Abstract
A massive study of ovarian cancer screening has determined that annual screening with ultrasound alone or CA125 testing and, if necessary, follow-up with ultrasound does not reduce mortality in the general population. Although more early cases came to light with screening, the increase was not large enough to yield a statistically significant survival improvement.
According to one of the largest clinical trials ever performed, ovarian cancer screening does not decrease disease mortality in the general population (Lancet 2021;397:2182–97). If population screening is to become a reality, researchers need to identify more sensitive methods to detect ovarian cancer earlier and in larger numbers of women, findings of the UKCTOCS study suggest.
Ovarian cancer screening is not standard practice, even among women at elevated risk, because there's scant evidence it improves survival. The only other large clinical trial to assess its effectiveness, which enrolled more than 78,000 participants, detected no benefit from ultrasound and tests for the ovarian cancer biomarker CA125 (JAMA 2011;305:2295–303).
In 2016, interim results from the UKCTOCS trial, which began in 2001, hinted that screening might work after all. Researchers enrolled more than 202,000 women in the UK and randomly assigned them to one of three groups in a 1:1:2 ratio: One was offered annual screening with ultrasound; the second was offered CA125 testing, with ultrasound follow-up for participants whose results suggested higher risk; and the third received no screening. Researchers tracked participants even after screening ended in 2011. With a median follow-up of 11.1 years, mortality rates from ovarian or fallopian tube cancer did not differ among the three groups (Lancet 2016;387:945–56). But the data suggested that after 7 years, death rates were leveling off in both screened groups, raising hopes that longer follow-up would uncover a benefit.
However, the latest UKCTOCS findings, reflecting a median follow-up of 16.3 years, were negative. Mortality rates remained identical across all three groups: 0.6%. Screening did catch more cancers at an early stage. For instance, in women who were offered CA125 testing, the incidence of stage I and stage IV cancers was 47.2% higher and 24.5% lower, respectively, than in the unscreened group. But this so-called stage shift was too small to yield a statistically significant decrease in mortality.
The trial showed for the first time that early detection of ovarian cancer is possible, says lead author Usha Menon, MD, of University College London in the UK. To save lives, however, “we require a screening test that detects ovarian cancer even earlier—and in more women—than the screening strategy we used,” she says. “General population screening cannot be supported or recommended at this point,” she adds, cautioning that the trial findings cannot be extrapolated to women at high risk due to family history or who carry BRCA1/2 mutations, for whom screening is sometimes recommended.
“It's an extremely well-designed study,” says Karen Carlson, MD, of Massachusetts General Hospital in Boston, who wasn't connected to the research. For instance, instead of setting a cutoff for CA125, the researchers used an algorithm that detects trends in serum levels over time, a better way to flag worrying results, she says. “This makes it very clear there is no benefit” from screening.Given the negative findings, “the data suggest earlier-stage cancers are still likely to metastasize and recur,” notes Kevin Elias, MD, of Boston's Brigham and Women's Hospital. “Improving survival will require interrupting these tumors even earlier in their development than previously thought.”
The rarity of ovarian cancer—which afflicts one of every 2,500 to 3,000 women—is an obstacle, says Barbara Goff, MD, of the University of Washington in Seattle. “When you have a cancer that is that uncommon, it becomes very hard to screen for,” meaning that researchers need to identify more specific biomarkers. Although some promising possibilities are under investigation, including tumor DNA and microRNAs from liquid biopsies, they won't be ready for clinical use anytime soon, she says. –Mitch Leslie
