NEW YORK (GenomeWeb) – A European consortium has delivered up promising epigenetic markers for catching ovarian and breast cancer early, enabling physicians to better personalize treatment for patients.
The Epigenetics For Female Personalised Cancer Care (EpiFemCare) project recently published its findings in two papers in the journal Genome Medicine, focused on breast cancer and ovarian cancer, respectively.
"Our research has demonstrated an important step forward to using cell-free DNA as a marker to detect breast and ovarian cancer much earlier in the process," said Martin Widschwendter, lead author on both publications and head of the EpiFemCare project.
"Subject to further studies, we feel that this is a very promising way of avoiding issues around misdiagnosis or over-diagnosis — tumours which are detected, but wouldn't have caused clinical symptoms — in the future," he said. "Importantly," Widschwendter added, "our test might have a real potential to diagnose in particular those cancers which, once diagnosed early, would not eventually lead to death anymore."
The European Commission funded the EpiFemCare project in 2012, awarding €7.6 million ($9.3 million) to support the effort for five years. Led by Widschwendter, who heads the department of women's cancer at University College London, the effort also has involved Konstanz, Germany-based GATC Biotech, as well as Basel, Switzerland-based bioinformatics firm Genedata.
The premise of the work was that changes in methylation patterns of linked CpGs analyzed in DNA fragments shed by cancers into the bloodstream could provide specific signals indicating the presence of cancer. Working with partners at Charles University in Prague and Ludwig-Maximilians University in Munich, the project looked at around 1,880 serum samples from women with breast or ovarian cancers, as well as those with benign conditions and from healthy volunteers.
In addition, they folded in data collected from more than 200,000 women from the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS).
As detailed in their ovarian cancer paper, the researchers looked at roughly 700 cancerous and non-cancerous tissues using the Illumina Infinium Human Methylation 450K Beadchip and reduced representation bisulfite sequencing, undertaken by GATC Biotech using Illumina's MiSeq or HiSeq 2500 instruments, to identify the most specific methylation patterns related to ovarian cancer. They then developed a three-marker panel using high-coverage bisulfite sequencing and ran it on 151 women. The researchers also validated it on 250 women with various conditions, in addition to about 170 controls from UKCTOCS.
The panel was able to discriminate high grade serous ovarian cancer patients from healthy patients or those with a benign condition with a specificity and sensitivity of 90.7 percent. Moreover, the researchers identified 57.9 percent of women who developed ovarian cancer within two years of sample collection with a specificity of 88.1 percent.
"There has been very little change in the past 40 years in terms of progress on the early detection of ovarian cancer, yet we discovered several DNA methylation markers that could potentially identify 57.9 percent of women with ovarian cancers two years in advance of their diagnosis," Widschwendter underscored.
As discussed in the project's second, breast cancer-focused paper, the authors used bisulfite sequencing on 31 tissues and established blood-based tests using 110 additional samples. The clinical use of one particular region, called EFC#93, was then validated in 419 patients from a clinical trial called SUCCESS (Simultaneous Study of Gemcitabine-Docetaxel Combination adjuvant treatment, as well as Extended Bisphosphonate and Surveillance), as well as more than 900 patients from UKCTOCS.
Based on this data, the investigators identified 18 breast cancer-specific DNA methylation patterns, six of which underwent additional testing, and EFC#93, deemed the best candidate, was validated for clinical use. According to the paper, the presence of EFC#93 indicated a diagnosis of fatal breast cancer for 42.9 percent of women within three to six months, and 25 percent of women within six to 12 months of sample donation, with a specificity of 88 percent. Given the results, the authors determined that EFC#93 could be used as a "new tool for early diagnosis and management of disseminated breast cancers."
While Widschwendter deemed the results "very positive," he said that more work is needed to realize the potential of using cell-free DNA to diagnose breast and ovarian cancer.
"A cell-free DNA based test will have to overcome several hurdles — in particular validation in a large prospective clinical trial — before it can be implemented in a population-based setting," said Widschwendter. "We remain very hopeful to make significant progress in this field for the women of the future who may be at risk of developing cancer."
While he confirmed that studies continue related to the findings of the EpiFemCare project, he said it was too early to comment further on those efforts.
GATC and Genedata
GATC Biotech and Genedata were the two industrial partners in EpiFemCare, with GATC Biotech tasked with sequencing the samples, while Genedata provided the informatics pipeline.
Tobias Paprotka, director of R&D at GATC Biotech, said that the company was responsible for developing the sequencing methods and with cell-free DNA preparation. "We established a highly sensitive protocol based on reduced representation bisulfite sequencing, and sequenced numerous samples," Paprotka said. "Subsequently, we developed a protocol for targeted bisulfite sequencing to check the candidate regions," he added.
While this work led to the panels of markers discussed in the papers, Paprotka acknowledged that creating the protocol was "very challenging" for GATC Biotech. "Cell-free DNA is already highly degraded DNA and only present in very low amounts," he said, and "bisulfite treatment … further degrades the DNA."
To overcome this, GATC Biotech compared various protocol setups to hone a sensitive approach that maintained the complexity of the sample, Paprotka said. It then adjusted its protocol for high throughput, as several thousand sequencing libraries had to be prepared for EpiFemCare. "We adopted automated liquid handling systems to achieve highest accuracy and precision," he said.
The work benefited GATC Biotech in that it gained additional experience related to bisulfite sequencing. "Bisulfite-treated, cell-free DNA is one of the most challenging materials for next-generation sequencing and we can use it now for several products and projects," said Paprotka.
EpiFemCare also presented challenges for Genedata. The company's scientists used the firm's Genedata Profiler software platform to process, manage, and analyze the data in the project. To accomplish this, given the diluted nature of tumor DNA fragments, the company's team had to craft new algorithms, which were subsequently integrated into workflows and made available to Genedata Profiler clients.
More specifically, Genedata's Expressionist for Genomic Profiling tool was used to map reads to the human genome, identify regions with tumor-specific methylation patterns, quantify the occurrence of those patterns, and then to calculate relative pattern frequencies per sample, according to the papers.
Pattern frequencies were calculated as number of reads containing the pattern divided by total reads covering the pattern region. The algorithm they subsequently developed scans the whole genome and identifies regions that contain at least ten aligned paired-end reads, the authors stated. The read bundles are then split into smaller regions of interest containing at least 4 CpGs. For each region and sample, the absolute frequency for all observed methylation patterns was then determined.
CEO Othmar Pfannes said that patient stratification and selection is a "key focus" for Genedata, and said the development of the new tools within the context of EpiFemCare show the "very clear value that the project has brought" to the company. He also noted that Genedata has been approached by potential customers that would like to carry out similar analyses in a variety of indications. "It has been a very successful collaboration," he said of EpiFemCare, noting that additional validation of the markers identified is continuing.
"The next step is to validate some of these results in other clinical studies," Pfannes said.