By Allison Proffitt
October 21, 2021 | Most rare diseases have a genetic component, yet most people with rare diseases remain undiagnosed, Thomas Defay, Alexion Pharmaceuticals, told the audience at the Bio-IT World Conference & Expo last month. Thankfully, more and more patients with rare diseases are getting sequenced.
Now is the time, Defay said, for measurable improvement in rare disease diagnoses. More genetic data increases clinical suspicion for a disease, genetic variants are more accurately characterized, newborn screening by whole genome sequencing appears poised to expand, and—most importantly—the time and cost of whole genome sequencing has dramatically decreased.
Alexion is committed to improving genetic support for the diseases it treats, Defay said, in three ways. First, by better capturing the existing knowledge around the genetics of disease, an area he conceded that Alexion has not always prioritized. They plan to maximize the existing literature around the genetics of disease, seek both broad and deep coverage, and consistently apply ACMG criteria.
Second, they hope to equip healthcare providers to respond to and use genetic findings, reviewing evidence for ACMG criteria and clearly summarizing and presenting supporting evidence to doctors. “These diagnostic efforts need to be done by healthcare providers across the nation and across the world; they’re not just in the most advanced labs,” Defay said. “Being able to interact with that data effectively is critical.”
Third, Alexion wants to add to the literature of knowledge around rare disease by targeted experimentation, work that Defay said represents a “significant investment” by the company.
Genomics Mastermind
At the Bio-IT World conference, Defay focused on the first two goals and highlighted Genomenon’s Mastermind product as the essential tool helping the company better capture existing scientific knowledge and then communicate the most relevant data to healthcare providers.
Alexion has partnered with Genomenon to build out curated dashboards specific to some of Alexion’s genetic diseases of interest.
Mastermind curates eight million full text genomic articles and 2.9 million supplemental datasets. “The supplemental datasets are important here,” Defay highlighted, “because a lot of the data isn’t in the article itself, but in the materials that come with the article.” Mastermind identifies variants found within the articles, using the findings to, “then decorate the genomic landscape for that disease with those articles,” he said.
Defay compared the Mastermind coverage to ClinVar. Mastermind covers more 19,864 genes and has data on 9.6 million variants. ClinVar contains data on 11,325 genes and 929,000 variants. “Now I don’t want to disparage ClinVar. I think ClinVar is an amazing resource. It’s been very valuable and will continue to be valuable for the community at large,” he said. “But this is really an opportunity to expand what you can do in terms of characterizing the variants that are out there actually in the literature.”
Researchers using Mastermind’s free search engine can search by variant name and receive a list of publications mentioning the variant. That free functionality has brought 12,000 users in 1,000 diagnostics labs in 120 countries to the free Mastermind search engine.
And it was that breadth of users, Defay explained, that convinced Alexion to do further development with Genomenon of the Mastermind platform. Alexion hoped to use the foundational Mastermind capabilities to develop a few specific goals: unbiased and consistent application of the ACMG criteria, data curated for use by healthcare providers and testing centers, and clearly summarizing and presenting genetic evidence to healthcare providers and labs.
The big idea, Defay said, is to use an AI curation engine in conjunction with human expert review to create a resource that is available to everyone—deeper curation of genes available to everyone who is using the Mastermind application.
Wilson Plan
One of the diseases that Alexion has focused on with Mastermind is Wilson disease, a rare recessive disorder known to involve the copper-transporting protein coded by the ATP7B gene, that generally presents in childhood or early adulthood. It is believed to be moderately penetrant, though Defay believes it is more common than we think thanks to misdiagnosis.
Alexion’s goal, he explained, was to speed diagnoses to reach patients before they progress to neurologic or hepatic disease. Using Mastermind, they explored the variants in the ATP7B gene and identified 537 variants from the published literature not available in ClinVar.
“To me this is a really big deal. This changes the game from a genetics point of view because you’re being able to start capturing enough genetic variants that you can reasonably increase the clinical suspicion for a high proportion of patients with Wilson Disease,” Defay said.
The next step was creating a dashboard for physicians offering clinical grade interpretation of the variants while allowing doctors to drill down into evidence and explore for themselves. Doctors can be alerted if evidence or interpretation on a particular variant changes.
By using the Mastermind platform, Defay reports “significant increase in patient identification” and he believes the improved annotation will help diagnose even more patients. Detailed, curated views like this—for Wilson and other diseases—will soon be made available to all Mastermind’s free version users, Defay said, “maybe as soon as November.”