CA-125 SENSITIVITY RATE OF 57.1% IN PREDICTING OVARIAN CANCER
This article was originally published in The Gray Sheet
Executive Summary
CA-125 SENSITIVITY RATE OF 57.1% IN PREDICTING OVARIAN CANCER is not sufficient to warrant its use alone as a screening test for the detection of ovarian cancer, Kathy Helzlsouer, MD, MHS, Johns Hopkins University, et al., conclude in a study published in the March 3 Journal of the American Medical Association. "The available evidence suggests that using CA-125 level alone as a screening test, particularly with a reference value of 35 U/ml, does not have a sufficiently high sensitivity to be recommended for routine screening for the detection of ovarian cancer," the study states. The case-control study was undertaken to evaluate blood serum CA-125 level as a potential screening tool for ovarian cancer. The study utilized blood serum collected over a four month period in 1974 from 11,009 female residents of Washington County, Maryland. Forty "incident ovarian cancer cases," diagnosed from 1975 to 1989, were identified from the group by linking the serum bank records with the Washington County cancer registry. Two controls were matched to each case based on race, age, the time of day the blood sample was collected, and menstrual cycle and menopausal status. Three cases and seven controls were subsequently eliminated because of inadequately matched menstrual cycles, leaving 37 cases and 73 controls available for analysis. Serum CA-125 levels were determined using Centocor's CA-125 radioimmunoassay serum tumor marker kit. The sensitivity and specificity of CA-125 levels as a screening test were then calculated over a range of reference values from 10 U/ml to 35 U/ml. The CA-125 levels were also evaluated with respect to the time interval from serum donation to diagnosis by dividing the 15-year followup period into three-year intervals. The 57.1% sensitivity rate was found using a serum CA-125 level of 35 U/ml in patients diagnosed within three years of their serum donation. Specificity at that reference level and time period was 100%. Both sensitivity and specificity went down as the time from serum donation to diagnosis increased. Over the whole 15-year follow-up period, sensitivity was 24.3% and specificity was 95.9% using the same 35 U/ml reference level. The researchers also found that if the reference value of CA-125 used to indicate a positive result was lowered, sensitivity was increased but specificity decreased. For example, using a 12-year follow-up period, sensitivity increased from 27.6% with a 35 U/ml level to 69% using a 10 U/ml level, while specificity dropped from 94.7% to 61.4%. The researchers believe that "even with a specificity as high as 99%, widespread screening will result in an unacceptable number of false-positive test results and low predictive value since ovarian cancer is a rare disease." A further concern about the potential use of CA-125 as a screening test was the fact that among postmenopausal women, the group at "highest risk" of developing ovarian cancer, sensitivity was lower than among premenopausal women (12.5% and 46.2% respectively over the 15-year period with the 35 U/ml value). Greater specificity shown in the postmenopausal group (97.9% as opposed to 92.3%) "may be explained by the lack of interfering conditions, such as endometriosis in older women, which are associated with elevated CA-125 levels," the study says. Although the researchers conclude that CA-125 is not sufficiently sensitive to use alone as a screen, they point out that "the use of multiple modalities including bimanual vaginal examination, transvaginal ultrasonography, and serum levels of CA-125 for the detection of ovarian cancer may be a means to improve sensitivity and maintain an adequate level of specificity." However, CA-125 measurement as part of a multimodality screening program "requires further evaluation" since "this application of CA-125 is not of proven benefit for the early detection of ovarian cancer."