Two-step Screening for Ovarian Cancer: A Scissor-step’ Forward?
Abstract & Commentary
By Robert L. Coleman, MD
Professor, University of Texas; M.D. Anderson Cancer Center, Houston
Dr. Coleman reports no financial relationships relevant to this field of study. This article originally appeared in the October 2013 issue of OB/GYN Clinical Alert.
Synopsis: An 11-year prospective screening trial in 4051 menopausal women (age 50-74) demonstrates that the risk of ovarian cancer algorithm probability index, along with ultrasound and gynecologic oncology consultation of "high-risk" cases, produces high specificity and positive predictive value leading to the identification of higher than expected incident early-stage ovarian cancers.
Source: Lu KH, et al. 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. Cancer 2013; Aug 26. Doi: 10.1002/cncr.28183 [Epub ahead of print].
Ovarian cancer is a highly lethal disease of low prevalence in the general population and is associated with advanced stage presentation in more than 70% of incident diagnoses. Strategies to identify these clinical features early in the disease process have included a focus on symptomatology, blood-based biomarkers, imaging (predominately ultrasonography), and physical exam with little success.
In this study, a two-stage ovarian cancer screening strategy was evaluated that incorporates change of CA125 levels over time and age to estimate risk of ovarian cancer. Women with high-risk scores were referred for transvaginal ultrasound (TVS). The prospective study included postmenopausal women (age 50-74) with at least one retained ovary, no personal history of ovarian cancer, and no family history of a first- or second-degree relative with breast or ovarian cancer. Patients were excluded if they had had another malignancy, other than breast cancer, within 5 years of enrollment. Participants underwent an annual CA125 blood test. Based on the Risk of Ovarian Cancer Algorithm (ROCA) result, women were triaged to a subsequent annual CA125 test if deemed low risk, a repeat CA125 test in 3 months if intermediate risk, or TVS and referral to a gynecologic oncologist if high risk. A total of 4051 eligible women participated over 11 years, accounting for 16,832 screen years. The average annual rate of referral to a CA125 test in 3 months was 5.8%, and the average annual referral rate to TVS and review by a gynecologic oncologist was 0.9%. Ten women underwent surgery on the basis of TVS triage, which identified four invasive ovarian cancers (one with stage IA disease, two with stage IC disease, and one with stage IIB disease), two ovarian tumors of low malignant potential (both stage IA), one endometrial cancer (stage I), and three benign ovarian tumors, providing a positive predictive value of 40% (95% confidence interval [CI], 12.2%, 73.8%) for detecting invasive ovarian cancer. The specificity was 99.9% (95% CI, 99.7%, 100%). All four women with invasive ovarian cancer were enrolled in the study for at least 3 years with low-risk annual CA125 test values prior to rising CA125 levels (CA125 change point). The authors concluded that ROCA followed by TVS demonstrated excellent specificity and positive predictive value in a population of U.S. women at average risk for ovarian cancer.
Some may recall the children’s game "Mother may I?" in which participants follow instructions from a leader (mother) trying to guess the progress of the participants (children) toward a goal (home). Granted requests, such as "baby steps" or "scissor steps," afford minor advancements toward progress. The current study, while offering hope and promise of bridging the gap between early detection and mortality from ovarian cancer, essentially represents a minor, albeit positive, step forward. It clearly demonstrates that the process will involve several key elements such as good biomarkers, good probability estimate algorithms, motivated follow-up, access to secondary triage and gynecologic oncology consultation, and ultimately validation by documentation of reduction in mortality. However, to reach this zenith, studies assessing the merits of the individual steps forward are necessary and build confidence in the process.
It is clear that a screening program’s best opportunity to reduce disease mortality is to identify a pre-invasive state where intervention will prevent disease.1 Currently, we have no such marker for ovarian cancer, which likely also consists of a high proportion of fallopian tube abnormalities that form the primary site of disease.2 Thus, the next best opportunity to affect mortality in ovarian cancer is "stage migration;" that is, diagnosing the disease at earlier stages than what is currently observed in the general population. This can measurably impact survival as early-stage disease is curable at rates two- to three-fold that of advanced disease. Since more than 70% of ovarian cancer diagnosed in 2013 is stage IIIC/IV, there is significant opportunity to vastly alter its natural history and expected disease-specific mortality under a screening program that results in significant stage migration. The performance of the two-step screening algorithm used in this trial is remarkably consistent with the initial prevalence report presented by the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS)3 group in 2009.4 Although incidence was not reported in that study, the general specificity and positive predictive value of the two-step algorithm (which is similar to the current study) were very consistent and, fortunately for screening participants, required only two to three operations from "positive screens" to identify cases of ovarian cancer — all in "early-ish" stages (Stage IC-IIB).
It was also of interest that two of the four cancers identified in this study had "normal" CA-125 values but met the criteria for a significant CA125 change point while remaining within the normal range. This highlights the value of establishing a normal baseline and the power of serial investigation. In addition, nearly 85% of all participants just went on getting annual CA-125 tests, demonstrating the low overall specificity of the program. The 200,000 person UKCTOCS trial’s primary endpoint is disease-specific mortality; it is hoped that the preliminary results will be confirmed, offering practice changing
policies in the management of screening menopausal women.
- Grimes DA, Schulz KF. Uses and abuses of screening tests. Lancet 2002;359:881-884.
- Crum CP, et al. The distal fallopian tube: A new model for pelvic serous carcinogenesis. Curr Opin Obstet Gynecol 2007;19:3-9.
- Menon U, et al. Sensitivity and specificity of multimodal and ultrasound screening for ovarian cancer, and stage distribution of detected cancers: Results of the prevalence screen of the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS). Lancet Oncol 2009;10:327-340.
- Coleman R. Screening for ovarian cancer: One scissor step’ made. OB/GYN Clinical Alert 2009;26:22-23.