Researchers shatter the speed record for diagnosing rare genetic diseases with DNA sequencing

About a year ago, Matthew Kunzman’s heart was failing, despite doctors’ best attempts to bolster it with every pump and gadget they could think of. But the 14-year-old has bounced back in large part due to super-speedy genetic sequencing that pinpointed the cause of his disease and helped doctors decide how to treat it — in just 11 and a half hours.

That speedy diagnosis — faster than any other medical team has previously reported — resulted from a new approach to DNA sequencing to help patients with deadly and rare diseases. On Wednesday, a team of Stanford researchers and collaborators published a letter in the New England Journal of Medicine reporting that they had sequenced 12 seriously ill patients and successfully diagnosed five of them (including Matthew). In all five cases, the information led to tangible changes in how patients were treated.

Typical turnaround time for diagnosis was around eight hours and as short as seven hours and eighteen minutes — less than half the current record. And the scientists are convinced they can cut that in half yet again. Such speed could be lifesaving for critically ill patients, according to Euan Ashley, a Stanford cardiologist and the study’s senior author.

“You can not only make care better, and help patients more, but do it cheaper, save money, save the system money,” Ashley said. “It seems like a win, win, win all around.”

There’s a lot to be learned by exploring your genetic code, which influences everything from your height and eye color to your likelihood of developing certain diseases. For doctors, knowing whether a patient’s symptoms are linked to specific DNA mutations — and, if so, which ones — can help them determine what treatments and surgical procedures to try and which ones to avoid.

But it typically takes weeks to run, process, and interpret sequencing results. That’s time some patients don’t have. And hospital stays spent chasing down the cause of an unknown disease can cost tens of thousands of dollars.

Ashley wanted to see how quickly he could speed things up. He and his team enrolled a dozen seriously ill patients admitted at Stanford, taking about half a teaspoon of blood from each of them for genetic sequencing. The participants, who ranged in age from 3 months to 57 years old, suffered from everything from seizures to cardiac arrest.

Throughout the six-month study, which kicked off in December 2020, researchers tweaked nearly every step of the sequencing process, from having someone run samples from the hospital to the lab to shortening the time needed to prep DNA for sequencing.

It was round-the-clock work. Ashley remembers walking into his lab one morning and seeing the sleeping bag of a lab member who’d stayed up to run samples all night long.

One of the main time-savers was that the team was sequencing up to 48 DNA samples from a single patient simultaneously, which allowed researchers to generate a whopping 200 gigabases of data within two hours. To do so, they relied on Oxford Nanopore’s sequencing technology, which can read tens of thousands of DNA letters at a time rather than sequencing teeny bits and stitching those small pieces back together into a whole genome, which is currently the predominant method.

“This is really starting to show the Oxford platform in a clinical setting,” said Shawn Baker, a sequencing consultant with more than 20 years of genomics experience, including a 12-year stint with sequencing-industry giant Illumina’s research and product marketing divisions. “And that’s really what’s been missing, up until very recently.”

But all that data left the researchers with a new challenge — quickly processing it. They turned to cloud computing to analyze raw sequencing data and spit out a recognizable string of A’s, T’s, G’s and C’s. They then used computing tools to scan those sequences for mutations in genes that could explain a patient’s symptoms, returning a list of about 20 to 30 candidate genes for a three-person team of genetics experts to review. All told, it typically took around eight hours to go from blood draw to diagnosis.

The Kunzman family experienced that process firsthand in January 2021. In the beginning, Matthew felt weak and short of breath and had a fever of around 102 degrees Fahrenheit. His parents figured he probably had Covid-19 or a flu, but a chest X-ray and additional tests revealed that not only was his heart enlarged, it was failing.

“It was horrible,” said his father, Matthew Kunzman Sr. “I served in Iraq. And this level of stress when your child’s life is at stake is way worse than anything I experienced overseas.”

Matthew was airlifted from Oregon to Stanford, where his condition worsened at first. But his parents consented to have researchers test Matthew’s blood. They were told that if his heart problems were caused by myocarditis, an inflammation of the heart muscle triggered by an immune attack, there might be drugs to help reverse his condition. But if the issue was genetic, he’d need a heart transplant.

Researchers learned Matthew’s issues were genetic, and he was quickly placed on the list for a transplant, which he received in March. He now says that he has mostly recovered —even if he’s not back to the 9-mile hikes he used to do with his family. His parents, meanwhile, want to get his three older brothers tested to see whether they have the same mutation.

Other medical systems have reported similar success stories in the past, namely Rady Children’s Genomics Institute in San Diego, which has sequenced thousands of sick infants from nearly 80 hospitals across North America. The institute held the previous record for diagnosis time at around 14 hours, though president and CEO Stephen Kingsmore was happy to hear that it’s been broken.

“Hearty congratulations,” he said. “We’re going to be hot on their heels.”

Kingsmore pointed out that the sequencing and data processing costs reported in the new study, which range from around $5,000 to $7,300 per sample, are higher than Rady’s approach, which uses Illumina technology. And he said he would have liked to see the authors test the accuracy of their method by running a sample with an already known sequence, a standard way to determine error rate.

Going forward, the Stanford team is working on reducing the total amount of sequencing data that it needs to collect, which would further speed up the process. And the plan is to soon offer rapid whole genome sequencing to more Stanford ICU patients, with diagnoses delivered in less than 10 hours.

But while that may be good news for Stanford patients, Ashley recognizes that plenty of patients and families don’t have access to a cutting-edge medical center, and that these technologies aren’t cheap. He said insurers are moving too slowly to cover genetic sequencing, despite studies from Rady and other places showing that the health care savings from whole genome sequencing outweigh its costs. Along similar lines, a U.K. study published in the New England Journal of Medicine in November found that many of the rare diseases spotted by whole genome sequencing would have otherwise gone undetected with standard, narrower genetic tests.

“I would love to see that we’ve moved beyond that moment where we’re still sitting on the phone and justifying every single test (to) where the test is ordered like other tests that we’ve known for years are both effective and cost effective,” Ashley said. “I really look forward to that.”