|
|
| |||||||||||||||||||||||||||||||
Background Inhibin is an ovarian hormone that inhibits the secretion of follicle-stimulating hormone (FSH) by the anterior pituitary gland. Women with granulosa-cell tumors of the ovary have elevated serum inhibin concentrations, but whether the concentrations are increased in women with other ovarian tumors is unknown.
Methods We measured serum inhibin and FSH concentrations before surgery in 212 postmenopausal women with suspected ovarian cancer and after surgery in 210 of them.
Results Eighteen of the 22 women (82 percent) with mucinous carcinomas (mucinous cystadenocarcinomas and mucinous borderline cystic tumors) of the ovary had elevated serum inhibin concentrations, whereas only 9 of the 53 women (17 percent) with serous carcinomas (serous cystadenocarcinomas and serous borderline cystic tumors) had elevated levels. Serum inhibin concentrations were also elevated in 2 of 12 women (17 percent) with clear-cell carcinomas, 4 of 26 women (15 percent) with undifferentiated carcinomas, 3 of 3 women (100 percent) with granulosa-cell tumors, and 5 of 27 women (19 percent) with other ovarian cancers. The serum concentrations of inhibin were increased in 2 of 28 women (7 percent) with nonovarian pelvic cancers and 11 of 41 women (27 percent) with benign ovarian diseases. All women but one with initially elevated serum inhibin concentrations had low values one week after surgery. Serum inhibin concentrations correlated negatively with serum FSH concentrations (P = 0.05) in women with granulosa-cell tumors but not in women with other tumors, suggesting that the inhibin secreted by tumors in the latter group has decreased biologic activity.
Conclusions Serum inhibin concentrations are elevated in most postmenopausal women with mucinous carcinomas of the ovary and in some women with other types of epithelial ovarian tumors. The concentrations fall after tumor removal.
CA-125 is a protein expressed on the cell membrane of normal ovarian tissue and more than 80 percent of nonmucinous epithelial ovarian cancers2. Serum CA-125 concentrations are elevated in some women with ovarian cancer, especially those with serous cancer. Serum CA-125 concentrations may also be elevated in women with nongynecologic cancers, liver disease, acute pancreatitis, and renal failure and in some seemingly normal women3.
Inhibin is a glycoprotein composed of an alpha and a beta subunit that is secreted by the granulosa cells of the ovary4. Its major physiologic action is to inhibit the secretion of follicle-stimulating hormone (FSH) by the anterior pituitary gland. Inhibin is secreted throughout the menstrual cycle and during pregnancy but not in postmenopausal women. We previously reported that women with granulosa-cell tumors of the ovary had elevated serum inhibin concentrations5 and that measurements of this hormone were useful as a marker of persistent or recurrent granulosa-cell tumors and of the response to therapy6.
In this study, we measured serum inhibin concentrations in postmenopausal women with epithelial ovarian cancer. The results indicate that serum inhibin concentrations are increased in women with several types of ovarian cancer, especially mucinous carcinomas, and that serum inhibin measurements may be useful as a tumor marker in some women with ovarian cancer.
Methods
Subjects
We studied 212 postmenopausal women admitted consecutively to all major gynecologic teaching hospitals in Melbourne for surgery for suspected ovarian cancer. They ranged in age from 47 to 90 years. Serum samples were obtained preoperatively from all women and one week after surgery from 210 women (2 died within the first postoperative week), of whom 143 proved to have ovarian cancer. All three women with granulosa-cell tumors were also studied two weeks after surgery, and one was studied three weeks postoperatively. The mucinous carcinomas and serous carcinomas were classified as invasive or borderline, the latter being defined as a tumor of borderline malignant potential or a carcinoma of low malignant potential because it did not have all the histologic features of carcinoma7,8.
The study protocol was approved by the research and ethics committee of all participating hospitals, and all women gave written informed consent for the studies.
Nongynecologic Cancers
As an additional control, serum inhibin was measured in 23 postmenopausal women with nongynecologic cancers. Among them, nine had breast cancer, seven hematologic cancers, six colon cancer, and one esophageal cancer.
Analytic Methods
Serum inhibin concentrations were measured in duplicate by radioimmunoassay with a polyclonal antiserum to 31-kd bovine inhibin and 125I-labeled bovine inhibin as tracer4. A pool of serum from women with ovarian hyperstimulation was used as the assay standard. The assay detects the biologically active inhibin dimer and cross-reacts with nonbiologically active peptides related to the alpha subunit; such peptides may circulate during the luteal phase of the ovarian cycle in premenopausal women. The results were expressed in arbitrary units in which 1 unit equaled the biologic activity of inhibin in 1 mg of ovine testicular lymph protein as measured by a pituitary-cell bioassay. Serum samples collected before and after surgery were usually analyzed in different assays.
The mean (±SD) sensitivity of the assay was 77 ±31 U per liter (range, 49 to 111 in 35 assays), and the within-assay and between-assay coefficients of variation in the region of maximal assay precision were 4.3 percent and 4.3 percent, respectively. In 72 normal postmenopausal women 40 to 67 years of age, all of whom had had amenorrhea for at least six months and serum FSH concentrations exceeding 22 IU per liter, previously established as the lower limit of normal for postmenopausal women in our laboratory, serum inhibin was undetectable in 67 women, between 119 and 122 U per liter in 3 women, and 252 and 409 U per liter in 2 women; the 5 women with detectable levels presumably had some residual ovarian function9. On the basis of these results we arbitrarily defined a serum concentration of more than 122 U per liter in postmenopausal women as being elevated.
Serum CA-125 concentrations were measured by radioimmunoassay kits supplied by Abbott Diagnostics, Sydney, Australia. A serum CA-125 value above 35 U per milliliter was considered abnormal2. Serum FSH concentrations were measured by radioimmunoassay4 with World Health Organization matched reagents, and serum estradiol concentrations were measured with kits supplied by Diagnostic Products, Sydney, Australia10. All these analyses were also done in duplicate.
Statistical Analysis
Differences between groups were evaluated by analysis of variance. Spearman's rank-order correlation coefficient was used to examine the relation between serum inhibin and FSH or estradiol in the various groups of women with ovarian tumors. The preoperative and postoperative serum inhibin and FSH concentrations in the women with granulosa-cell tumors were compared by the paired t-test.
Results
The preoperative serum inhibin concentrations in the 212 postmenopausal women thought to have ovarian cancer are shown in Table 1. The concentrations were elevated in 54 (25 percent) of these women. Among the 143 women who proved to have ovarian cancer, 41 (29 percent) had elevated values. The mean values in women with mucinous carcinomas (cystadenocarcinomas and mucinous borderline cystic tumors) and granulosa-cell tumors were higher than those in women with all other primary cancers (P<0.001). Eight of the nine women (89 percent) with mucinous cystadenocarcinomas had elevated concentrations, as did 10 of the 13 women (77 percent) with mucinous borderline cystic tumors (Figure 1). Seven of the 22 women with mucinous carcinomas had serum CA-125 concentrations that exceeded 35 U per liter. There was no correlation between the serum inhibin and CA-125 concentrations in these women (Spearman rs, 0.22; P>0.30). Among the 50 women with serous cystadenocarcinomas, 9 (18 percent) had elevated serum inhibin concentrations; none of the 3 women with serous borderline cystic tumors had elevated concentrations. All three women with granulosa-cell tumors had elevated serum inhibin concentrations.
|
|
All women with mucinous cystadenocarcinomas and elevated preoperative serum inhibin concentrations who were studied one week after surgery had values below 122 U per liter. The concentrations were also considerably lower one week after surgery in most of the women with the other types of ovarian cancers and those with benign tumors (Figure 1).
Endocrine Correlations
In women with granulosa-cell tumors, serum inhibin concentrations measured at any time during their disease were negatively correlated with serum FSH concentrations (Spearman rs, -0.57; P = 0.05) (Figure 2). In contrast, there was no correlation between the preoperative serum inhibin and FSH concentrations in the women with mucinous carcinomas (cystadenocarcinomas and borderline cystic tumors) (Figure 2).
|
Nongynecologic Cancer
One of the 23 postmenopausal women with nongynecologic cancer (4 percent) had elevated serum inhibin concentrations. This woman, whose serum inhibin concentration was 250 U per liter, had a carcinoma of the colon.
Discussion
We found elevated serum inhibin concentrations in postmenopausal women with a variety of ovarian diseases, especially mucinous carcinomas. The high frequency of elevated values in women with mucinous carcinomas (82 percent) is particularly important because there is no hormonal marker for this type of cancer, which constituted 15 percent of the ovarian cancers in this report. Moreover, the most widely used marker for ovarian cancer, serum CA-125, was less helpful in identifying mucinous borderline cystic tumors than some other types of carcinoma. Among the women with mucinous borderline cystic tumors, 1 of the 9 (11 percent) tested had an elevated serum CA-125 concentration, whereas 10 of the 13 women (77 percent) with this tumor who were tested had elevated serum inhibin concentrations. The finding of elevated serum inhibin concentrations in women with mucinous borderline cystic tumors suggests that the capacity to secrete inhibin is increased early in the process of malignant transformation.
Although other markers have been used to detect ovarian cancer,11 such as CA 15-3, alpha-fetoprotein, ovarian antigen, and cancer-associated antigen, none have the physiologic importance of inhibin. This study extends previous work identifying the value of serum inhibin concentrations in women with granulosa-cell tumors of the ovary. In this group, serum inhibin concentrations were negatively correlated with serum FSH concentrations, but there was no correlation between the two hormones in the women with mucinous or other ovarian carcinomas. These results suggest that the inhibin secreted by granulosa-cell tumors is biologically active, whereas that secreted by mucinous tumors may consist mainly of inhibin peptides related to the alpha subunit that are detected in the immunoassay used but have little or no biologic activity. The precise biochemical forms of inhibin, or its alpha or beta subunits, secreted by ovarian tumors are not known.
Although we have demonstrated that women with ovarian cancers, particularly mucinous cystadenocarcinomas and mucinous borderline cystic tumors, have elevated serum inhibin concentrations, more study will be needed to determine the sensitivity and specificity of serum inhibin measurements as a screening test for ovarian cancer. American authorities have emphasized that screening for preinvasive ovarian cancer by yearly pelvic examination, ultrasonography, and serum CA-125 measurements would add over $13 billion yearly to the U.S. health care bill, with no guarantee of lowering the death rate from ovarian cancer11,12. There is no reason to believe that serum inhibin measurements would be more effective for this purpose. The finding of rapid reductions in serum inhibin concentrations after surgery, however, does suggest that inhibin measurements could be useful as a tumor marker.
Supported in part by the National Health and Medical Research Council of Australia and the Anti-Cancer Council of Victoria.
We are indebted to the consultants who referred patients for this study; to Dr. Philip McCloud, Monash Department of Mathematics, for statistical advice; to Dr. Beatrice Susil, Department of Anatomical Pathology, Monash Medical Centre, for pathological advice; and to Ms. Sue Pankridge for medical artwork.
Source Information
From the Department of Obstetrics and Gynecology, Monash University (D.L.H., P.M., T.J., A.J.D.), Prince Henry's Institute of Medical Research (H.G.B., M.B.), Royal Women's Hospital (M.Q., R.R.), Mercy Hospital for Women (P.G.), and the Peter MacCallum Institute (J.J.C), all in Melbourne, Australia.
Address reprint requests to Professor Healy at Monash Medical Centre, Department of Obstetrics and Gynecology, 246 Clayton Rd., Clayton, VIC 3168, Australia.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
