TThe Food and Drug Administration issued an emergency authorization on Friday for a new test to detect Covid-19 infections – one that stands out from the hundreds already authorized.
Unlike tests that detect fragments of SARS-CoV-2 or antibodies against it, the new test, called T-Detect COVID, looks for signs of previous infections in the body’s adaptive immune system – in particular, the T cells that help the body to remember what your viral enemies are like. Developed by Adaptive Biotechnologies, based in Seattle, it is the first test of its kind.
Adaptive’s approach involves mapping antigens to their corresponding receptors on the surface of T cells. They and other researchers had previously shown that the mold of T cells floating in an individual’s blood reflects the diseases they encountered, in many cases years later. The next step is to try to unlock this information to help diagnose these past infections.
advertising
This challenge is extremely heavy on data. “When you think about the patient level, we are looking at 300,000 to 400,000 T cells on average,” said Lance Baldo, medical director at Adaptive. “When you look at the population level, we see hundreds of millions and, finally, billions of T cells. So it ends up being a problem of scale of the web. “
Type Microsoft. In 2018, Adaptive developed a partnership with its giant technology neighbor to build the cloud infrastructure and machine learning models needed to handle these reams of data – in particular, to build a complete map of which T cells bind which antigens.
advertising
“Microsoft wants and wants to enter the healthcare field,” said Baldo. “Adaptive needed experience in cloud computing, machine learning and AI. So it was a pretty ideal fit. ”Teams from both companies worked together one day a week, at Adaptive’s Seattle office or at Microsoft in Redmond.
When the virus started to pick up speed, they quickly directed a large part of this team to work on Covid-19. In June, they were able to access blood samples from people who had been infected with the coronavirus and sequence the genomes of the T cell receptors present. Then they could compare that data set with their control group – the T cell receptor sequence database they worked on for years – and in two months, they collected enough data to publish their first results.
The machine learning models needed to develop the T-Detect test, in the end, were relatively simple. “For me, this is really a big advantage,” said Jonathan Carlson, senior director of immunomics at Microsoft and leader of the partnership with Adaptive. “It is a viral infection and leads to a violent T cell response, and it turns out that you can find exactly the same T cell receptors in many people. And that allows you to use a very simple statistical approach. ”The test reported a sensitivity of 97.1% and specificity of 100%.
The USA issued by the FDA is a reflection of that first approach – but it is not the end of the test’s evolution. “When we filed with the FDA, we did something called a ‘classifier block’,” said Baldo, the algorithm that determines whether the T cell receptors in a blood sample say “Yes Covid” or “No Covid”.
These T-cell responses, however, may vary depending on the version of the virus to which you are exposed.
“We have already discussed this with the FDA,” said Baldo. “You have mutations and other variants coming.” Therefore, Adaptive and Microsoft continue to improve the classifier. “Models get better often,” said Carlson. “Weekly, monthly.” The question that remains, then, is: “when is it better enough? That’s where Adaptive really spends a lot of time thinking. “
At some point, when the test reaches a new limit on sensitivity and specificity, they plan to present a second version of the test for FDA review.
This is just one of Adaptive’s three areas of focus in the coming months, said Baldo. “One of the pillars is to improve the current algorithm and ensure that we continue to have a great test, as the virus continues to mutate,” he said. The second is to direct the company’s T-cell experience to other issues surrounding Covid-19, including the impacts of long Covid and the effectiveness and durability of the immune response elicited by different vaccines.
The third is to continue his work on other diagnoses, for diseases such as celiac disease and multiple sclerosis. Before the pandemic, the company was focused on developing a proof-of-concept diagnosis for Lyme disease, which was announced in November 2019.
This distributed focus will force the company to continue building not only its biological capabilities, but also its machine learning approaches. Although his approach to Covid-19 screening is relatively straightforward, said Carlson, “I don’t expect it to work for all diseases.”