequent analysis. Total RNA was isolated from frozen tissue using a commercially available extraction method (Isogen; Nippon Gene Inc, Tokyo, Japan). Normal placental tissues were used as positive control for AM gene expression. RT–PCR RT–PCR for determination of AM gene expression was performed according to the method previously described (Arao et al., 1994; Hata et al., 1999, 2000). Briefly, complementary DNA (cDNA) was prepared by random priming from 500 ng of total RNA using a First-Strand cDNA Synthesis kit (Pharmacia-LKB, Uppsala, Sweden). Primers for the AM gene (GenBank accession number D14874) were: ACATGAAGGGTGCCTCTCG (upstream) and GGTAGATCTGGTGTGCCAGC (downstream); and the PCR product was 162 bp. Primers for the human β2-microglobulin gene (GenBank accession number V00567) were: ACCCCCACTGAAAAAGATGAG (upstream) and ATCTTCAAACCTCCATGATGC (downstream); and the PCR product was 120 bp. PCR was carried out in a Thermal Cycler (Perkin-Elmer Cetus, Northwalk, CT, USA) with a mixture consisting of cDNA derived from 5 ng of RNA, 10 pmol of upstream and downstream primers for the sequences of the AM gene and 5 pmol of primers for the β2-microglobulin gene, 200 µmol of deoxynucleotide triphosphate, and 0.1 IU of Taq DNA polymerase with reaction buffer (Life Technologies, Rockville, MD, USA) in a final volume of 10 µl. The condition for PCR was denaturation at 94°C for 1 min, annealing at 58°C for 1 min, and extension at 72°C for 1 min. For each specimen, 34 cycles of PCR were performed, and the products were separated on 9% polyacrylamide gels. Then bands were visualized by ethidium bromide staining. National Institute of Health analysis software (version 1.61; National Institute of Health, Bethesda, MD, USA) was used to scan the RT–PCR polyacrylamide
gels after photographic documentation. This software measures relative mean density over a fixed grey scale range after correction for background. AM expression was described in terms of the relative yield of the AM gene to that of the β2-microglobulin gene. Histological grading Histologically we classified tumours into four grades including a noninvasive grade as previously described (Arao et al., 1994). The brief concept of grading used in the present study is as follows. A tumour of grade 1 is low malignant potential (LMP), which is non-invasive to surrounding tissues. Tumours of grades 2–4 correspond to well differentiated, moderately differentiated, and poorly differentiated carcinomas respectively. Immunochemical staining Formalin-fixed and paraffin-embedded tissue sections (5 µm) were stained for AM using the DAKO Envision Polymer System, horseradish peroxidase (HRP) method (Dako A/S, Carpinteria, USA). Briefly, tissue sections were dewaxed in xylene, rehydrated in graded alcohol down to water, washed in phosphate-buffered saline (PBS; pH 7.25) for 5 min and quenched in peroxidase blocking reagent for 5 min to remove endogenous peroxidase activity. Subsequently, slides were incubated with the rabbit anti-human AM primary antibody (Peptide Institute Inc, Osaka, Japan) at a 1:20 dilution for 30 min at room temperature. The secondary antibody conjugated with the HRP-labelled polymer was used for 30 min at room temperature following the primary antibody. Finally, bound antibody complexes were stained for 10 minutes with 0.05% diaminobenzidine and 0.01% hydrogen peroxide in 0.05 mol/l Tris–HCl buffer (pH 7.6). The slides were washed between each step with PBS (3⫻10 min). The sections were counter-stained with Methyl Green. A normal placental section was used as positive control. To assess the immunospecificity of each antibody, normal rabbit serum (negative control) was used instead of primary antibodies. Statistical analysis Mann–Whitney U-test and Kruskal–Wallis one-way analysis of variance by ranks were used as appropriate for the evaluation of differences between end-points. Kaplan–Meier curves were compared by means of homogeneity test and the log-rank test. The Cox proportional hazards model was used in survival analysis. Maximum likelihood parameter estimates and likelihood ratio statistics (LRS) in the Cox proportional hazards models were obtained with the use of a statistical package, EPICURE (Preston et al., 1990). All P values presented were twosided; P ⬍ 0.05 was considered to be statistically significant.
Results AM gene expression To determine the number of PCR cycles appropriate for quantification, PCR was performed from 22–40 cycles at increments of two cycles. The expression ratios of AM to β2-microglobulin were reasonably constant from 24–38 cycles (data not shown).
Figure 1. Representative adrenomedullin (AM) gene expression analysed by reverse transcription–polymerase chain reaction (RT–PCR) and compared with β2-microglobulin (β2-MG) expression. Lanes 1 and 2 ⫽ ovarian cancer stage I, lane 3 ⫽ ovarian cancer stage II; lanes 4 and 5 ⫽ ovarian cancer stage III; lane 6 ⫽ ovarian cancer stage IV, lane 7 ⫽ normal placenta.
K.Hata et al.
Therefore, in the subsequent experiments the values at 34 PCR cycles were defined as the expression of target genes. The representative profile of AM gene expression by RT–PCR is shown in Figure 1. In addition, the sequence of PCR products were analysed and they were identical to the sequence of AM gene. AM gene expression and clinico–pathological features The AM gene expression in ovarian cancer tissues is summarized in Table I (median 0.36, range 0.04–1.57). The gene expression values are mean values from at least three independent RT–PCR experiments. The values of AM gene expression were classified according to the patients’ age at diagnosis, stage of disease, residual tumour mass after initial surgery, histological subtype and grade (Table II). Only the histological grade was significantly associated with AM gene expression (P ⫽ 0.027) (Figure 2).
AM gene expression and prognosis In a follow-up study of 41 cases after complete resection of the primary tumours by surgical operation, AM gene expression of the nine cases which experienced recurrence (median 0.44, range 0.06–1.53) was higher than that of 32 cases without recurrence (median 0.24, range 0.04–1.44), however, the difference was not significant. When the patients are divided into three groups according to the level of AM gene expression, we found that the prognosis of the three groups were different (P ⫽ 0.028, LRS test for homogeneity). The prognosis of the patients with medium
Table II. Clinico–pathological features and adrenomedullin (AM) gene expression Clinico–pathological features Age at the time of diagnosis 艋50 (n ⫽ 28) ⬎50 (n ⫽ 32) FIGO stage I–II (n ⫽ 28) III–IV (n ⫽ 32) Residual disease 艋2 cm (n ⫽ 41) ⬎2 cm (n ⫽ 19) Histological subtype serous (n ⫽ 29) mucinous (n ⫽ 14) endometrioid (n ⫽ 13) clear cell (n ⫽ 3) Histological grade low malignant potential (n ⫽ 10) well differentiated (n ⫽ 9) moderately differentiated (n ⫽ 25) poorly differentiated (n ⫽ 16) a
versus c: P ⫽ 0.020;
AM gene expression median (range)
P value 0.651
0.46 (0.04–1.54) 0.33 (0.06–1.57) 0.106 0.24 (0.04–1.53) 0.47 (0.06–1.57) 0.063 0.30 (0.04–1.53) 0.53 (0.09–1.57) 0.097 0.33 0.23 0.75 0.38
(0.04–1.49) (0.05–1.47) (0.07–1.54) (0.27–0.74)
0.20 0.12 0.47 0.58
(0.09–0.85)a (0.06–0.74)b (0.04–1.54) (0.05–1.57)c
versus c: P ⫽ 0.009.
Figure 2. Adrenomedullin (AM) gene expression in relation to histological grades of ovarian cancer: 1 ⫽ low malignant potential; 2 ⫽ well differentiated; 3 ⫽ moderately differentiated; 4 ⫽ poorly differentiated. Horizontal line indicates the median value.
Figure 3. Comparison of survival between three groups; low adrenomedullin (AM) gene expression (range 0.04–0.20, n ⫽ 20), medium AM gene expression (range 0.22–0.54, n ⫽ 20), and high AM gene expression (range 0.58–1.57, n ⫽ 20), according to the Kaplan–Meier method.
Adrenomedullin in epithelial ovarian cancer
and high levels of AM gene expression was significantly worse than that of those with low AM gene expression. The P values obtained from log-rank test were 0.018 and 0.02 respectively (Figure 3). FIGO stage (III–IV) (P ⫽ 0.0001), residual disease (⬎2 cm) (P ⫽ 0.006), histological grade (poorly differentiated) (P ⫽ 0.042), and AM expression (value) (P ⫽ 0.019) were found to be significantly associated with a poor prognosis in univariate Cox regression analysis (Table III). Multivariate Cox regression analysis revealed that FIGO stage (III–IV) is an independent prognostic factor (P ⫽ 0.004) (Table IV). Cellular AM expression To evaluate the localization of AM in tumour tissue, we performed immunochemical staining using the anti-human antibody for AM. AM was localized in the outer cell membrane or the cytoplasm of the carcinoma cells. Moreover the endothelial cells in the tumour stroma were immunoreactive for AM (Figure 4).
Table III. Results of univariate Cox regression analysis Parameters
Age (years) at the time of diagnosis 艋50 (n ⫽ 28) referent ⬎50 (n ⫽ 32) 0.61 FIGO stage I–II (n ⫽ 28) referent III–IV (n ⫽ 32) 15.14 Residual disease 艋2 cm (n ⫽ 41) referent ⬎2 cm (n ⫽ 19) 4.55 Histological subtype serous (n ⫽ 29) referent others (n ⫽ 31) 0.86 Histological grade others (n ⫽ 44) referent poorly (n ⫽ 16) 2.99 AM gene expression value 3.37 AM ⫽ adrenomedullin.
95% confidence interval
Discussion This is the first study on AM gene expression determined by RT– PCR in epithelial ovarian cancers. AM gene expression was significantly associated with histological grade of the tumour. Elevated AM gene expression was significantly correlated with reduced survival when examined by univariate analysis. Multivariate analysis, however, demonstrated that AM gene expression is not an independent prognostic factor among the clinico–pathological parameters studied. Moreover, the AM gene expression level of the cases with recurrence after complete resection was higher than that without recurrence, but no significant difference was noted. It has to be noted that the number of cases was limited. This molecular evaluation indicates that AM gene expression level may define a more aggressive tumour phenotype in epithelial ovarian cancer. One stage I patient (case 46) died 43 months after operation. The AM gene expression was remarkably high at 1.53. The high AM gene expression in this primary ovarian cancer tissue might be an indicator of its poor prognosis, despite the fact that ovarian cancer was at an early clinical stage. Therefore, even in epithelial ovarian cancer presenting with early disease, an extensive surgical treatment does not guarantee a cure. The patient of case 48, whose samples displayed a high
Table IV. Results of multivariate Cox regression analysis Parameters 3.06–273.80
95% confidence interval
FIGO stage I–II (n ⫽ 28) III–IV (n ⫽ 32) Residual disease 艋2 cm (n ⫽ 41) ⬎2 cm (n ⫽ 19) Histological grade others (n ⫽ 44) poorly (n ⫽ 16) AM gene expression value AM ⫽ adrenomedullin.
Figure 4. Immunohistochemical staining for adrenomedullin (AM) in a section of a serous cystadenocarcinoma (original magnification ⫻100). AM was localized in the outer cell membrane or the cytoplasm of the carcinoma cells. In addition, the endothelial cells in the tumour stroma were also immunoreactive for AM.
K.Hata et al.
AM gene expression, with a similar type and histological stage to that of case 46, is alive. Intensive care is necessary in her surveillance for recurrent disease. New molecular variables, e.g. AM gene expression, correlating with the malignant potential of the cancer cells would be clinically valuable in the identification of high-risk patients in early stage ovarian cancer. The localization of AM in the endothelium of blood vessels was expected because of the involvement of this peptide in blood pressure regulation (Ishiyama et al., 1993; Lippton et al., 1994). The expression of AM in ovarian carcinoma cells suggests possible new roles for this molecule in ovarian cancer, although, to confirm this speculation, further physiological studies are needed to determine the relationships between AM expression and the secretion and distribution of the specific receptors in ovarian cancer. AM expression in rat cultured granulosa cells of the ovary have suggested an autocrine function for AM in these cells (Abe et al., 1998). One group (Miller et al., 1996) demonstrated that AM is expressed in a large variety of human tumour cell lines and that it can function as an autocrine growth factor capable of driving a self-perpetuating state in malignant disorders. Responding tumour cell lines were shown to express AM receptors and showed peptide-mediated increases in intracellular cAMP. The growth and progression of tumours is dependent on the process of angiogenesis. This process begins when a pinpoint colony of tumour cells expands to a size where simple diffusion of nutrients (and wastes) is insufficient. New capillaries are elicited, and the tumour then enters a phase in which perfusion becomes the mechanism by which nutrients arrive and metabolic wastes are carried away (Folkman, 1971). Therefore, blood supply is a critical factor for the growth and progression of the tumour. An alternative or synergistic survival method might be the production and secretion of vasodilatory substances such as AM to enhance the availability of nutrient factors to the tumour bed. Recently, it has been shown that AM is a novel growth factor for endothelial cells and is angiogenic in vivo in the chick chorioallantoic membrane assay (Zhao et al., 1998). This study has shown that AM gene expression might be an indicator of the prognosis of epithelial ovarian cancer, although further study is needed to assess how the expression of this gene is associated with an unfavourable prognosis of this carcinoma. The RT–PCR method we used for determination of AM gene expression is convenient because it does not require radioisotopes or relative large amount of tumour tissue. Because even small amount of samples obtained from specimens during the operation are sufficient for the evaluation of AM gene expression, RT–PCR detection of AM gene expression may make it possible to identify epithelial ovarian cancer patients with poor prognoses before chemotherapy. In those patients, it is anticipated that antiAM therapy could be applied in conjunction with the conventional cytotoxic chemotherapy. Indeed, it has been demonstrated that blocking the biological activity of AM by the use of a monoclonal antibody in neoplastic cell lines results in significant suppression of tumour growth (Miller et al., 1996).
References Abe, K., Minegishi, T., Tano, M. et al. (1998) Expression and effect of adrenomedullin on rat granulosa cell. Endocrinology, 139, 5263–5266.
Arao, S., Suwa, H., Mandai, M. et al. (1994) Expression of multidrug resistance gene and localization of P-glycoprotein in human primary ovarian cancer. Cancer Res., 54, 1355–1359. Cannistra, S.A. (1993) Cancer of the ovary. N. Engl. J. Med., 329, 1550–1559. Folkman, J. (1971) Tumor angiogenesis: therapeutic implications. N. Engl. J. Med., 285, 1182–1186. Hata, K., Kamikawa, T., Arao, S. et al. (1999) Expression of the thymidine phosphorylase gene in epithelial ovarian cancer. Br. J. Cancer, 79, 1848–1854. Hata, K., Fujiwaki, R., Nakayama, K. et al. (2000) Expression of TP and TIE2 genes in normal ovary with corpus luteum and in ovarian cancer: correlation with ultrasound-derived peak systolic velocity. Mol. Hum. Reprod., 6, 319– 323. International Federation of Gynecology and Obstetrics (FIGO) (1987) Changes in definitions of clinical staging for carcinoma of the cervix and ovary. Am. J. Obstet. Gynecol., 156, 263–264. Ishiyama, Y., Kitamura, K., Ichiki, Y. et al. (1993) Hemodynamic effects of a novel hypotensive peptide, human adrenomedullin, in rats. Eur. J. Pharmacol., 241, 271–273. Ishizaka, Y., Ishizaka, Y., Tanaka, M. et al. (1994) Adrenomedullin stimulates cyclic AMP formation in rat vascular smooth muscle cells. Biochem. Biophys. Res. Commun., 200, 642–646. Ishizuka, J, Townsend C.M., Bold, R.J. et al. (1994) Effects of gastrin on 3⬘, 5⬘cyclic adenosine monophosphate, intracellular calcium, and phosphatidylinositol hydrolysis in human colon cancer cells. Cancer Res., 54, 2129–2135. Kacinski, B.M. (1992) CSF-1 and its receptor in ovarian and other gynaecological neoplasms. In Sharp et al. (eds), Ovarian Cancer. Vol. 2. Biology, Diagnosis and Management. Chapman & Hall Medical, London, UK, 115 pp. Kitamura, K., Kawamoto, M., Ichiki, Y. et al. (1993) Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma. Biochem. Biophys Res. Commun., 192, 553–560. Kitamura, K., Ichiki, Y., Tanaka, M. et al. (1994) Immunoreactive adrenomedullin in human plasma. FEBS Letts., 341: 288–290. Kosary, C.L. (1994) FIGO stage, 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., 10, 31–46. Lippton, H., Chang, J., Hao, Q. et al. (1994) Adrenomedullin dilates the pulmonary vascular bed in vivo. J. Appl. Physiol., 76, 2154–2156. Martinez, A., Miller, M.J., Unsworth, E.J. et al. (1995) Expression of adrenomedullin in normal human lung and in pulmonary tumours. Endocrinology, 136, 4099–4105. Miller, M.J., Martinez, A., Unsworth, E.J. et al. (1996) Adrenomedullin expression in human tumor cell lines. J. Biol. Chem., 271, 23345–23351. Moody, T.W., Zia, F., Draoui, M. et al. (1993) A vasoactive intestinal peptide antagonist inhibits non-small cell lung cancer growth. Proc. Natl Acad. Sci. USA, 90, 4345–4349. Preston, D.L., Lubin, J.H. and Pierce, D.A. (1990) EPICURE: risk regression and data analysis software. HiroSoft International Corporation, Seattle, USA. Richardson, G.S., Scully, R.E., Nikrui, N. et al. (1985a) Common epithelial cancer of the ovary. Part 1. N. Engl J. Med., 312, 415–424. Richardson, G.S., Scully, R.E., Nikrui, N. et al. (1985b) Common epithelial cancer of the ovary. Part 2. N. Engl. J. Med., 312, 415–424. Satoh, F., Takahashi, K., Murakami, O. et al. (1995) Adrenomedullin in human brain, adrenal glands and tumor tissues of pheochromocytoma, ganglioneuroblastoma and neuroblastoma. J. Clin. Endocrinol. Metab., 80, 1750–1752. Zhao, Y., Hague, S., Manek, S. et al. (1998) PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a non estrogenic mechanism in the human endometrium. Oncogene, 16, 409–415. Received on January 17, 2000; accepted on July 11, 2000