A 2-Stage Ovarian Cancer Screening Strategy Using the Risk of Ovarian Cancer Algorithm (ROCA) Identifies Early-Stage Incident Cancers and Demonstrates High Positive Predictive Value

Abstract

Diagnosis of ovarian cancer at an early stage is associated with a rate of survival of 75% or higher. Most women with this cancer are diagnosed at a late stage when long-term cure rates are less than 30%. Currently, there are no proven screening strategies for the early detection for ovarian cancer. A definitive diagnosis requires invasive surgery and removal of the ovaries. Any screening strategy for this cancer must minimize false positives to decrease the number of unnecessary operations. Measurement of fixed cut-point levels of carbohydrate antigen 125 (CA-125), an ovarian tumor marker, and an abnormal transvaginal ultrasound (TVS) finding have been used as a screening strategy for ovarian cancer; however, neither has sufficient specificity. Recent studies have shown that the sensitivity and specificity of screening with CA-125 can be improved when rising CA-125 levels, even within a reference range, are used to prompt TVS. The Risk of Ovarian Cancer Algorithm (ROCA) is a computer algorithm that quantifies the risk of developing ovarian cancer in postmenopausal women at normal risk who are screened for this cancer. The aim of this single-arm prospective study was to determine if a 2-step strategy that incorporates change of CA-125 levels over time and age has sufficient specificity and positive predictive value for estimating the risk of ovarian cancer in a population of postmenopausal women at normal risk. A small fraction of women with high-risk ROCA scores were referred for TVS as the second step of this strategy. Participants had an annual CA-125 blood test. Based on the ROCA score, women were triaged into the 1 of following 3 groups: a low-risk group (which would require return in 1 year for a CA-125 test), an intermediate-risk group (which would require women to repeat their CA-125 test in 3 months), or a high-risk group (which would require TVS and referral to a gynecologic oncologist). The decision for surgery was made by the gynecologic oncologist. The final study sample was composed of 4051 participants who were evaluated over 11 years (2001–2011). The average annual rate for triage to a CA-125 test in 3 months (intermediate-risk group) was 5.8%, and the average annual rate for triage to TVS and review by a gynecologic oncologist (high-risk group) was 0.9%. Ten women with suspicious TVS findings underwent surgery: 3 of these women had benign ovarian tumors, 2 had stage I ovarian serous tumors of low malignant potential, 4 had early-stage high-grade invasive ovarian cancers, and 1 patient was subsequently found to have endometrial cancer. The specificity with the 2-stage screening strategy was 99.9% (95% confidence interval [CI], 99.7%–100%). All 4 of the invasive ovarian cancers were found in women with normal, low-risk ROCA scores who had returned for annual CA-125 tests for at least 3 years prior to the rising CA-125 values. These data show that ROCA followed by TVS identifies early-stage incident ovarian cancer with excellent specificity and positive predictive value in a population of postmenopausal women at normal risk for ovarian cancer.