Search for Ovarian Cancer Screen Must Look Beyond Ultrasound

August 1, 1995

WASHINGTON--Only a screening test that can reliably find stage I tumors will have any real impact on overall ovarian cancer mortality, and transvaginal ultrasound does not appear to fulfill that requirement. Although the technique can detect stage I ovarian cancers, its specificity is not high enough to make it useful as a general screening test, Beth Y. Karlan, MD, said at the American Cancer Society National Conference on Gynecological Cancers.

WASHINGTON--Only a screening test that can reliably find stageI tumors will have any real impact on overall ovarian cancer mortality,and transvaginal ultrasound does not appear to fulfill that requirement.Although the technique can detect stage I ovarian cancers, itsspecificity is not high enough to make it useful as a generalscreening test, Beth Y. Karlan, MD, said at the American CancerSociety National Conference on Gynecological Cancers.

Dr. Karlan said that proven screening tests such as Pap smearsand mammography work because the biology of the tumors is wellunderstood, and the tumor sites are more accessible than the ovaries.

With ovarian cancer, the molecular events leading up to the developmentof the cancer are not well known, nor is the history of any precursordisease. However, studies of p53 mutations indicate that the diseasebegins with a single cell that then proliferates and metastasizes.

A monoclonal origin for ovarian carcinomas would theoreticallypoint the way to detection of early neoplastic changes in theovary. "We know how it starts, but we still don't know howlong it takes to get from stage I to stage III," said Dr.Karlan, of the department of obstetrics and gynecology, Cedars-SinaiMedical Center, Los Angeles.

"We need to better understand early tumor biology in termsof what allows these cells to transform and proliferate, in orderto find a good screening test that would diagnose early-stagedisease," she added.

Two Screening Studies

Dr. Karlan reported on a University of Kentucky trial of 3,220asymptomatic postmenopausal women given annual transvaginal ultrasoundscreens for 5 years to find early alterations in the ovary. Forty-fourwomen had positive transvaginal ultrasound scans; three provedto be malignant tumors (two were stage I), and 41 were benign.

A screening study at Cedars-Sinai (the Gilda Radner Ovarian CancerDetection Program) used transvaginal ultrasound, color Doppler,and tumor biomarkers, including CA 125, to screen 1,100 women(40% postmenopausal) who were at increased risk because they hadfirst- degree relatives with a history of breast, ovarian, endometrial,or colon cancer.

Initial ultrasound screening showed simple cysts in 20% of thepostmenopausal women, but the majority resolved on follow-up.Thus, Dr. Karlan suggests rescanning patients with this finding6 to 8 weeks after the initial scan before using any invasivediagnostic procedures.

Although 18% of the premenopausal women had CA 125 levels over35 U/mL, most of these values declined on a second screening.Dr. Karlan emphasized that serial CA 125 testing is more usefulthan single screens, in order to look for a rate of rise overtime. For example, a doubling of the CA 125 concentration over3 to 6 months would be significant.

Along with ultrasound, color Doppler flow imaging may be usedto detect angiogenesis, an early event in tumor growth. "Thisis a help in looking at ovarian tumors in postmenopausal women.But drawing lines around wave forms is a subjective science, andexperience is vital for accuracy," Dr. Karlan said.

"We don't want to use a screening test that would offer falsereassurances or false concerns to the public," Dr. Karlansaid. "At the moment, transvaginal ultrasound has not beenproven to be effective as a screening tool."

Future Directions

Rather than trying to enhance the efficacy of older ovarian screeningmodalities, for example, by developing new ways to analyze colorflow Doppler or building improved ultrasound machines, Dr. Karlanbelieves that researchers should focus on developing new modalitiesbased on the molecular and cellular biology of the disease.

One new modality under study, she said, is a molecular test thatuses a technique known as differential hybridization to look fordifferences in DNA between normal epithelial tissue and ovariancancer.