CHICAGO – When Genomics plc announced earlier this month that it had completed a $30 million financing round, the Oxford, UK-based genome analysis company said it would apply the proceeds to support growth in precision medicine, specifically what it calls genomic prevention.
The company's cofounder and CEO Peter Donnelly, professor emeritus of statistical science at the University of Oxford, described genomic prevention as the application of polygenic and other risk factors to identify people who are at high risk of certain diseases before they actually develop an ailment. This would then guide clinicians to initiate preventive interventions such as earlier diagnostic screening, lifestyle changes, and prophylactic medications.
"The key idea is to use sophisticated prediction tools, powered by genomics, to identify when people are still healthy [and] identify subsets of individuals who are at high risk," Donnelly said. "Then the medical system in a population health way can work out what to do to reduce the risk, either to prevent disease or to catch it early so outcomes are much better."
He said this contrasts with genomic medicine, which is for people who are already sick. "Genomic prevention … is an opportunity for healthcare systems moving forward to get much better at understanding risk and then working out how to act early to stop disease development."
Underlying that effort — as well as the firm's other focus area of drug and target discovery — is polygenic risk scoring, which gives Genomics a reliable means of quantifying the genetic component of health risk for common conditions such as cardiovascular disease, type 2 diabetes, and certain cancers like breast and prostate cancer.
"We know [that for certain] diseases there are many, many tens or hundreds of thousands of positions in the genome which each contribute to risk, but individually contribute a tiny amount to risk," Donnelly explained. "The idea of a polygenic risk score is, you can identify those and aggregate their effects within an individual."
Because drug discovery and precision health are such massive challenges, Donnelly does not expect the company to expand its scope all that much in the next several years.
"There is a huge appetite [around the world] for a shift in healthcare from treatment to prevention, and we believe very strongly that the kind of tools that we're developing, with genomics at the heart … give us a chance to really do prevention in a clever way," Donnelly said.
Genomics has built an analysis engine that integrates well more than 100 billion data points, helping users match genetic variation to changes in 25,000 different measurements in humans from genome-wide association studies. The company also has access to sequencing data through public sources including the UK Biobank, as well as from its business partners.
The firm's pipeline automates quality control, alignment, and integration of data to support analysis via its algorithms. Donnelly said that the algorithms, which can analyze all 25,000 measurements at once, are "at the core of our company."
The database also contains functional genetic data for drug and target discovery. This includes information on chromatin and regulatory regions, RNA-seq data from single-cell studies, and data from CRISPR screens. "At a high level, it's about using genetics to link changes to biology — perturbations — to outcomes," Donnelly said.
Lately, Genomics has been applying its algorithms to attempt to predict likelihood of COVID-19 infection and severity of the disease in those who are infected, but Donnelly said there still is not enough genetic data to make reliable forecasts. The company is focused more on the genetics of humans who get infected rather than on variants of the SARS-CoV-2 virus, he explained.
Genomics' customer base includes large healthcare systems, pharmaceutical firms, biotech companies, and some national programs like the UK's National Health Service. The company has also worked with Genomics England in the past.
This month, Genomics researchers had a hand in a paper that appeared in the American Journal of Cardiology and another article in Circulation: Genomic and Precision Medicine.
The Circulation paper showed that genetics is as strong a predictor of cardiovascular disease as blood pressure or cholesterol, so the risk of developing heart disease is more accurately predicted when polygenic risk scores are integrated into more traditional predictive models.
"When you do that, you improve performance. You get better at [identifying] the individuals who are at high risk for, in this case, heart disease," Donnelly said.
The article in the AJC looked at a similar integration for predicting coronary artery disease.
In both cases, Genomics worked with clinical partners to bring genetics-based disease prediction into primary care.
"It's really critical that we're doing this in primary care," said Euan Ashley, director of the Center for Inherited Cardiovascular Disease at Stanford Medicine, because so much screening happens during annual physicals and other routine testing ordered by general practitioners.
Ashley and Donnelly were among the authors of both articles, and in January, the center that Ashley runs switched to clinical whole-genome sequencing for all of its in-house gene panel tests, covering different classes of inherited conditions.
Stanford is now ramping up a pilot to generate polygenic risk scores from low-pass cardiovascular sequencing data for patients in primary care. The California institution turned to Genomics to perform that risk scoring.
"When we start to move things onto the clinical side, we really do need enterprise solutions to do enterprise-depth problems," Ashley said, adding that this is a competency Stanford does not possess in house.
Genomics is also running a pilot with the NHS in England's North East region to predict cardiovascular disease from screening in primary care clinics, as discussed in the Circulation paper.
While Genomics calls it genomic prevention, Stanford's name for its pilot is preventive genomics, but Ashley said that the principle is the same.
"Of course, there's been this phenomenal impact of genomics on rare disease through sequencing the genome and sequencing the exome, and that's been very exciting," Ashley said. But his team has been wondering for the decade that Stanford has been involved in clinical genomics when sequencing could start to have an effect on preventive care.
"For almost every year in the last 10 years, I've said, 'not quite,'" Ashley said. "I've changed now. I really, genuinely believe that the time is now and that these scores are valid at a level that we weren't able to ascertain before because we didn't have big enough studies."
Some of the research to date has not included many diverse populations, according to Ashley, but he said that Stanford is working to close that gap and to demonstrate the effect of polygenic risk scores on multiple ethnic and racial groups.
"We need to, one, pay attention to that, two, get more diverse cohorts, and three, actually work on algorithmic approaches to maximize the potential with the data we currently have," Ashley said.
"Just picking up a score and porting it into our population and expecting to perform the same [for everyone], that's not going to happen. Picking up the score, porting it to another population, and making sure it's optimized for that population, that's where we should be today as we grow out the diversity of the cohorts on which we train."
Ashley said that the two recent papers, as well as the ongoing pilots, are important because they integrate polygenic risk scores into regular physician workflows.
Stanford primary care physicians assess patients with an American College of Cardiology screening tool called the Atherosclerotic Cardiovascular Disease Risk Estimator Plus, which screens for 10-year risk of heart attacks based on phenotypic factors such as smoking, diabetes, and cholesterol that come from medical records. But the recently published research shows that integrating polygenic risk into that score makes for a more precise risk estimate.
"We're building [that] into the workflow so that the doctor doesn't do anything different," Ashley said.
Stanford and Genomics are working together to integrate this information into the academic medical center's Epic Systems electronic health record.
The pilot involves as many as 10,000 patients, but Ashley said that the clinical genomics program expects to extend it across the entire Stanford Medicine enterprise starting later this year. The first expansion will be into oncology, followed by specialties where faculty members emerge to champion the effort within their departments.
"I think the value proposition in cardiology is very clear, but I also don't have any doubt that the value proposition will become clear in other areas," Ashley said.
Genomics' Donnelly said that the opportunity for genomics-based preventive care is substantial because of rapidly escalating healthcare costs.
"A key way of addressing [costs] is to move the focus from treating people later on in a disease to either stopping the disease or catching it earlier," Donnelly said. "We believe that genomic prevention will be key to that."
He said that the company is lucky to be in the UK because prevention is part of that country's national genomics strategy.
This type of prevention is about identifying high-risk people "who are currently invisible to the system, and working out how to target resources more effectively to them," Donnelly said. He said that the Circulation paper in particular "showed clearly that combining the genetics component of risk with the current tools that we use routinely substantially improves prediction."
With its new investment, the startup has moved fully into a growth phase, according to Donnelly. Earlier investors Foresight Capital and F-Prime Capital led the new financing round, with participation from Oxford Sciences Innovation and Lansdowne Partners.
Genomics previously raised £33 million ($45.8 million) in a Series B round in late 2018.
One earlier investor is Vertex Pharmaceuticals, which is in the final year of a three-year R&D partnership — with an option to extend for two more years — with Genomics to apply genetics and machine learning to help the pharma company identify new targets for precision medicines. Any novel targets identified through the collaboration will result in Vertex making additional milestone and royalty payments to Genomics, the companies said.
Donnelly declined to say whether Genomics and Vertex would exercise the option.