Detecting cancer earlier is an important component in the mission to increase cancer survival rates. Treatment is much more likely to be successful if cancers are caught earlier, before they are at late stage or have spread to other parts of the body.
There are a number of strategies that aim to shift cancer diagnoses earlier. These include regular monitoring of people with pre-cancerous conditions that put them at higher risk of developing cancer and national screening of asymptomatic individuals to test for breast, cervical and colon cancer.
The cancer tests used in monitoring and screening programmes have been optimised to be specific to avoid the incorrect diagnosis of cancer in healthy individuals, and sensitive to not miss cancers that are present. However, using current technology, these tests can each only look for one type of cancer, and multiple different tests are necessary to check for a range of cancers. In the context of screening an asymptomatic population, testing for different cancers separately is time-consuming and expensive for healthcare providers, and likely to have lower participant uptake. Furthermore, there are only a few limited cancers for which an effective screening test exists. For many cancers, including those with particularly poor outcomes such as pancreatic and oesophageal cancers, patients need to be symptomatic in order to trigger the diagnostic process, by which point, the cancer may already be well advanced. Scientists are now looking to develop a new type of test that can simultaneously detect a range of cancers.
A new approach through liquid biopsies
In order to develop a multi-cancer test, many researchers, including those from Oxford, have turned to liquid biopsies, using blood, urine or saliva. These are an attractive option since they are clinically minimally invasive and have the potential to inform on multiple cancers in the same test. Blood contains small fragments of DNA – cell free DNA, cfDNA - that have been released into the bloodstream from many parts of the body. Of particular interest is the circulating DNA that originates from tumour cells – circulating tumour DNA, ctDNA – in patients with cancer.
Since ctDNA is present in very low levels in the blood, one of the major challenges is detecting this small amount of ctDNA, and differentiating it from cfDNA from normal cells to assess whether there is a cancer somewhere in the body. The next challenge is to then identify from which tissue the ctDNA originated in order to be able to locate the cancer and do further tests to confirm the diagnosis.
There are two types of information that can be extracted from circulating DNA to help with these two challenges. The first is the sequence of the ctDNA, which can reveal DNA mutations and copy number changes that indicate the presence of cancer. However, there are only a handful cancer-specific mutations in the whole cancer genome that can distinguish ctDNA from all the cfDNA from normal cells and the probability of detecting the small numbers of ctDNA fragments that contain these mutations is relatively low. This fact combined with the issue of variability between the mutations found in cancer cells, even within the same tumour, means that it is difficult to confidently and sensitively identify ctDNA, and therefore cancer, using the DNA sequence alone. Additionally, it is difficult to determine the ctDNA’s tissue of origin, especially for early stage cancers, with this approach.
This is where the second type of information comes in. DNA can be chemically modified by methylation on cytosine bases. These modifications have important roles in several cellular regulatory processes but are frequently altered in cancer cells and preserved in ctDNA. An advantage of measuring methylation is that the altered methylation patterns that distinguish cancer from normal DNA are spread across hundreds of sites, increasing the chances of differentiating ctDNA from normal cfDNA. Importantly, previous work by Oxford’s Dr Chunxiao Song (Ludwig Institute for Cancer Research) has shown that DNA methylation has characteristic patterns in cancer cells from different organs such as the lung, pancreas and liver, therefore it should be possible to use blood-derived DNA methylation patterns to detect the presence of cancer in the body and in which organ the cancer is located.
Developments in DNA methylation-based cancer testing
In an exciting new development, a blood test has been developed based on DNA methylation that can detect over 50 types of cancer. Published in the Annals of Oncology, a team of clinicians, academics, and the biotech company GRAIL developed a cancer test that has very high specificity (99.3%) and good cancer location accuracy (93% in the 96% of cancers for which the test predicted the tissue of origin). However, the sensitivity of the test varied for different types of cancers and stages. Very late stage cancers were detected in 93% of cases but early stage cancers in only 18%. This reflects the fact that smaller cancers release less DNA into the blood, making them harder to detect. The difference in sensitivity between cancer types could also be attributed to differential shedding of ctDNA between organs.
Overall, this publication is a huge step forwards, providing the proof of principle that DNA methylation can be used as the basis of a test for multiple cancers. Particularly promising is the 63% sensitivity for early stage pancreatic cancers, a cancer in much need of an assay suitable for screening asymptomatic individuals. However, the low sensitivity for most other early stage cancers means that this technology needs further development for utility in the screening setting.
TAPS: a new technology to improve sensitivity
At the Ludwig Institute for Cancer Research, Oxford Branch, Dr Chunxiao Song has developed a method called TAPS (TET-assisted pyridine borane sequencing) that might be able to tackle this issue of sensitivity. The Annals of Oncology study used the standard way of measuring DNA methylation called bisulphite sequencing. Bisulphite is a harsh chemical that causes severe degradation of the DNA sample, thus lowering the amount of DNA available for detection and therefore the test’s sensitivity. Further, bisulphite sequencing detects DNA methylation indirectly, converting all non-modified cytosine to thymine. This decreases sequencing complexity, reducing sequencing and mapping quality, and increases sequencing costs.
By contrast, TAPS is a bisulphite-free sequencing method for DNA methylation. TAPS biochemistry selectively and directly converts methylated cytosine to thymine in a mild reaction that preserves DNA integrity and is effective at very low DNA concentrations.
Figure: Overview of the TAPS chemical biology method for bisulphite-free, base-resolution, and direct sequencing of DNA methylation. Ten-eleven translocation (TET) enzyme oxidation of methylated cytosine (mC) to 5-carboxylcytosine (5caC) is followed by pyridine borane reduction of 5caC to dihydrouracil (DHU). Subsequent polymerase chain reaction (PCR) converts DHU to thymine, enabling a C-to-T transition of mC. By comparing to a reference genome sequence, it is possible to identify the cytosine bases that have been converted to thymine during TAPS and therefore those that were methylated.
Due to its unique direct mechanism, TAPS generates fewer errors and is cheaper than bisulphite sequencing. TAPS can also detect genetic mutations and copy number variations, so it is possible to combine the two types of information described above, increasing the data that can help distinguish cancer-originating from non-cancer-originating DNA. Furthermore, additionally understanding the mutational landscape of the tumour can help inform on possible treatment strategies.
Figure: TAPS can detect mutations, copy number variations and DNA methylation profiles in the same sequencing run.
With its increased sensitivity over bisulphite sequencing and the richness of information gained, TAPS has great potential for a combined DNA methylation and sequence-based cancer blood test.
The work by the GRAIL consortium has made significant headway by demonstrating the utility of DNA methylation as a diagnostic biomarker for cancer. I believe that the different chemistry used in the TAPS method here in Oxford will further increase the sensitivity of such approaches and allow combination of epigenetic with genetic information for an improved power of detection. - Dr Chunxiao Song, Principal Investigator at the Ludwig Institute for Cancer Research
Future directions for TAPS technology
TAPS technology has been developed by Chunxiao and his team in Oxford, who now intend to take it to the next level by further optimising and validating TAPS for use in a clinical setting. They have founded a new company called Base Genomics that is developing TAPS into a product and expediting its path towards the clinic.
Genomic technologies with the power, simplicity and broad applicability of TAPS come along very infrequently. In my previous role I experienced the transformative effect that such breakthrough technologies can have on science and healthcare. I’m excited to be working with the team at Base Genomics to help TAPS to make a similar impact, deepening our understanding of epigenetics and directly enabling improved outcomes for patients. - Vincent Smith, Base Genomics’ Chief Technology Officer and former Vice President of Consumables Development at Illumina
Base Genomics has also been focusing on the clinical applications of TAPS and has been working closely with Professor Anna Schuh, Director of the Oxford Molecular Diagnostics Centre and Honorary Consultant Haematologist.
As a clinician, it is hugely exciting to be involved in the optimisation of the TAPS methodology for the clinic. Technology such as this has the potential to make a real difference to how cancers are diagnosed. - Professor Anna Schuh, Director of the Oxford Molecular Diagnostics Centre and Honorary Consultant Haematologist
Dr Françoise Howe, May 2020