Sequencing whole genomes helps diagnose far more rare diseases, study shows

When the U.K.’s National Health Service started to use whole genome sequencing, doctors were able to determine diagnoses for more people with rare diseases — including some for whom other genetic tests had failed to turn up an answer, researchers reported Wednesday.

The new paper, published in the New England Journal of Medicine, described the results of a pilot study from what’s known as the 100,000 Genomes Project. Whole genome sequencing led to diagnoses for 25% of the thousands of participants in the study thought to have a rare genetic condition — 14% of whom would not have been diagnosed by different methods, the scientists said.

The study echoes what other research has found about the power of whole genome sequencing, and in this case demonstrated its utility across a wide variety of genetic disorders. It adds to the evidence that, in some cases, sequencing a person’s whole genome is required to identify their condition, as opposed to just reading certain genes or key portions of their DNA.

“What’s novel here is that they took a very large population of rare disease cases across many different specialities,” said Heidi Rehm, the chief genomics officer at Massachusetts General Hospital and a researcher at the Broad Institute, who was not involved in the new study. “This supports the notion that genomic approaches are probably the best approach to tackle rare disease.”

The 100,000 Genomes Project, started in 2013, was intended to sequence that many whole genomes from patients as the NHS built out a genomic medicine service. The project also has arms looking at cancer and infections.

Even though this pilot study focused on rare diseases was just published, the NHS has been using whole genome sequencing for certain patients for some time, linking clinical information about their conditions with their genetic information to identify the roots of their disorders. The first diagnoses from the project came in 2015.

In the United States, insurance companies often balk at covering whole genome sequencing, arguing that looking at someone’s unabridged genetic information will uncover too many variations in a person’s DNA whose meaning researchers don’t understand — variants of uncertain significance, as they’re called — and won’t help narrow down the actual cause of the person’s condition. Instead, doctors can often only test a selected panel of genes, or turn to a test that reads the portion of the genome that codes for proteins, called exome sequencing.

But in the U.K. pilot study, 14% of the diagnoses wouldn’t have been uncovered with tests like exome sequencing. Instead, the disease-causing mutations were found in, say, the stretches of the genome that don’t encode proteins or the bits of DNA that reside in organelles called mitochondria.

“That’s really compelling data that we need genomic approaches,” Rehm said. Panel tests are cheaper and faster, she said, but their success depends on doctors choosing the right genes.

The leaders of the study framed genome sequencing as an intervention with advantages both for individual patients and for health systems. One in four patients who received a diagnosis through the study saw an immediate benefit, including being matched to treatment or enrolling in a clinical trial. The researchers said that providing patients with firm diagnoses — and shortening the often years-long “diagnostic odyssey” that many people with rare disease have to embark on to identify their conditions — can help tailor their treatments and avoid unnecessary tests and appointments. Whole genome sequencing now costs just a few hundred pounds, the researchers said.

“Publishing this today gives a platform for the rest of the world to adopt this,” said Sir Mark Caulfield, one of the authors of the paper and the former chief scientist at Genomics England, who is now at Queen Mary University of London. “We’ve obviously run for the finish line on NHS adoption, but it’s our duty to get this out there so others can use this to bring these benefits to everyone across the world with rare disease.”

The new study also underscored the complexity of certain genetic conditions — and how much researchers still have to uncover about the connections between our genes and our health.

While whole genome sequencing provided diagnoses — what researchers call the diagnostic yield — in 25% of cases overall, the yield reached 35% when the cause was thought to be tied to just one gene. The yield was only 11% if the condition was likely rooted in multiple genes. Similarly, 40% to 55% of types of hearing, vision, or intellectual disorders were diagnosed, but the rates were much lower for gastroenterologic or rheumatological disorders.

The researchers also had better luck diagnosing conditions in people who had more family members participating, which allowed the scientists to identify the genetic differences among them that pointed to the cause of the person’s disorder.

The pilot study included 2,183 people with rare disorders — both children and adults — as well as 2,477 family members. The vast majority — 88% — were of European ancestry, with people of South Asian ancestry accounting for 7% of participants and other racial and ethnic groups making up smaller segments.

The 25% overall diagnostic yield for participants in the study also fits with what other studies have found. You might think that reading a person’s whole genome could spit out a diagnosis at higher rates, but the fact is researchers still have much to learn about the role of certain genes in health, how genes influence each other, and how multiple genes can contribute to disease — particularly those in the stretches of DNA that haven’t been given much attention until recently.

Researchers hope that building out sequencing data from more people in tandem with reviewing their clinical information can help steer them to more connections between genetic variants and disease. And with new technology, scientists are getting better at whittling down all the genetic noise produced by looking at someone’s whole genome to home in on what’s causing these conditions.

In one example from the study, Caulfield described a patient who had 6 million genetic variants in the 3.3 billion base pairs of their DNA that diverged from a baseline genome. Of those variants, 677,000 were considered rare, 2,826 changed a protein, 67 were different from the person’s parents’ DNA — and one was the genetic cause of the person’s condition.