|
Frequent serum CA125 testing in tandem with transvaginal sonography may be an effective screening method to detect ovarian cancer early in women at increased risk of the disease.
When serial CA125 levels were analyzed using the previously established risk of ovarian cancer algorithm (ROCA), the overall specificity was 99.7% and positive predictive value 13% among 2,343 high-risk women in a prospective pilot study.
The results are encouraging because no proven strategy exists for these patients, but larger studies are needed to validate the findings, Steven J. Skates, Ph.D., reported. "Serial CA125 testing essentially establishes a baseline value for each woman that personalizes the interpretation of each new CA125 result to determine if ultrasound is indicated at a particular time," he said in an interview.
The ROCA has previously shown greater positive predictive value and sensitivity than a single CA125 test when screening healthy postmenopausal women for ovarian cancer (J. Clin. Oncol. 2005;23:7919-26).
The current study defined high-risk women as those with a BRCA1 or BRCA2 mutation, or first- or second-degree relatives with a BRCA mutation or multiple breast or ovarian cancers, or women of Ashkenazi heritage and at least one first-degree or two second-degree relatives with breast or ovarian cancer. Women with a prior diagnosis of ovarian cancer were excluded.
Participants underwent CA125 testing every 3 months, and the risk of having ovarian cancer was recalculated after each test based on the CA125 profile. Women with a greater than 1% risk were referred to ultrasound, and those with a greater than 10% risk to a gynecologic oncologist.
Between July 2001 and September 2006, 19,549 CA125 tests were performed, totaling 6,284 woman-years of screening. Compliance was high, and the average number of CA125 tests was three per year. Of 628 referrals made to ultrasound, 414 were performed, resulting in 38 women (9%) undergoing study-indicated surgeries.
Nine ovarian cancers were identified during screening—three were prevalent (one early stage, two late stage) and six were incident (five early stage, one late stage), reported Dr. Skates, ROCA Screening Study Group, Massachusetts General Hospital Cancer Center, Harvard Medical School, in Boston.
Three incident carcinomas were found on prophylactic oophorectomy in early stage. The ROCA detected two of the three remaining incident cases in early stage, and three of three prevalent cases.
The positive predictive value of ROCA was higher at 22% in postmenopausal women, but is acceptable at 13% in high-risk women considering prophylactic oophorectomy after childbearing, Dr. Skates said. He received grant support from Fujirebo Diagnostics Inc., a cosponsor of the study with the National Cancer Institute.
Skates S.J. et al. A prospective multi-center ovarian cancer screening study in women at increased risk. Abstract 5510.
Commentary |
Ovarian carcinoma remains the major killer of women among gynecologic cancers, accounting for more than 60% of all deaths. This study by Skates et al. evaluated the efficacy of serial measurement of CA-125 every 3 months and use of these values to calculate a risk of ovarian carcinoma. Among 38 women referred for surgery, nine ovarian carcinomas were discovered. Five were found in an early stage by the ROCA (Risk of Ovarian Cancer Algorithm), yielding a positive predictive value of 13%. It needs to be confirmed in a study large enough to find 30 cases. This pilot study is too small to draw definitive conclusions about the ROCA. We must be certain a screening test is truly of value in detecting early disease because a false-positive screening test for ovarian carcinoma usually results in a major invasive abdominal procedure. Use of serial CA125 measurements to screen for ovarian carcinoma cannot be recommended for routine clinical use at this time. — J. Tate Thigpen, M.D.
|