Results from a pilot study performed on participants of the 100,000 genomes project confirm that using genome sequencing as part of the diagnostic process could significantly improve the number of rare disease patients who receive a diagnosis.
Writing in the New England Journal of Medicine, the researchers report that 25% of the participants of the pilot received a new diagnosis after whole genome sequencing, 14% of which would have been missed using conventional genetic tests.
As well as helping to reduce long diagnostic odysseys for the families involved, this has the potential to save medical costs in the long run, according to the researchers, by minimizing unnecessary hospital costs through earlier diagnosis and, in some cases, treatment.
“Our findings were important in that they were used to create a National Genomic Test Directory, …Whole genome sequencing is now available and over the next four to five years, the NHS will sequence up to 500,000 whole genomes,” explained Mark Caulfield, Professor of Clinical Pharmacology at Queen Mary University of London and former Chief Scientist at Genomics England, who co-led the study, during a press briefing.
The original 100,000 genomes project began in 2013 and was completed at the end of 2018. Part of the original remit of the project was to provide more information on rare diseases, to help improve diagnoses and treatments, and also some types of cancer and infectious diseases.
“Rare inherited diseases affect about 6% of the UK population… We think there are up to 10,000 different disorders that could be classified as a rare disease, and 80% of those have some sort of genetic basis,” commented Caulfield. “Many of the people who have a rare disease either have very long diagnostic odysseys to get an answer for why they’re like they are, or they do not achieve an answer in their entire lifetime.”
To try and improve this process and really focus in on rare disease, the investigators assessed the value of whole genome sequencing for diagnosis in a group of 4660 participants from the 100,000 genomes project from 2183 families. These individuals had been identified as having 161 different rare disorders during previous medical visits, but their exact diagnoses were unknown and could not be determined using standard testing.
Following sequencing analysis, 25% of the group obtained a confirmed diagnosis. This was even more successful for some types of disorders. The team also found three new disease genes and 19 new associations.
“Across the cohort, hearing disorders, metabolic disorders, intellectual disability, neurological disorders, and various eye diseases had higher rates than that, some are up to 40 to 45% of diagnostic yield,” said Caulfied. “This is beginning to be really successful.”
Although not all cases have achieved a diagnosis, some of the ones that have, have been very successful. For example, following the death of a 4-month year old baby, an analysis revealed that he had suffered from a severe metabolic disorder due to inability to take vitamin B12 inside his cells. When his mother had a new pregnancy, this condition was picked up in his younger sibling at birth and he was saved by weekly B12 injections.
Caulfield explained that not only do these diagnoses help the families and sometimes provide a treatment for the previously unknown condition, they can also save a lot of money spent on unsuccessful hospital visits and treatments.
On average, “the individuals affected by these rare diseases spent 6 years attending 68 hospital appointments prior to diagnosis,” he explained.
“People are in and out of the system the whole time, they’re using a lot of resource. And if we could focus that on getting the diagnosis, and that’s what we hope from this program, then we can use less resource to monitor them and maybe move in some cases to treatment.”
Damian Smedley, a professor specializing in computational genomics at Queen Mary University of London, and co-lead author on the study helped create a program called ‘Exomiser’ that can find possibly pathogenic gene variants from whole-exome or whole-genome data by integrating phenotype information. This program helped improve the number of diagnoses made during the study.
“Exomiser lets us collect detailed clinical phenotype data. We can automate comparing that to previous knowledge about phenotypes that are associated with previous patients. And we can really help use that information to narrow down the variants,” said Smedley, during the briefing.
“We’re looking for a single causative variant in most cases. So, by making use of this phenotype data, in this very sort of controlled and automated way, this really helped us up our diagnostic yield,” he explained.
The more data that can be collected on rare disease phenotypes and genetics the more likely it is that a higher number of diagnoses will be made in the future. The researchers are now feeding the information from this study into the new Genomic Medicine Service in the UK, being run as part of the NHS.
The researchers note that the 75% of participants without a diagnosis are not being forgotten. “Often, the most frequent reason that we suddenly find an answer that we didn’t know maybe two years ago is because the collective international knowledge has improved,” said Rich Scott, clinical geneticist and Chief Medical Officer, Genomics England.
“Even doing the simple thing of just running the reanalysis every three months with exactly the same software means we are likely to discover lots of new diagnoses,” adds Smedley, explaining that tens to hundreds of new disease genes are discovered each year.