SGO: Ovarian Cancer Tests Increase Negative Biopsies

SAN DIEGO, March 6 -- Ovarian cancer screening in a major randomized trial is leading to large numbers of biopsies that find a relatively low number of malignancies, researchers said here.

The Prostate, Lung, Colorectal and Ovarian screening trial (PLCO), four years into its evaluation of transvaginal ultrasound and CA-125, has not had encouraging results so far for these ovarian cancer screens, according to Edward Partridge, M.D., of the University of Alabama at Birmingham.

It will take "a couple of more years" before the landmark National Cancer Institute-sponsored study is able to say whether the women randomized to screening will live longer than the women in the control group, Dr. Partridge told a plenary session of the Society of Gynecologic Oncologists meeting here.

But in a preliminary look at the women in the screening arm, Dr. Partridge said researchers have found that the testing "leads to substantial number of surgeries that are perhaps unnecessary." Specifically:

• An abnormal transvaginal ultrasound is more likely to lead to biopsy than an elevated CA-125.
• The proportion of biopsies that actually find cancer is higher when the surgery follows an elevated CA-125.
• Ultrasound is more likely to discover a stage I or II cancer, while CA-125 is more likely to uncover tumors of stage III or IV.

In the study, women were randomized to six annual CA-125 tests, four annual ultrasound tests, or to routine care. All told, Dr. Partridge said, 33,993 women with ovaries were randomized to the screening arm.

Dr. Partridge said so far 95 cancers have been detected in the cohort, 67% of them after one or the other method of screening.

The researchers found that 17% to 41% of positive ultrasound tests led to a biopsy, depending on the year of the study, compared with between 8.9% and 15% for positive CA-125 tests.

It's not clear why that should be so, Dr. Partridge said, although it may be a function of the concreteness of the test. "Transvaginal ultrasound is a visible abnormality on the ovary that you and the patient can see," he said, while elevated CA-125 "is just a number."

On the other hand, 19% to 32% of the biopsies performed after CA-125 testing actually found cancer, he said, compared to less than 5% in any year of the study for ultrasound.

Overall, the cancer yield of the two tests combined ranged from 3.5% in the first year to 9.6% in the most recent year. Most experts think that a yield of 10% is the lowest level that would make a screening test acceptable.

On the other hand, Dr. Partridge said, there is no acceptable level unless a test can be shown to reduce mortality. "Everything else is unimportant," he said.

Most cancers detected by ultrasound -- 77% -- were stage I or II, while 90% of the cancers found by CA-125 testing were stage III or higher, he said.

Although it is too early for the study to give definitive results on mortality, the researchers should be praised for their "expansive effort," said Paul DePriest, M.D., of the University of Kentucky, in a formal discussion of the report.

Among the strengths of the study, he said, are its prospective design and large, multicenter nature. Unfortunately, the study also allows variations in triage and diagnosis that have led to the variation in surgical intervention rates, he said.

He also noted that of the cancers detected following screening, a large proportion - 45 of 63 - were stage III or IV, which raises the question of whether the two tests are useful for early detection of ovarian cancer.