Copy-Number Profiling and Methylation-Based Tumor Typing during Prenatal Cell-Free DNA Screening.

BACKGROUND

Prenatal cell-free DNA (cfDNA) screening is primarily designed to detect fetal chromosomal abnormalities, but can also identify aneuploidies derived from other tissues, including cancer. The identification of aneuploidies of unknown origin during prenatal cfDNA screening can lead to time-consuming multistage investigations and anxiety for the expecting mother.

METHODS

To expedite the identification of the origin of copy-number aberrations and guide clinical management of such profiles suggestive of maternal malignancy, we developed a methylation and aneuploidy-aware prenatal screening pipeline. Plasma cfDNA is enzymatically converted to identify the methylated cytosines during sequencing. The tissue of origin is predicted by leveraging a cell-type-specific methylome atlas into our methylation-based deconvolution algorithm, MetDecode.

RESULTS

We demonstrate that aneuploidy profiling on enzymatically converted cfDNA enables the identification of placental and cancer-derived aneuploidies with similar accuracy compared with conventional prenatal cfDNA screening. The methylation-based deconvolution pinpointed the tumor origin correctly in 91.67% of the pregnant women with a tumor fraction >3%.

CONCLUSION

Methylome and aneuploidy-aware cfDNA screening could substantially improve the diagnostic processes, pinpoint the origins of aneuploidy and improve cancer management during pregnancy.