Study Suggests Short Time Interval for Ovarian Cancer Screening

PHILADELPHIA—To be effective, the time interval for ovarian cancer screening should be no more than a year, a screening trial of 22,000 postmenopausal women suggests. In this study, 67 women were identified with ovarian cancer. For 28 of these women, ovarian cancer was detected within 1 year of their last CA 125 blood test.

Overall, the investigators found, based on review of longitudinal CA 125 tests, that ovarian cancer started an average of 1.9 years before it was found clinically. The results were reported at the annual meeting of the American Association for Cancer Research (AACR).

A 2-Year Window

“This means there is about a 2-year window between the development of ovarian cancer and its clinical detection,” said Steven J. Skates, PhD, one of the study investigators. Dr. Skates is assistant professor of biostatistics, Massachusetts General Hospital.

The screening trial was conducted at St. Bartholomew’s Hospital, London, between 1986 and 1995, under the direction of Ian Jacobs, MD.

In the study, all volunteers underwent an initial prevalence CA 125 test, Dr. Skates reported. Then they were randomized to a further three annual screens or a control arm. If CA 125 levels exceeded 30 U/mL, the woman was recalled for a pelvic ultrasound and 3-monthly CA 125 tests for 1 year. If the ultrasound exam was abnormal, a gynecologic consult was arranged.

All of the study volunteers underwent a follow-up on cancer status in December, 1997, through the UK National Cancer Registry, Dr. Skates said.

“Screening for ovarian cancer,” Dr. Skates said, “is an appealing approach to reducing mortality from this disease because most cases are currently detected in late-stage disease, resulting in a very low survival rate. A crucial component of designing trials to test approaches to screening for ovarian cancer is the interval between regular tests.”

Proposed trials, he observed, have intervals ranging from every year, to every 1½ years, to once every 3 years.

“Estimates on the preclinical duration of disease provide vital information for this design issue,” Dr. Skates said. “Now we have an empirical estimate of the preclinical duration of ovarian cancer. The fact that it is a short 2 years means that if we are going to screen for ovarian cancer and detect it at an early stage, screening every year is a requisite—and screening every 2 years will miss many of the early stages of ovarian cancers.”