Transvaginal Ultrasonography as a Screening Method for Ovarian Cancer

Women from every Kentucky county participate in the Kentucky Ovarian Cancer Screening Program. Screening sites include: Maysville, Prestonsburg, Greenup, Elizabethtown, Somerset, Paducah and Lexington. Offsite participants account for 14% of the screening population with 86% being screened in Lexington. The long-term survival (20 year) of women with screen-detected ovarian cancers is twice that of unscreened women (65% vs 32%). Separation of cases into Type 1 and Type 2 ovarian cancer shows that screening improves the survival of both Type 1 and Type 2 ovarian cancers. Type 1 ovarian carcinomas for the screened and unscreened populations were defined based on these WHO criteria: mucinous carcinomas all grade, clear cell carcinomas all grades, endometrioid carcinomas grades 1 \& 2, serous carcinomas grades 1 \& 2, and malignant Brenner's tumors all grades. Type 2 ovarian carcinomas for the screened and unscreened populations were defined based on these criteria: undifferentiated carcinomas, endometrioid carcinomas grade 3, serous carcinomas grade 3, and carcinosarcomas. While long-term 20-year survival of women with Type 1 ovarian cancers detected by screening was significantly better than for unscreened women (81% v 46%, respectively), the survival benefit was even more pronounced for Type 2 ovarian cancers detected by screening of Kentucky women compared to unscreened Kentucky women (55.7% vs. 0.3%, respectively) or unscreened women at UK Hospital (12%). Screen-detected cases of Type 2 invasive ovarian cancers had better survival than unscreened cases when those detected had early- or late-stage disease. However, better survival was achieved when Type 2 ovarian cancers were detected at an early (72%) compared to late-stage (46%). Our data support the effectiveness of the screening protocol at the University of Kentucky, and subsequent treatment in accordance with National Comprehensive Cancer Network guidelines. The significance of these findings is that our approach has resulted in the detection of both early-stage Type 1 and Type 2 ovarian cancers and these cases have had improved survival when compared to that of unscreened cases, indicating that the screen-detected cases are associated with a potential survival advantage even for aggressive ovarian carcinomas. The primary objective of this study is to prospectively evaluate the false positive (FP) percentage generated by the ovarian screening algorithm and determine whether serial transvaginal ultrasonography can lower the FP percentage as demonstrated in the retrospective analysis. The aim of serial ultrasonography is to decrease FP percentage to 0.32% (positive predictive value of 24%) without adversely impacting the results for true positives and false negatives. on a "per woman screened basis" since this corresponds to a minimally acceptable positive predictive value of 24% or higher. This assumes an average of three screening years for each new woman entering the program.