NEW YORK — DNA methylation signatures found within cervical cells may be able to identify women at risk of ovarian and breast cancer.
Researchers from the University of Innsbruck in Austria have identified two DNA methylation signatures reflecting ovarian cancer and breast cancer risk. Genetic factors like inherited mutations in the BRCA1 or BRCA2 genes as well as non-genetic factors like environmental exposures or lifestyle contribute to ovarian and breast cancer risk, and epigenetic signatures may be able to capture a combination of both those risk sources.
The two signatures, dubbed the Women's risk Identification for Ovarian Cancer, or WID-OC-index, and the Women's risk Identification for Breast Cancer, or WID-BC-index, are detected within cervical cells, which are biologically related to the tissues from which ovarian and breast cancers typically arise but are more easily accessible for sampling.
"Our aim is to identify the vast majority of women who are at risk of developing a women-specific cancer — irrespective of genetic or non-genetic factors," senior author Martin Widschwendter, a professor at Innsbruck, wrote in an email.
As they reported in the first of the two Nature Communications papers appearing on Tuesday, Widschwendter and his colleagues relied on cervical smear samples from a cohort of 242 women with ovarian cancer and 869 women without ovarian cancer to generate the WID-OC-index. The WID-OC-index includes a large number of CpG sites, reaching optimal performance with 14,000 CpGs.
In a validation cohort of 47 ovarian cancer cases and 225 controls, this WID-OC-index could identify women with ovarian cancer with an accuracy of 76 percent. The signature, the researchers noted, did not appear to be due to the presence of tumor DNA within the cervical smear.
In their second paper, the researchers described using a similar approach to develop the WID-BC-index. In particular, they collected cervical liquid-based cytology samples from 329 women with primary breast cancer with features linked to poor prognosis like large tumor size or lymph node involvement and 869 women without breast cancer. This WID-BC-index includes about 29,000 CpGs.
In a validation cohort of 225 controls and 113 breast cancer cases, the WID-BC-index could identify women with breast cancer with an accuracy of 81 percent. Further, women in the top quartile of the WID-BC-index had a 15.7-fold increased risk of breast cancer, independent of other risk factors.
The findings suggested to the researchers that epigenetic programming defects occur among women at increased risk of developing cancer.
According to Widschwendter, the WID-OC-index and the WID-BC-index could be used as first-line screens for disease with follow-up testing based on the risk levels they reveal. For instance, he said that women with high WID-OC-index levels could subsequently undergo CA-125 and plasma cell-free DNA testing, while women with high WID-BC-index levels could undergo breast MRI and analysis of cell-free DNA.
Widschwendter and his colleagues have also developed two other WID indices, one for endometrial and one for cervical cancer. The next step, he said, is to assess prospectively how well all four indices work on one cervical sample and whether they, in combination with other measures, enable the early detection of cancer.
At the same time, the researchers also plan to examine whether selective progesterone receptor modulators can prevent highly aggressive breast and ovarian cancers, and if epigenetic signatures can be deployed to gauge how well that approach works.