NEW YORK (GenomeWeb) – A new consortium of researchers, funded with €15.4 million ($18.8 million) from the European Union's Horizon 2020 program, plans to use clinical expertise combined with genomics and other technologies to help thousands of patients with unsolved rare diseases obtain a molecular diagnosis.
The initiative, called Solve-RD, officially kicked off earlier this month and will run for five years. It involves participants from 21 organizations in Germany, the UK, the Netherlands, France, Italy, Spain, Portugal, Belgium, the Czech Republic, and the US.
The project is coordinated by Olaf Riess and Holm Graessner, both at the University of Tübingen in Germany, with co-coordinators Han Brunner from the Radboud University Medical Center in Nijmegen, the Netherlands, and Anthony Brooks from the University of Leicester in the UK.
According to Graessner, an estimated 30 million Europeans suffer from a rare disease, and about 80 percent of these likely have a genetic cause. Of those patients, about half, or 12 million, don't have a molecular diagnosis yet and might thus benefit from the initiative.
The consortium is working closely with four existing European Reference Networks (ERNs) for rare diseases, virtual networks that each focus on a different disease type and bring together healthcare providers across Europe. Last year, the EU launched the first 24 ERNs, which have more than 900 participants from over 300 hospitals in 26 European countries.
The four reference networks — ERN-RND for rare neurological diseases, ERN EURO-NMD for neuromuscular diseases, ERN ITHACA for congenital malformation and rare intellectual disability, and ERN GENTURIS for genetic tumor risk syndromes — will provide Solve-RD with patient data and clinical expertise, in particular exome sequencing data from more than 19,000 patients who remain without a molecular diagnosis.
In addition, Solve-RD has a number of affiliated partners — including six other ERNs and two existing rare disease networks — that will be able to access its infrastructure once it is set up.
Despite the fact that exome and genome sequencing has helped researchers find many novel rare disease genes — often the first step in developing a therapy — the diagnostic yield of a clinical exome test still hovers between 20 and 70 percent, according to Alexander Hoischen, an assistant professor of immunogenomics at Radboud who leads one of the sub-projects of Solve-RD. This means that between 30 and 80 percent of patients who receive such a test don't get a diagnosis from it.
Solve-RD aims to change that. Initially, the consortium plans to reanalyze more than 19,000 patient cases, collected by the four partner ERNs, where exome sequencing did not reveal a disease-causing mutation but where the disease is believed to be genetic in origin. The hope is that a technical reanalysis of the exome data with improved mapping, variant calling, and annotations might yield additional diagnoses.
One of the core tools for this is an exome analysis pipeline established by RD-Connect, another EU-funded project that provides a platform to connect databases, registries, biobanks, and clinical bioinformatics for rare disease research. The consortium also plans to use unified phenotype data for the reanalysis and has a number of phenotype experts involved through the ERNs.
In addition, reanalyzing data from so many patients in concert will increase the odds of finding two cases with defects in the same gene. Up until recently, unsolved exome cases were stored "in their own little silos" at centers across Europe, Hoischen said, but this has now changed. "It could be that one of the reasons we don't find the disease gene or the mutation is that if you only compare the 500 exomes we have here in Nijmegen, we will never have a second case," he said. "But once we join up with 500 cases from Stockholm and 500 from Tübingen, it may just be that we finally find the second patient that has a mutation in the very same gene."
What is still being discussed is how much of the data can be shared outside the consortium. "We have all the disease experts from the ERNs that agreed to share all their data, which is quite unique already. There is still debate whether we can share the data publicly because it is private patient data that is not necessarily consented in such a way," Hoischen said, adding that much of the data could probably be shared in aggregated format.
Overall, Solve-RD hopes that the reanalysis could solve at least 3 to 5 percent of the unsolved exome cases, and could uncover a number of novel disease genes.
In addition to reanalyzing exomes, the consortium plans to explore whether whole-genome sequencing can solve additional cases in selected patients cohorts totaling a few thousand, using both short-read and long-read sequencing technologies.
Genome sequencing delivers large numbers of non-coding variants, which can be hard to interpret. But for diseases that tend to be caused by de novo mutations, such as intellectual disability with no family history, the non-coding variants can be filtered down to smaller numbers by comparing the genomes of patients and their parents. "This is not very new – we have done this for exomes, we have done this for arrays. But we still think that trick may work," Hoischen said.
Long-read genomes might be particularly suitable for diagnosing patients with diseases like ataxia, he added, which are often caused by repeat expansions that exome sequencing and short-read genomes miss. "Whether we're right or wrong about that, the future will tell," he said.
But Solve-RD also plans to move beyond sequencing and employ other omics technologies, depending on the availability of affected tissue samples from patients. For example, the consortium has muscle biopsies for some patients with neuromuscular disease phenotypes that will be analyzed by RNA-seq, using mostly short-read, but in selected cases long-read technologies. For other groups of patients, the consortium wants to employ mass spectrometry-based proteomics, metabolomics, or epigenomics.
For a handful of disorders, dubbed the "unsolvable syndromes," the consortium plans to pull out all the stops. "We and some other centers had the issue that there are some syndromes and rare diseases that are very well defined clinically, but they have remained unsolved throughout the years, despite a lot of work by exome and genome sequencing," Hoischen said. Those syndromes, which include Gomez-Lopez-Hernandez syndrome, Aicardi syndrome, and Hallermann-Streiff syndrome, tend to be clinically recognizable but do not have a genetic cause yet, and the consortium has collected about a dozen cases for each.
"We will just use all the omics tools we have and we will try everything on those to see whether they remain unsolvable or whether we can crack them," Hoischen said. This will include trio long-read sequencing, deep exome sequencing to search for somatic mutations, short-read RNA-seq, epigenetics, metabolomics, and proteomics.
What makes these syndromes attractive, he added, is that they might lead to new paradigms that could help explain other rare diseases. "Maybe we will understand much better how non-coding mutations work, or maybe it's a somatic mutation that we missed so far, or it's something digenic that we didn't understand before," he said.
Sequencing and other omics services for Solve-RD will be solicited through European tenders that will go out in the near future and will be open to both commercial providers and academic facilities. "For the short-read genomes, it will be relatively easy to find well-suited partners," Hoischen said. "Long-read genomes are just slowly coming up, so that may be a little bit more challenging."
In addition to analyzing patient data and samples, Solve-RD plans to establish an online brokerage system that connects researchers who have discovered novel disease genes with scientists who have been studying those genes in model systems and can validate their function. The system will be modeled after an existing Canadian platform, called Rare Diseases: Models and Mechanisms Network. "The link is made through a database where the expertise is being pooled, and then seed funding is provided to the functional group to at least perform initial experiments," Graessner explained.
Finally, the consortium plans to develop a targeted gene sequencing panel, based on molecular inversion probe technology, for up to 50 novel disease genes it discovers in order to be able to rapidly screen thousands of additional patient samples. "We aim to reach out to every clinician connected to the respective ERN to quickly screen all the samples they have," he said. "That will be way beyond the 20,000 patients we currently have."
The ultimate goal of Solve-RD is to come up with new best practices for how to tackle rare diseases in a diagnostic setting. "For this type of rare disease, the best approach may be always to do a genome, while for this rare disease, the best approach may be to do a trio exome, and for yet another disease, it may be slightly different," Hoischen said. "That's exactly what we have to find out – what is the best test."