Does screening for ovarian cancer using an algorithm that customizes interpretation and follow-up of cancer antigen (CA) 125 testing reduce ovarian cancer mortality in average-risk women?
Like the U.S. Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, this large and long U.K. study found no reduction in ovarian cancer mortality with screening using ultrasonography or a multimodal CA 125–based screening strategy. (Level of Evidence = 1b)
Study design: Randomized controlled trial (double-blinded)
Funding source: Government Allocation: Concealed Setting: Population-based
The U.K. Collaborative Trial of Ovarian Cancer Screening was a randomized trial that recruited approximately 200,000 average-risk women 50 to 74 years of age. They were randomized in a 1:1:2 ratio to multimodal screening, annual ultrasonography, or usual care. Multimodal screening was based on CA 125 testing, but instead of a single cutoff for all women, there were individualized cutoffs. Significant changes from each woman’s baseline triggered additional blood tests and if necessary, ultrasonography and biopsy.
Women were recruited between 2001 and 2005 and were screened regularly until 2011. An interim report of this trial after a median of 11 years of follow-up initially claimed to have detected a reduction in disease-specific mortality in a post-hoc analysis of incomplete data.
This interpretation was criticized in an editorial from the Ovarian Cancer Research Fund Alliance and the Banbury Conference Writing Group because the investiga- tors have a financial stake in the patented CA 125 algorithm. The results in the full report are based on follow-up for a median of 16.3 years with complete follow-up for 95% of the cohort. Overall, 0.9% of women in each group were given a diagnosis of invasive epithelial ovarian or tubal cancer.
There was no significant difference between groups in ovarian cancer–specific mortality (0.58% in the multimodal screening group vs. 0.61% in the control group; P = 0.58). Second- ary analyses looking at only early and late deaths found no difference in mortality, and ultrasonography had no benefit. There was a stage shift with multimodal screening: 47% more women had a diagnosis of stage I disease, and 25% fewer had stage IV disease, but this did not translate into a mortality benefit. The incidence of disease stages III and IV combined was only 10% lower. Harms in terms of the number of biopsies and surgeries were not reported. Based on calculations from supplemental materials, there was no difference in all-cause mortality between groups.