Julianna LeMieux Ph.D. Senior Science Writer GEN
Identifying the Methylation Patterns of cfDNA Aids in Identifying Cancer’s Origin
There will be roughly 1.7 million new cases of cancer diagnosed in the United States this year. For each one, the earlier the diagnosis the better.
Although some cancers have routine screening methods such as colon cancer, multiple barriers prevent one-third of eligible people from being screened. For many cancers, with no routine screening methods, the cancer is frequently recognized past the point of available treatment options. Taken together, these hurdles make methods to detect cancer in its earliest stage, in as many people as possible, two major goals of cancer research.
One of the most exciting areas of development in early cancer detection lies in cell-free DNA (cfDNA). These short fragments of DNA circulate in the blood after being released from cells, including cancerous cells.
One way cfDNA can signal the presence of cancer is through genetic variation. Because the DNA sequence in cancer cells has differences from non-cancer cells, those sequence variations, present in cfDNA, can be used to detect the presence of cancer.
However, as Kristina Warton Ph.D., a researcher within the Gynaecological Cancer Research Group at the University of New South Wales in Sydney, Australia points out, “identifying cancer based solely on its DNA mutations requires a preexisting idea of all of the mutations associated with the cancer. This is difficult without prior access to tissue from each particular tumor. This is work that is being done, but is both laborious and expensive.”
A cheaper and easier way to use cfDNA in cancer screening is to harness the information held in the epigenetic marks on the DNA. Epigenetic marks, such as the addition of methyl groups to DNA, are highly tissue specific and also dysregulated in cancer. Therefore, when the methylation pattern on the DNA is analyzed, it identifies from which tissue the cfDNA has been released, uncovering the location of the cancer.
A cfDNA Success—FDA-approved Colorectal Cancer Screening Test
One area where cfDNA has been implemented clinically is in colorectal cancer, the second-leading cause of cancer death in the United States. Epi proColon (made by Epigenomics,) the only FDA-approved blood-based cfDNA test for screening, is based on the detection of tumor-specific methylation on the SEPT9 gene (septin-9 protein). In the DNA of colorectal cancer tissue, but not normal tissue, the cytosine residues on the SEPT9 gene are methylated. This cancer specific methylation can be harnessed to detect the presence of tumor cells in the colon.
Theo deVos, Ph.D., the director of development and commercial operations at Epigenomics, notes that the company is tackling multiple different cfDNA-based projects, in addition to advancing earlier detection of colorectal cancer. Another goal is to develop tools to be used in conjunction with other cancer screenings that may give inconclusive findings, such as CAT Scans for lung cancer. deVos’s goal is to “add information to make these findings more relevant without having to go in to do a biopsy.” The researchers at Epigenomics are also identifying new methylation biomarkers the company can add to cancer screening panels.
With regard to the power of using methylation for cancer screening, deVos says the choice of DNA sequence versus DNA methylation “strongly depends on what you are trying to do.” He notes, “if the clinical application is treatment selection, genetics are important because a lot of cancer treatments are designed around mutations.”
However, he adds that “for broader applications like screening, methylation allows us to find a biomarker that covers a whole range of cancers.” One example of the crossover of SEPT9 between cancers was detailed in a paper published in April, 2018 in EBioMedicine, by Oussalah et al. In this research, Oussalah and team used SEPT9 to identify hepatocellular carcinoma (HCC) in patients with cirrhosis.
In the field of cfDNA cancer screening, many eyes are focused on Illumina spinout Grail, due to their deep pockets and scientific bandwidth. Grail is striving to become the leader in the liquid biopsy field, as evidenced by the enormous undertaking of their ongoing Circulating Cell-free Genome Atlas (CCGA) Study. This project is designed to establish a baseline of cell free nucleic acid (cfNA) profiles in an estimated 15,000 people to be recruited by the end of 2018, including both cancer patients and healthy controls. To date, 11,000 people have enrolled.
This approach of gathering large amounts of data is an important one, according to Hao Wu Ph.D., an associate professor at the Department of Biostatistics and Bioinformatics at Emory University. “A major challenge right now is to collect more data, especially from large-scale, population level studies,” Wu says. “We should also collect data from normal people with different demographics, so that we can establish a baseline.”
In the initial discovery phase of CCGA, Grail took a three-pronged approach to gathering information about blood-based tests as a screening tool for early detection. The first looked at DNA mutations (single nucleotide variants and small insertions and/or deletions), the second used whole genome sequencing to detect abnormal number of copies of genes, and the third used bisulfite sequencing to analyze cfDNA methylation patterns.
Charlotte Arnold, head of communications at Grail, notes that “We are now optimizing our assays and machine learning algorithms to determine the most informative genomic features for continued development and validation of a blood test for early detection of multiple cancer types.”
Although Grail did not need more clout in the field, it gained some when it merged with Cirina in 2017, the company started by one of the founders of the cfDNA field, Dennis Lo, M.D., Ph.D. Lo pioneered the field of cfDNA through the discovery of cell-free fetal DNA circulating in maternal blood, work that deVos cites as, “the basis of some of the best work in the whole field.”
The Process is Easy, If You Have the DNA
The methylation pattern on cfDNA can be determined using bisulfite treatment which is easy, quick, and cheap. Bisulfite conversion kits can be purchased from many different vendors for a few hundred dollars for dozens of reactions.
Treating DNA with bisulfite changes cytosines (one of the four nucleotide DNA bases) into uracils, which are normally found in RNA. However, the change only happens in unmethylated cytosines. The cytosines that have methyl groups attached to them (5-methylcytosine), will not be converted to uracil. After bisulfite treatment, the DNA is sequenced. In the sequence, the methylated cytosines which were protected from conversion will still be read as a “C” whereas the unmethylated cytosines will have been changed. Therefore, because the bisulfite treatment creates methylation dependent changes in the DNA sequence, that analysis of the sequence after the bisulfite treatment reveals the methylation patterns on that piece of DNA.
With the kits available, bisulfite treatment of DNA seems as easy as any standard molecular biology protocol. However, many researchers in the field agree that the biggest challenge to working with cfDNA is getting enough of it. Not only is cfDNA found in extremely low concentrations, but it is also fragmented—and the bisulfite sequencing process fragments it further. Therefore, cfDNA can be difficult to obtain in adequate quantities.
It is a problem that is not easily solved. Increasing the signal seems a likely solution, but is not a panacea. When screening, deVos says “you want to have as high a sensitivity as possible at as high a specificity as possible. But, when you push the sensitivity up, things come up in the background.” The addition of more markers is one way to improve the process. However, technical advances in this process would improve cancer screening dramatically.
A second challenge, noted by Warton, is the ability to obtain blood that has been biobanked appropriately in order to use it clinically. Some of the cfDNA tests (like the FDA-approved Epi proColon) can use plasma, while others require cfDNA from serum. Although frequently mistaken for each other, serum and plasma cannot be used interchangeably. Serum contains more DNA than plasma, but the DNA is a contaminant from lysed leukocytes in the blood—another issue holding back clinical implementation that will require further technical development.
A Blood Test for Cancer Screening Is in the Future
Despite the work that needs to be done, Warton remains optimistic that there will be a day in the future when a routine screening blood test will identify both the presence of cancer and its location. And, she thinks that it may not be that far away. She says that any doubts can be dismissed by looking at how cfDNA has revolutionized prenatal testing. “The speed at which that work moved into the clinic happened at a breathtaking pace,” she says. “If cancer diagnosis were to follow down the same path, it will happen quickly.”
A recent development by Epigenomics is a step in the right direction. The company has designed a product to allow researchers to bring a plasma sample to bisulfite-?labeled DNA with one kit. With this, researchers and clinicians can run next generation sequencing (NGS) panels in their own laboratories, bringing the power of using cfDNA for cancer screening to anyone who wants to use it.
David Bull, the director of marketing at Epigenomics, wants to see a day when everyone is screened for colon cancer. Because, as he says, “screening tools are only useful if they are used.” He notes that “guaiac fecal occult blood test (gFOBT) is not a great test, but, studies have proven that when used annually this imperfect screening test reduces colon cancer mortality and morbidity.” Bull says that we have to look at these tests from a program standpoint—not a point in time. But at this point in time, the cfDNA field is at the beginning of a turning point that may end with a routine blood test at annual check-ups to screen for cancer.
This article was originally published in the July/August 2018 issue of Clinical OMICs. For more content like this and details on how to get a free subscription, go to www.clinicalomics.com.