The 100 000 Genomes ProjectBMJ 2016; 353 doi: https://doi.org/10.1136/bmj.i1757 (Published 13 April 2016) Cite this as: BMJ 2016;353:i1757
- Mark Peplow, freelance journalist, Cambridge
Nobody could deny that Mark Caulfield, chief scientist of Genomics England, has a bold vision. “This will bring genomic medicine across the healthcare system,” he enthuses. “It is, in essence, an NHS transformation programme.”
He’s talking about England’s 100 000 Genomes Project, which is now ramping up into high gear. Overseen by Genomics England, it is one of the biggest whole genome sequencing projects in the world. And it is working to a breathtaking timetable: most of these genomes will be sequenced by the end of next year.
The genetic material will come from patients with rare diseases or common cancers and their families (box 1). By identifying any genetic anomalies, and linking them to participants’ medical histories for the rest of their lives, the project aims to build up a unique database for treatment and research. “It will allow us to find things in the data that we might not notice in ordinary clinical care,” says Caulfield. That should offer better diagnoses and more targeted therapies. It also gives scientists a treasure trove of information that could help to develop more effective drugs.
Box 1: The road to 100 000 genomes
Roughly 25 000 cancer patients will each contribute two genomes: their own and that of their tumour
About 17 000 people with rare diseases, plus two blood relatives of each patient, will add another 50 000 genomes
The project is also sequencing genomes from a smaller number of patients with severe infections
That remit is impressive enough. But the project’s broader goals are to kickstart a national genomics industry and make the UK the first country to routinely use DNA sequencing in mainstream healthcare. “If we get this right, our ambition is to see new treatments, new diagnostics, coming to patients in the UK first,” says Caulfield.
The project is already having clinical impact among people with rare diseases, with the first child participants receiving a genetic diagnosis in January. There are about 7000 known rare diseases, and roughly 1 in 17 people (about three million in the UK) are affected at some point in their lives.1 “Collectively, the burden is high,” says Caulfield. “They are a huge cause of disability, and the toll on individuals is huge.”
More than 80% of rare diseases are suspected to have a genetic component. But their rarity makes diagnosis a huge challenge, which can be time consuming and frustrating for patients and clinicians. “Many of these disorders are totally unique,” says Beverly Searle, chief executive of Unique, the rare chromosome disorder support group (www.rarechromo.co.uk). “That’s why it’s so important to have these projects where you gather large datasets.” With tens of thousands of genomes from patients with rare diseases, it becomes much more likely to find a statistically robust association between genetic variants and a particular disease.
Once that link is established, the experiences of those who share the same genetic anomalies can be compared to predict how a particular patient’s condition might develop in the future and which treatments are likely to be more effective. Not only could this improve clinical outcomes, it could also save time and money. “It’s important that expectations are managed—not every family will get a diagnosis,” cautions Searle. “But we’ve got the potential for one test to give you an answer.”
The other arm of the project focuses on common cancers, including those in the lung, breast, colon, prostate, and ovary, where a genetic diagnosis could affect treatment options. About half of melanomas are caused by a mutation in the BRAF gene, for example, and these can be treated with a drug that specifically targets the BRAF protein. “A mutation can help to predict a medicine’s effectiveness,” says Caulfield.
Collecting and using the data
Most of the project’s participants arrive via one of the 13 NHS Genomic Medicine Centres that were established last year around England. People give a small blood sample, and (if they have cancer) a small piece of their tumour, which can have a substantially different genome.
The project has already sequenced more than 7000 genomes and is recruiting more than 200 patients with rare diseases per week. But the pace of sequencing will quicken in the coming months, when a dedicated facility at the Wellcome Genome Campus in Hinxton, Cambridgeshire, opens. The American company Illumina is setting up a world class sequencing facility there, stuffed with machines that can read an entire genome in about a day, for less than £1000 (box 2). Caulfield hopes that most of the sequences will be completed by the beginning of 2018.
Box 2: Sequencing revolution
A human genome contains about 3.2 billion base pairs of DNA. Most variations are harmless, but sometimes a change to a single base can lead to disease
Previous sequencing efforts have tended to focus only on DNA that codes for proteins. But this ignores more than 95% of non-coding DNA in the genome that may have important functions, such as decreasing the chances of mutations, or switching other genes on or off
The Human Genome Project took 13 years to deliver the first full human genome sequence in 2003, at a cost of more than £2bn. Today, a whole genome sequence costs less than £1000 and takes about a day, depending on the accuracy needed
In the 100 000 Genomes Project, each genome from a patient with a rare disease will be read 30 times and those from cancer patients 75 times
The information will be stored at a secure government data centre in Corsham, Wiltshire, along with each participant’s health records. Any findings that are relevant to the participant’s condition are sent back to their doctors, to help with diagnosis and treatment. Unexpected findings—for example, a genetic mutation that has no immediate health impact, but may affect their children—are also communicated if the patient has stated a wish to receive this information.
This first phase of the project—gathering samples, generating genomes, curating the data, and providing diagnoses where possible—should cost around £200m, funded by government. The second phase, supported until 2020 with an additional government commitment of £250m involves a deeper analysis of the data to tease out associations between genetic variation and disease.
This analysis will be done through the Genomics England Clinical Interpretation Partnership (GeCIP), which already includes more than 2250 researchers from the NHS and universities. Researchers can access the data for free as long as an independent review committee approves their research proposal. A dozen companies have already joined in, too, including GlaxoSmithKline and AstraZeneca, although they have to pay for the privilege: £250 000 for big drug companies.
Caulfield is adamant they are not selling access to the data, merely recouping the project’s costs. But this commercial element has raised concerns. “These are NHS data that can be accessed for a fee, whatever language you use,” says Stephan Beck of University College London’s Cancer Institute.
Indeed, Genomics England itself is a company, owned by the Department of Health. Caulfield says that this structure helps them to operate faster and more flexibly than a conventional academic research consortium. Genomics England will own the intellectual property that the project generates, for example, so it will be easier for companies to negotiate with them, as a single point of contact, than with multiple university research partners. Any profit that Genomics England makes goes back into genomic medicine in the NHS, Caulfield adds.
Acting as a company also helps to demonstrate that the UK is fertile ground for the genomics industry. Illumina, for example, is not simply contracted to carry out the sequencing work—it is a fully fledged research partner, and the project was a key factor in the company’s decision to base its new European headquarters in Cambridge. “It’s about making England the place to bring your genomics project,” says Caulfield. “The ambition is to generate commercial opportunities—not just for us, but for universities and the NHS itself.”
“There’s an aversion to mixing health and commercial interests,” says Mark Sheehan, a bioethicist at the University of Oxford's Ethox Centre, who is involved in an ethical and social science study within GeCIP that will monitor the project and provide ongoing advice. For now, he is maintaining a “cautious scepticism,” but he believes that as long as commercial relationships are handled appropriately and transparently, they don’t necessarily pose ethical problems: “At the end of the day, we need the companies to do the research.”
In the early stages of the project, some had voiced concerns about the risks of patient data being misused or even stolen. Could an insurance company or employer use a patient’s genomic data to deny them coverage or other rights? “We raised issues when the project was first mooted, but so far the way it’s being developed is really good,” says Searle. All of the data are pseudonymised, so individual patients cannot be identified by the researchers analysing the data.
“We’re very honest with patients about this—it’s very hard to guarantee anonymity—but we’re doing everything we can,” says Caulfield. As an additional security measure, researchers will be able to access the data only at the Corsham site: “You can’t take the data away,” say Caulfield firmly.
This, however, raises another problem, says Beck, as it severely restricts the number of researchers able to use the data. “The process is too cumbersome, too slow,” he says. Beck leads the UK arm of the Personal Genome Project, an effort that also has its sights set on sequencing 100 000 genomes from volunteers. The big difference is that all of the information gathered will be open access, so researchers around the world can immediately dive into the database. He acknowledges that there are valid concerns about genetic privacy, but that participants should be able to make their own judgment about the risks involved. “We’re very open about the risks, but people can look at that and balance the risks for themselves,” he says.” Beck notes that in the United States, where the project began in 2005, no participants have had their data used against them. About 10 000 people have registered to take part in the UK, but only 10 sequences have been completed so far.
Caulfield hopes that by 2020, the 100 000 Genomes Project will have instituted a culture change in the NHS, raising the profile of genomic medicine and providing the infrastructure necessary for it to flourish. He believes that in the next five years it will become routine to order a genome sequence for many patients presenting with a rare disease or cancer.
“I hope there will be greater engagement with the possibilities of genomic medicine,” says Christine Patch, a consultant genetic counsellor at Guy’s Hospital in London, who has been recruiting patients with rare diseases to the project. “There will be more tests based on DNA analysis, and clinicians need to understand the power—and the limitations—of the tests,” she says. Genomics England is now working with the NHS to deliver 700 person years of training in genomic medicine in the coming years.
Meanwhile, other UK nations are showing interest in joining the 100 000 Genomes Project, and Caulfield says that the country is well positioned to lead the world in genomic medicine. That’s partly thanks to the NHS, which makes it possible to link the genome to each participant’s medical records over time, something that would be harder to do in the more fragmented US healthcare system. “It gives us a uniquely rich data set,” he says. “Wherever we go, we are the envy of the world—they’re in awe of what we’re doing.”
Competing interests: I have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.
Provenance and peer review: Commissioned; not externally peer reviewed.