SAN FRANCISCO (GenomeWeb) – As the UK's 100,000 Genomes Project wraps up this year, the National Health Service is getting ready to implement diagnostic whole-genome sequencing for patients with rare disease and certain cancers in the fall.
Last November, the NHS said it planned to start offering diagnostic WGS routinely in October of 2018. According to Anna Schuh, director of molecular diagnostics in the University of Oxford's department of oncology, this will go along with a "complete restructuring of the NHS genetics labs." Rather than each individual NHS lab having a WGS protocol, sequencing will be centralized and performed at the same Genomics England facility used in the 100,000 Genomes Project. Meantime, other types of genetic diagnostics will be phased out, Schuh said.
Likely, around seven centers will continue to perform conventional testing, with an increasing number of samples sent to Genomics England's Cambridge facility. Genomics England is responsible for choosing the sequencing technology and it recently signed a three-year contract with Illumina, Schuh said.
"With centralization comes logistical issues that still need to be ironed out," she added. For instance, researchers are trying to reduce turnaround times and figure out how much DNA needs to be submitted. For the 100,000 Genomes Project, input DNA requirements were relatively high, but Schuh said for a clinical service, the goal is to reduce requirements to 200 nanograms, which should be technically possible. The impetus for reducing DNA requirements is that often, it is difficult to get sufficient amount of DNA from a biopsy and "for a patient who lives far from the centralized laboratory, you can't just go back and say 'I need more DNA.'"
The goal for turnaround time is 21 days, Schuh said, but that's from the time the sample leaves the local hospital to results being turned. However, she said, there can often be delays on the hospital pathology side after a patient biopsy is taken. That process alone can take up to three weeks, she added.
According to Schuh, the NHS will initially pay for diagnostic whole-genome sequencing for rare disease patients and for cancer patients with acute myeloid leukemia, acute lymphocytic leukemia, and sarcomas, as well as all pediatric cancers.
Schuh said that those will be the first cancer types eligible for testing because cases tend to be more urgent and genomic information is particularly helpful for prognostics and to make treatment decisions. "There's a more urgent need for the acute leukemias," she said. "We're using so many different tests and there are so many abnormalities that are clinically actionable that it makes sense to do this with one test."
Diagnostic sequencing will be rolled out for other cancers later on in a stepwise fashion and eventually be available for most, if not all, cancers. Chronic lymphocytic leukemia and ovarian cancer will likely comprise the second wave, Schuh said. "We have a plan to gradually introduce whole-genome sequencing as a routine diagnostic across all cancers," she said.
In preparation for the rollout of diagnostic WGS for cancer, researchers at the University of Oxford published a clinical validation study of the whole-genome sequencing protocol for blood-based cancers this week in the British Journal of Hematology.
In the study, the researchers evaluated tumor and germline samples from 64 patients with CLL using WGS, and results were compared to orthogonal technologies.
The researchers looked at SNVs and indels in 11 known driver genes for CLL that are already routinely tested in the UK and compared whole-genome sequencing with a targeted NGS panel. They also compared genome-wide copy number alterations discovered by WGS with those detected by fluorescence in situ hybridization and microarrays. Concordance was 96 percent for SNVs and indels compared with targeted NGS, and for copy number alterations it was 87 percent compared with FISH and 93 percent compared with array.
Discordant calls were due to a number of factors. For instance, three variants identified by WGS could not be confirmed by targeted sequencing due to a lack of sample, the authors noted. Another reason for discrepancies was differences in the bioinformatics pipelines, the authors noted, a problem "that urgently calls for stringent standardization of bioinformatics algorithms and filters, not least to improve accurate calling of chromosomal aberrations," they wrote.
Schuh noted that this validation study is the first of many studies that would be published in the upcoming months, validating sequencing for other tumor types. However, such validation studies are just a "small part of the overall work that needs to be done."
Schuh was part of a team that developed and validated clinical NGS panel tests for cancer under the UK's Stratified Medicine Program in 2013. But as the 100,000 Genomes Project got underway, researchers began amassing evidence that for many applications, whole-genome sequencing could offer better results, and last year, began returning some results to participants.
The number of clinically relevant biomarkers for cancer has grown tremendously over the last several years, such that a targeted gene panel is not always able to capture all relevant alterations. As such, Schuh said that for many cancers, multiple tests were being run sequentially, which was not efficient.
"A small panel of genes is a good step in the right direction, but it's not going to give you the full picture," she said. When researchers began looking genome-wide, they discovered markers such as DNA damage signatures, microsatellite instability, and genomic complexity — defined as a genome marked by lots of large chromosomal aberrations. Such genomic changes cannot be assessed via a small gene panel. For hematological malignancies alone, Schuh said, there are now 287 diagnostic tests that can potentially be run on 32 different machines. "And then a lot of the time you don't get the results because you run out of DNA," she added. Moving to WGS will enable the total number of tests to be reduced. She added that the NHS test directory would be reviewed annually to determine which tests should be phased out.
With diagnostic sequencing, patients will also have the option of receiving actionable germline results. As part of the 100,000 Genomes Project, the researchers honed their consent process to make it both comprehensive but not overly burdensome, Schuh said.
Researchers are also currently conducting cost-effectiveness studies. Such studies have proven to be difficult, she said, because "they require a systems approach rather than just some micro-costing of sequencing consumables." For diseases like AML and ALL, the studies have shown that WGS is cost effective, since there are currently so many different tests that are run and that could all be replaced with WGS. "Then there is a group of diseases where a large panel currently gives all the clinically actionable mutations, but might require additional tests in the future, so here the WGS would be implemented at a later stage," she said.
Another issue that the researchers will have to address is data storage. Right now, Schuh said, the genomic data is being stored indefinitely in a data center under the same security measures as data stored at the UK Ministry of Defense. The data is very secure, she said, but it's not in the cloud, which adds some challenges.
Longer term, Schuh said, implementing WGS will also help further drive drug development, and keeping genomic data will be a key piece in spurring those discoveries. "We think that the provision of WGS for patients recruited into clinical trials and long-term storage of the genomic and clinical data is future-proof for biomarker discovery," she said. In addition, because Genomics England is owned by the UK Department of Health, revenue from providing pharmaceutical companies access to genomic data would flow back to the NHS.