The largest ever prospective randomised trial of ovarian cancer screening are largely disappointing
The results of the largest ever prospective randomised trial of ovarian cancer screening are largely disappointing, with no significant differences in mortality found between screened and unscreened women.
Earlier trials had failed to show any survival benefits even when ovarian cancer was detected early, and prognosis has improved little in 30 years, said the authors of the study published in the Lancet.
Designed to establish the effect of screening postmenopausal women on ovarian cancer mortality, the study randomised over 202,000 women randomised to undergo transvaginal ultrasound, ultrasound with CA215 (multimodal screening or MMS), or no screening.
After 14 years, there was a 15% reduction in mortality in the multi-modal screening group and an 11% reduction in the ultrasound group, but the findings were not significant.
This was despite multimodal screening leading to early detection of ovarian cancer in 39% of patients, compared to a 26% rate in the no screening group.
The authors said the findings were nevertheless encouraging because, once prevalent cases were excluded (where women were found to have ovarian cancer at recruitment), a significant mortality reduction of 23% was found in years seven to 14 in women in the MMS group, the authors said.
“Findings from this trial suggest that for 641 women screened annually using the multimodal strategy for 14 years, one ovarian cancer death is prevented”, the authors wrote.
By comparison, “for 1000 women invited to biennial mammography screening for 20 years from 50 years of age, two to three women are prevented from dying of breast cancer,” they said.
The study authors said the results were encouraging for various reasons, including the leveling off of the mortality rate for both screened groups, but not in the control group.
An accompanying editorial said the focus should now be on unraveling the mechanism behind the findings so screening could be improved.
“Awareness and symptom recognition for diagnosis of ovarian cancer at an early stage will be difficult to improve upon,” the editorial said.
And looking for new tumour markers was probably not a high priority, since of 28 new biomarkers tested in previous studies, none were better than CA125 at predicting ovarian cancer risk.