Skip to main content
Premium Trial:

Request an Annual Quote

Liquid Biopsy Tool Detects Epigenetic Signs of Cancer From Cell-Free DNA

This article has been updated to include that the EPINUC tool uses SeqLL's tSMS platform.

NEW YORK — Researchers have developed a liquid biopsy-based tool to analyze cell-free DNA for epigenetic and other signs of cancer.

The tool, called epigenetics of plasma-isolated nucleosomes (EPINUC) relies on a single-molecule fluorescent imaging system, which can image combinations of histone modifications, such as the ones associated with nucleosomes isolated from plasma samples. As cell-free DNA found in the blood from nucleosomes keep their tissue- and cancer-specific epigenetic states, the approach could be used to detect cancer biomarkers and diagnose disease from a small blood sample.

"Many of the conventional methods clinically available today to detect and diagnose cancer are invasive and unpleasant," senior author Efrat Shema from the Weizmann Institute said in a statement. This approach would instead take less than a milliliter of blood.

EPINUC, which the researchers described in Nature Biotechnology on Thursday, can detect six active and repressive histone modifications — including the trimethylation of H3K4me3, the acetylation of H3K9ac, and the trimethylation of H3K9me3 — as well as the ratios of these modifications.  At the same time, they extended the single-molecule approach to also detect protein biomarkers like carcinoembryonic antigen (CEA) and metalloproteinase-1 (TIMP1) — both of which are found at increased levels in the blood of colorectal cancer (CRC) patients — MST1, and p53. To generate these multiplexed single-molecule biomarker measurements, the approach relies on SeqLL's tSMS platform, a single-molecule sequencing tool.

In their proof-of-concept analyses, the Weizmann-led team first applied EPINUC to a set of 46 plasma samples from 40 individuals with late-stage CRC and 33 plasma samples from healthy individuals. They found that, as expected, CEA levels were higher in individuals with CRC and fell in CRC patients after resection. Meanwhile, individuals with CRC had higher levels of H3K27me3-, H3K9me3-, H3K9ac-, and H3K4me1-modified nucleosomes and a higher ratio of H3K9ac-to-H3K4me1 nucleosomes than healthy individuals.

When they extended their analysis to a set of 17 plasma samples from individuals with early stage CRC, the researchers again uncovered epigenetic differences, though not as many, but still indicating that epigenetic changes may occur early in disease.

The researchers additionally applied their EPINUC approach to 10 plasma samples from individuals with pancreatic ductal adenocarcinoma (PDAC) to find that their epigenetic profiles differed not only from healthy individuals but also from individuals with CRC. In particular, individuals with PDAC had very high levels of H3K4me3- and H3K27me3-modified nucleosomes. In a principal components analysis, early stage CRC samples fell between samples from healthy individuals and individuals with late-stage CRC, suggesting early-stage CRC could be a transitional stage, while samples from individuals with PDAC clustered separately from the CRC samples.

Further, the researchers reported that a machine learning algorithm based on EPINUC could classify CRC samples with a sensitivity of 92 percent, specificity of 85 percent, and precision of 92 percent. These findings suggested to the researchers that EPINUC could detect and diagnose cancer.

"We've achieved a successful proof of concept for our method, which now needs to be confirmed in clinical trials," Shema said. "In the future, our multiparameter approach may serve to diagnose not only various cancers but also additional diseases that leave traces in the blood, such as autoimmune disorders or heart disease."

The researchers noted that their findings were without the inclusion of sequencing data, which they said could add specificity and sensitivity. For instance, when they folded in single-molecule sequencing data from a small set of late-stage CRC samples and pancreatic samples — dubbing that approach EPINUC-seq — they could identify their tissues of origin. This method could be used, the researchers added, to trace the origins of cancers of unknown primary.