NEW YORK – University Hospital Tübingen in Germany is embarking on a new project, called Ge-Med, that will replace diagnostic exome sequencing with whole-genome sequencing across a variety of genetic conditions, including familial cancer syndromes and rare genetic disorders. The program is also among the first to explore polygenic risk scores (PRS) in a routine clinical setting.
"Polygenic risk scores are not yet part of the diagnostic setting, so we have to learn how to handle them," said Olaf Rieß, a medical geneticist and director of the hospital's Institute of Medical Genetics and Applied Genomics that is spearheading the project.
Ge-Med follows a two-year pilot study of whole-genome sequencing for three indications that recently wrapped up at the institute. It adds to similar efforts elsewhere that are implementing whole-genome sequencing clinically and have plans to return PRS, for example at Stanford Medicine's Center for Inherited Cardiovascular Disease and at institutions led by Weill Cornell Medicine.
The Tübingen institute has been serving as a diagnostic provider for all types of genetic conditions for many years and also performs somatic tumor mutation profiling for cancer patients. Almost five years ago, the lab switched all of its gene sequencing panels, which Rieß said always seemed to be in need of updates, to diagnostic exome sequencing, with the exception of a somatic cancer panel.
A year and a half ago, the institute launched a project called Genome First that looked into the utility of diagnostic whole-genome sequencing. Illumina provided in-kind support for the effort by contributing certain consumables. For that study, Rieß and his colleagues chose three disease indications — retinal disease, childhood cancer syndromes, and developmental delay — and required that patients did not have any previous gene panel or exome sequencing tests. Along with the project, the lab was accredited by DAkkS, Germany's accreditation body, for clinical whole-genome sequencing.
In total, the Tübingen team sequenced about 1,500 genomes, which included several hundred parental genomes from patient-parent trios. While the data analysis for structural variants and copy number variations is still ongoing, and the first publication is pending, the project was a success, Rieß said. For retinal disease, for example, the diagnostic rate was almost 80 percent, which he said was about 10 percent higher than with retinal-specific gene panels alone.
One aspect that set the Genome First project apart from some of its counterparts in the US was that at least part of its costs were carried by health insurance. German public health insurance, which covers the majority of the population, does not yet reimburse for diagnostic exome or genome sequencing, Rieß explained, but it does pay for diagnostic gene panels ordered by a medical geneticist, and it reimbursed the institute at a rate equivalent to such panels. "There was not a single patient where insurance did not reimburse some amount of money," he said.
Overall, consumables and other costs for whole-genome sequencing are still higher than for exome sequencing, he said, but it's easier and overall less expensive to call different types of variants from a single genome assay than having to combine exome sequencing with other assays that require different types of equipment and additional personnel.
The Ge-Med project is building on the Genome First pilot study but is opening up whole-genome sequencing to all genetic indications, including patients who have had other genetic testing in the past. "Any patient who comes to us who has either a familial cancer syndrome or a rare disease, a genetic condition basically, gets the genome independently of whether they had a panel, exome, or whatever before," Rieß said.
German health insurance pays for such additional testing after a year has gone by, he explained, as long as the first test didn't result in a diagnosis. The new project, for which Illumina is covering some of the personnel costs, also no longer sequences the genomes of parents, he added, which would have been too costly in the long run.
Aside from potentially increasing diagnostic yield, and standardizing workflows on a single type of assay, one major reason why the institute is switching to clinical whole-genome sequencing is that it now can, for the first time, report polygenic risk scores (PRS) to patients who elect to receive them. This will build on the experience the Tübingen group has gained with returning secondary findings from its exome sequencing tests, which Rieß said has been well received by patients.
The first PRS to be included in the analysis is for breast cancer and will be reported to patients undergoing breast cancer risk gene testing at the institute. While this normally involves a panel of about a dozen genes, having the whole genome will allow for a PRS to be reported in addition, with the hope that it can explain familial breast cancer risk in some patients who test negative for the high-risk genes.
This summer, the researchers also plan to add a PRS for diabetes, and later this year, another one for cardiovascular disease. Both conditions can potentially be prevented by lifestyle changes, Rieß said, so patients with a high PRS can receive specific recommendations for lowering their risk.
However, implementing PRS reporting has not been trivial. For example, the informed consent needed to be adapted so patients would understand what a high PRS means, and medical specialists had to be trained on their use. There have also been discussions on the cutoff for placing patients in a high-risk group. In addition, there needed to be a management plan for patients found to have a high PRS. "We've been discussing this for a year," Rieß said. "That's the reason we just started now."
Over time, the Tübingen group hopes to replace all diagnostic exome sequencing tests — currently about 3,000 per year — with whole-genome sequencing, he said, and to grow testing to at least 5,000 genomes annually a couple of years from now. For now, only local patients who are seen in Tübingen will receive the polygenic risk scores, though, to ensure proper follow up.
Olivier Elemento, director of the Englander Institute for Precision Medicine at Weill Cornell Medicine, which is also implementing clinical whole-genome sequencing, said that the main challenge with polygenic risk scores is that very few are currently actionable.
For breast cancer PRS, for example, no clear reporting guidelines or clearly defined prevention strategies currently exist for individuals with a high score, he said, though "that will likely change in the near future."
At his institution, he and his colleagues are seeing a lot of interest among clinicians in reporting polygenic risk scores, he added, for example for coronary heart disease, multiple sclerosis, diabetes, and prostate cancer.
"I think it's a very exciting moment for clinical genomics when we can start to apply more fully our longstanding appreciation of the complex genetic architecture of diseases across the variant frequency spectrum," said Euan Ashley, director of the Center for Inherited Cardiovascular Disease at Stanford Medicine, which plans to start reporting a PRS for cardiovascular disease in the near future. "This means that for diseases like coronary artery disease, we can integrate genome-wide PRS with traditional risk factors to improve the accuracy of our prediction and better direct therapies like statins."