Epithelial ovarian cancer accounts for approximately 90% of all cases of ovarian cancer. The tumors are derived from the coelomic epithelium.
Ovarian cancer is the fifth most common malignancy in American women. In the United States, it is the leading cause of death from gynecologic cancer. Ovarian cancer is the seventh most common malignancy worldwide.
Incidence and mortality
Ovarian cancer currently has the highest fatality to case ratio of all the gynecologic malignancies. In 2000, approximately 23,100 new cases of ovarian cancer were diagnosed, and 14,000 women died from their disease. For women in the United States, lifetime risk of developing the disease is approximately 1 in 70, or 1.4%. The occurrence of epithelial ovarian cancer is unusual in women younger than 40 years and peaks in women aged 60–64 years. Although the overall incidence of ovarian carcinoma decreases in women older than 64 years, the age-adjusted incidence continues to rise. The incidence of ovarian cancer shows wide geographic variation, with the highest incidence rates (11.5–15.3 per 100,000 women) in Scandinavian countries, Israel, and North America, and the lowest rates (3.3–7.8 per 100,000 women) in developing countries and Japan. The incidence of ovarian cancer has remained stable over the last three decades in high-risk (developed) countries, whereas increasing incidence has been reported in low-risk (developing) countries.
The causes of ovarian cancer are poorly understood, but several factors have been associated with an increased or decreased risk of the disease.
Individual risk factors
Age over 40 years, white race, nulliparity, infertility, history of endometrial or breast cancer, and family history of ovarian cancer consistently have been found to increase the risk of invasive epithelial cancer. Higher parity, use of oral contraceptive pills (OCPs), history of breast feeding, tubal ligation, and hysterectomy have been associated with a decreased risk of ovarian cancer. The relationship of a number of additional factors to the risk of ovarian cancer has not been well elucidated; these factors include age at menarche, age at menopause, use of fertility drugs, use of estrogen replacement therapy, talc use, dietary factors, lactose intolerance, and history of mumps and other infectious diseases.
Patients with a family history of ovarian, breast, endometrial, or colon cancer are at increased risk of developing ovarian carcinoma. Although the lifetime risk of developing ovarian cancer in American women is estimated to be 1.4%, the risk increases to 5% in women with one first-degree relative with the disease and further to 7% when two first-degree relatives have ovarian cancer.
Hereditary familial ovarian cancer accounts for approximately 5% of all newly diagnosed cases. Of these women, a large proportion have a mutation in the BRCA1 or BRCA2 gene. A subset of women with two first-degree relatives with ovarian cancer have one of the three distinct autosomal dominant syndromes that have been termed familial ovarian cancer: site-specific ovary, breast-ovary, and Lynch type II syndromes. (Lynch syndrome type II is a hereditary condition characterized by nonpolyposis colorectal cancers and associated endometrial and ovarian cancer.) These syndromes can involve the breast, ovary, colon, and endometrium. For a woman who has two or more affected first-degree relatives and whose family carries familial ovarian cancer, the risk of developing the disease has been estimated to be 40–50%. Hereditary ovarian cancer syndromes are rare, accounting for less than 1% of all reported ovarian cancer cases.
The recent identification of the breast and ovarian cancer susceptibility genes ( BRCA1 and BRCA2) is an important advance in the field of genetic epidemiology. These genes have been linked to familial breast cancer and are linked to the breast-ovary and site-specific ovarian cancer syndromes. Women who are positive for BRCA1 or BRCA2 may have as much as a 50% risk of developing ovarian carcinoma. Furthermore, these women develop the disease at an earlier age than those women without the BRCA genes. Several genes for Lynch II syndrome have also been identified. Screening tests for these genes are available.
Environmental factors may play a role in ovarian cancer. An association between diet and ovarian cancer was suggested on the basis of the geographic variation of ovarian cancer incidence and mortality. Some have suggested that diets high in animal fat increase the risk of ovarian cancer. Decreased risk has been associated with higher total intake of vegetables, vitamin A, and vitamin C. An association between talc use and ovarian cancer has been proposed; the epidemiologic data, however, remain inconclusive.
Reproductive factors play an important role in ovarian cancer risk. Increasing parity decreases the relative risk of developing ovarian cancer. Conversely, nulliparity has been associated with an increased risk of ovarian cancer. The use of OCPs also has been associated with a decreased relative risk. Increasing duration of OCP use has been associated with decreasing risk, and evidence suggests that the protective effect of OCP use persists for 10 or more years after discontinuation of use. Women with a history of breast feeding have been reported to have a lower risk of ovarian cancer than nulliparous women and parous women who have not breast fed. Women with infertility have an elevated risk of ovarian cancer, independent of nulliparity. Although fertility drugs have been implicated in the development of ovarian cancer, their association has not been clearly separated from the risk that nulliparity and infertility confer. Tubal ligation or hysterectomy with ovarian preservation both lower the risk of ovarian cancer. 3.
The etiology of ovarian cancer is multifactorial. Three main hypotheses have been proposed to explain the pathogenesis: the incessant ovulation, gonadotropin, and pelvic contamination theories. The first two theories have received the most support from epidemiologic data. The possibility that a combination of these hypothetical causes occurs during the complex process of carcinogenesis has also been suggested, and future results from studies should help to clarify the issue.
The incessant ovulation hypothesis
This hypothesis postulates that repeated minor trauma to the epithelial surface of the ovary, caused by uninterrupted ovulatory cycles, increases the likelihood of ovarian cancer. The protective effects of ovulation suppression by parity, OCP use, and lactation support this hypothesis. The differences in protective effect provided by pregnancy and OCP use, however, undermine this hypothesis.
The gonadotropin hypothesis
This hypothesis postulates that exposure of the ovary to continuously high circulating levels of pituitary gonadotropin increases the risk of malignancy. This theory is supported by the protective effects of parity and OCP use and the increased risk observed with fertility drugs. This theory is undermined by the fact that premenopausal women develop ovarian carcinoma. In addition, the protective effect of lactation, which increases follicle-stimulating hormone, is inconsistent with the gonadotropin hypothesis. The lack of association of ovarian cancer with late age at menopause also is inconsistent with both the incessant ovulation hypothesis and the gonadotropin hypothesis.
The pelvic contamination theory
This theory suggests that carcinogens may come into contact with the ovary after passing through the genital tract. Although the protection afforded by tubal ligation and hysterectomy supports this hypothesis, the significant role of other reproductive factors is not explained by the pelvic contamination theory.
Detection and prevention
Criteria for effective screening tools
Early ovarian cancer is a silent and often asymptomatic disease. An effective screening test must have sufficient specificity, sensitivity, and positive predictive value for the disease it is used to detect. To be used to screen large populations of women, the screening test must also be cost effective. A sufficiently high positive predictive value is difficult to achieve for a disease with as low an incidence as ovarian cancer (compared with breast cancer, which occurs in 1 of 9 women, ovarian cancer occurs in 1 of 70). Currently, no available screening test has sufficient positive predictive value for early-stage ovarian cancer.
Routine yearly pelvic examination
Routine yearly pelvic examination is currently in use for the general population as a screening tool. The rectovaginal pelvic examination, however, does not have sufficient sensitivity to diagnose early disease.
Cancer antigen 125 (CA-125)
CA 125, first described in the 1980s, is an antigen expressed by 80% of nonmucinous epithelial ovarian cancers. A level higher than 35 U/mL is considered abnormal. In premenopausal women, however, CA-125 levels may also be elevated in a number of benign conditions, including pelvic inflammatory disease, endometriosis, pregnancy, and hemorrhagic ovarian cysts. In addition, approximately 50% of women with early ovarian cancer have a normal CA-125 level. Other markers for ovarian cancer, such as carbohydrate antigen 19-1 (CA 19-9), CA-15-3, OVX1, Her-2/neu, and human chorionic gonadotropin (hCG) fragments, have been investigated. However, an appropriately sensitive and specific combination has not been identified.
Transvaginal ultrsound has also been considered as a screening tool, in combination with color Doppler imaging. However, transvaginal ultrasonography has a poor positive predictive value when used to screen the general population. If the use of ultrasonography as a screening tool is limited to high-risk populations (postmenopausal patients or women with a significant family history), then 17 surgeries are required to find 1 case of stage I cancer. Multimodal screening using CA-125 measurement with sonography has been evaluated and found not to be efficacious as a general screening tool.
Current recommendations for screening
These include a comprehensive family history and annual rectovaginal pelvic examination for women with no significant family history of ovarian cancer. Women with two or more first-degree relatives with ovarian cancer have a 3% chance of having a familial ovarian cancer syndrome, which carries a 40% lifetime risk of ovarian cancer; these women should be examined and counseled by a gynecologic oncologist. Women with a familial ovarian cancer syndrome should undergo annual rectovaginal pelvic examination, CA-125 determination, and transvaginal ultrasonography. Most authors agree that, after women at high risk of ovarian cancer complete their childbearing, they should consider a prophylactic oophorectomy. Participation in clinical trials for screening should be encouraged. BRCA1/BRCA2 testing helps to identify patients at highest risk of hereditary ovarian cancer.
If a woman is undergoing pelvic surgery for other reasons, prophylactic removal of the ovaries may be considered to eliminate her risk of future ovarian cancer almost entirely (although the risk of peritoneal cancer still exists after removal of both ovaries). The associated risks of premature menopause, with its potential for bone loss and heart disease, must be weighed against the risk of developing ovarian cancer, as discontinuation of hormone replacement therapy is common.
Because their use has been shown to be protective, a number of authors suggest encouraging the use of OCPs. High-risk groups may be the most appropriate candidates for OCP prophylaxis.
Investigational methods. No chemoprophylactic agents are currently available for ovarian cancer. Research on numerous agents, however, including retinoids, may improve our armamentarium against ovarian cancer.
Pathology of epithelial ovarian cancer
These are thought to arise from the surface epithelium of the ovary, which is closely related to the coelomic epithelium lining the peritoneal cavity. The tumors are classified by cell type and behavior as benign, atypically proliferating, or malignant. Atypically proliferating tumors are associated with a small risk of recurrence, may have invasive implants, and may decrease survival. Malignant tumors recur, metastasize, and decrease survival.
The serous histologic subtype is the most common, accounting for 46% of all epithelial ovarian tumors. The mean age of patients at diagnosis is 56 years. The epithelium resembles the normal lining of the fallopian tube. Seventeen percent of serous tumors are atypically proliferating, and 33% are malignant. Thirty-three percent of atypically proliferating tumors are bilateral, and carcinomas are bilateral in 33–67% of cases.
These are lined by cells that resemble the cells of the endocervical glands. Eighty percent of all ovarian mucinous tumors are benign; malignant mucinous ovarian tumors are quite rare. Less than 5% of mucinous tumors are malignant, and 16% are atypically proliferating tumors. The mean age of patients diagnosed with malignant tumors is 52 years. Mucinous tumors can become quite large, filling the abdominal cavity. Malignant tumors are commonly bilateral, although atypically proliferating mucinous tumors are bilateral in only 8% of cases.
These account for 6–8% of epithelial tumors, and most are malignant. Twenty percent may be atypically proliferating. The mean age of patients diagnosed with malignant tumors is 57 years. A significant proportion of cases have been associated with adenocarcinoma of the endometrium, and 10% of cases are associated with endometriosis.
Clear cell carcinomas
These account for 3% of ovarian cancers. These tumors are the ovarian neoplasms most commonly associated with paraneoplastic syndromes, including hypercalcemia and hyperpyrexia. Histologically, hobnail-shaped cells are characteristic of the clear cell carcinomas. The mean age at diagnosis is 53 years. Bilaterality occurs in 13% of cases. Clear cell carcinomas also have been associated with endometrial cancer and endometriosis.
Malignant Brenner tumors
These are very rare and are defined as a benign Brenner tumor associated with an invasive component of another type of carcinoma.
Pattern of metastasis
Ovarian cancer can spread by direct extension, by exfoliation of cells into the peritoneal cavity (transcoelomic spread), via the bloodstream, or via the lymphatic system. The most common pathway of spread is believed to be transcoelomic. Cells from the tumor are shed into the peritoneal cavity and circulate, following the path of the peritoneal fluid up the paracolic gutters, along the intestinal mesenteries, and up to the diaphragm.