By Allison Proffitt
May 11, 2020 | Adaptive Biotechnologies and Microsoft are working on a new type of COVID-19 diagnostic: measuring the body’s immune system for evidence of a SARS-CoV-2 infection.
There are currently two types of tests for COVID-19: PCR tests indicate the presence of virus from a nose or throat swab, and serology tests measure antibodies in the blood.
“What we’re looking at is a different type of diagnostic, essentially a third type: looking at the T cells, the cellular part of your immune system, to diagnose COVID-19. If you ever were to be exposed to SARS-CoV-2 and infected with COVID-19, your body would mount—thankfully!—an immune response,” Lance Baldo, Chief Medical Officer at Adaptive Biotechnologies told Diagnostics World.
“T cells are often—if not always—the first cells to respond. They respond by identifying that virus and then expanding: reproducing many, many times to try to get ready to kill that virus. This happens really early, usually within a day of getting that SARS-CoV-2 exposure. It’s that T cell receptor sequence and that clonal expansion or multiplying that we ultimately look at to try to identify a diagnostic for COVID-19.”
To gather more data on how the body’s immune system responds to SARS-CoV-2, Adaptive and Microsoft began enrolling a virtual clinical study, ImmuneRACE, to gather samples from 1,000 diverse participants earlier this week. The study is now open to individuals across the country who have COVID-19, have recovered, or have been exposed.
Read more about the study including how to enroll at Clinical Research News.
Adaptive Biotechnologies offers quantitative immunosequencing assays to reveal the breadth and depth of the immune repertoire for diagnosis and cancer tracking. In the ImmuneRACE study, samples will be used to measure changes and markers in immune response that would help indicate who has a SARS-CoV-2 infection now or has had one in the past. The company uses multiplex PCR amplification, high-throughput sequencing, and sophisticated bioinformatics to tease out what an immune response to the SARS-CoV-2 virus looks like specifically.
The study will rely on self-reporting from 1,000 participants, but Baldo isn’t concerned that messy data will muddy results. Adaptive’s bioinformatics platform along with machine learning and AI capabilities from Microsoft will be used to reveal signals within the data.
“Your body doesn’t need to know the difference between what’s SARS-CoV-2 and what’s influenza. But what we can do with our technology and ultimately machine learning, is identify the receptor sequences that seem to be specific to SARS-CoV-2. Then I could look at your blood and say, ‘You don’t have them at high enough levels to be able to say that you had SARS-CoV-2,’… or I could look at your blood and say, ‘Yep, you actually have those T cell receptor sequencer sequences and they’re present in a sufficient amount that we believe you either had the disease or have the disease.’”
Partnerships For Diagnostic Discovery
Since 2018, Adaptive has had a partnership with Microsoft to use machine learning to map the immune system response to many different diseases, including infectious diseases, autoimmune disorders, and cancer. The most developed work so far has been for Adaptive’s immunoSEQ Dx diagnostic assay for Lyme disease. In March, the companies announced an expansion of their existing partnership to study COVID-19.
“We took those existing relationships and the existing ways of working and the existing tools and technologies which were already being developed in infectious disease… We were able to take all those learnings and apply them to what we’re doing with COVID-19,” Baldo said. “The good news from the adaptive immune system is that it doesn’t make a distinction—from a T cell perspective—between bacteria and a virus or anything else. We’re able to use that same approach that we used with Lyme disease with COVID-19. That’s how we were able to put this deal together in about 24 hours.”
Eventually, a diagnostic test for COVID-19 measuring T cells could serve as complementary or alternative diagnostic testing for individuals with known exposures or symptoms, it could help triage patients and inform treatment strategies based on risk, or it could facilitate ongoing immunity surveillance testing of the population to inform decisions on restrictions, the companies theorize.
“One hypothesis is that there will be temporal relationship—a time-based relationship—between these T cell receptors sequences and the time of infection or clearance [recovery],” Baldo said. “Ideally we’d be able to identify someone who had potentially an acute infection vs someone who has cleared infection based on the T cell receptor frequencies and their abundancies at any given time.”
A future test would likely fall under the immunoSEQ Dx umbrella. But “we’re not even thinking commercial things and branding right now,” Baldo said. Step one is gathering the data.
In addition to data collected from ImmuneRACE study participants, Adaptive Biotechnologies and Microsoft are also gathering retrospective COVID-19 data from healthcare organization partners all over the world. These data will all be deidentified, combined, and made freely available on Azure. Data will likely be shared on a modified version of Adaptive’s immunoSEQ Analyzer tool, part of the bioinformatics and visualization platform that supports the immunoSEQ Assay.
Baldo expects the first dataset to be released next month. And the accuracy of the immune response signature will be continually improved and updated online in real time as more study samples are sequenced.
“Our attitude is that whether we have 100 or 400 or 500 or frankly 1,000 subjects sequenced…, time is not on our side here,” Baldo said. “We need to make data available sooner rather than later.”