This article is more than 5 years old. Information may no longer be current.
Questions remain for more aggressive ovarian cancer screening
Havrilesky LJ. Cancer. 2010;doi:10.1002/cncr.25624.
Click Here to Manage Email Alerts
Annual screening for ovarian cancer in the general US population could lead to a mortality reduction of approximately 10% to 15%, according to recent statistical model results. This modest reduction could be further offset by detection of indolent cancers.
Researchers said the aim of the study was to estimate the effect of a “two-phenotype paradigm of epithelial ovarian cancer on the mortality reduction achievable using available screening technologies.” The one-phenotype Markov model of ovarian cancer natural history was modified to a two-phenotype model that included aggressive and indolent phenotypes based on histopathologic criteria.
Data for stage distribution, incidence rates and mortality outcomes for individual cases were calibrated to findings from the NCI’s SEER database.
The findings of the current analysis were validated by multimodality prevalence screen from a Monte Carlo micro-simulation, based on 1 million events, of the U.K.’s Collaborative Trial of Ovarian Cancer Screening. Based on an annual screening model, mortality reduction and positive predictive value were estimated.
When validated against data from the U.K. findings, the current model predicted that 41% of screen-detected stage I cancers and 47% of screen-detected stage II cancers would be diagnosed.
The predicted proportion of screen-detected early stage cancers was 40.9% for the two-phenotype model and 40.8% for the one-phenotype model; these findings were compared with a rate of 47.1% (95% CI, 29.8-64.9) of early-stage screen-detected cancers in the U.K. trial.
A 1-year screening period for patients with a mean age of 60 years yielded a 27.4% positive predictive value for the two-phenotype model and a 26.5% positive predictive value for the one-phenotype model; comparatively, the U.K. trial yielded a positive predictive value of 35.1% (95% CI, 25.6-45.4).
Annual screening of an average-risk US population of adults aged 50 to 85 years resulted in a model-estimated positive predictive value of 14%, according to the results. The model also predicted five lifetime false-positive tests per 1,000 women linked to annual screening in this age population, based on both the one-phenotype and two-phenotype models.
Compared with no screening, annual screening of women aged 50 to 85 years was linked to a mortality reduction of 14.7% in the one-phenotype model and 10.9% in the two-phenotype model. Regardless of screening frequency or test sensitivity, the two-phenotype model was associated with a possible achievable mortality reduction of 3% to 6%.
The two-phenotype model indicated that aggressive tumors would account for 62% of detected cancers and 68% of deaths, according to the results.
“The current analysis suggested that reductions in ovarian cancer mortality using available screening technologies on an annual basis are likely to be modest,” the researchers wrote. “A model that incorporated [two] clinical phenotypes of ovarian carcinoma into its natural history predicted an even smaller potential reduction in mortality because of the more frequent diagnosis of indolent cancers at early stages.”
 |
Follow HemOncToday.com on Twitter. |