Researchers at London?s St. Bartholomew?s Hospital and Royal London Hospital conducted the study in collaboration with scientists at Harvard Medical School and Massachusetts General Hospital. It was published in the April 10, 1999, edition of The Lancet.

Study Methods

About 22,000 women were included in the study. All were at least 45 years of age, postmenopausal, had no history of ovarian cancer, and still had their ovaries. A simple blood test was used to measure the women?s levels of CA-125, a blood protein found at increased levels in about 80 percent of women who have ovarian cancer.

After the initial baseline screening, half the women were randomly selected to receive three more annual tests, while the other half received no further tests but were followed by researchers. If a woman?s CA-125 levels were elevated, she received pelvic ultrasound examinations, and if abnormalities were detected, she was referred to a gynecologist for a surgical biopsy.

Findings

In the group assigned to receive annual screenings, 468 women had elevated CA-125 levels, and 29 were ultimately referred for a surgical biopsy. Of those referred for biopsy, six were diagnosed with ovarian cancer, while the other 23 had false-positive results, or results that suggested ovarian cancer although the women did not have it. Ten more women in the screening group developed ovarian cancer during the eight years they were followed by the researchers. In the control group (the group not receiving annual screening) 20 cases of ovarian cancer were identified.

Women in the screened group lived longer than those in the control group, 72.9 months and 41.8 months, respectively.

The researchers, led by Ian J. Jacobs, MD, of St. Bartholomew?s Hospital, explained, "We knew that the power to identify a survival difference between the screened and control groups would be limited ? the trial was not done with the primary intention of assessing the impact of screening on mortality."

Currently, two large studies are underway to confirm the benefit of regular screening for all women. Researchers expect the first results may be available in three or four years.

Conclusions

The authors said the trial was designed primarily to address issues of feasibility and screening performance and concluded, "These results show that a multimodal approach to ovarian cancer screening in a randomized trial is feasible and justify a larger randomized trial to see whether screening affects mortality."

"This study did indeed show that women in the general population were willing to participate in an ovarian cancer screening study," said Debbie Saslow, PhD, director of breast and cervical cancer for the American Cancer Society (ACS). "It did not show, however, that they should [get screened]."

The study?s results point to the poor accuracy of the screening methods used, Dr. Saslow added. "In other words, the screening tests missed too many existing cancers and falsely detected too many cancers that did not in fact exist," she said.

"In this particular study, for each of the six women who were diagnosed with ovarian cancer as a result of screening, four additional women underwent surgery unnecessarily," she said. "An additional 10 women who were screened developed ovarian cancer within eight years although the test did not detect any cancer. The poor accuracy of CA-125, even in combination with ultrasound, is the primary reason why the American Cancer Society does not recommend screening [with CA-125] for women at average risk."

ACS Screening Recommendations

The ACS recommends women at high risk ? those who have a family history of the disease ? have access to ovarian cancer screening, as well as women who have symptoms that are associated with ovarian cancer, said Dr. Saslow.

"[For high-risk women] a combination of ultrasound, CA-125, and bimanual pelvic examination is currently thought to be the most effective approach. Unfortunately, no existing test or combination of tests is conclusive. A diagnosis must be confirmed by surgery," she said.

The ACS estimates ovarian cancer will be detected in 25,200 women in the US this year, and there will be 14,500 deaths from the disease.

• Ovarian Cancer
• Women and Cancer (book)