NEW YORK – With a recently signed multiyear deal with Teva Pharmaceuticals, immune profiling firm Immunai is targeting its technology to earlier points in the drug development process.
This move to earlier-stage work could heighten the impact of the company's platform, said Immunai CEO Noam Solomon, suggesting that the shift reflects growing confidence in its utility among the company's pharma customers.
The Teva deal, financial details of which were not disclosed, represents a move for Immunai beyond oncology into autoimmune disease. It follows a multiyear partnership with AstraZeneca the company announced in September that focuses on clinical trial optimization. Immunai will receive $18 million for the initial phase of that collaboration.
Founded in 2018, New York City-based Immunai uses multiomic analysis of immune cells combined with AI modeling to aid in drug development, helping pharma firms with tasks like optimizing selection of patients for clinical trials or validating a drug's mechanism of action.
At the core of the company's offerings is its Annotated Multiomic Immune Cell Atlas (AMICA), a collection of multiomic immune cell profiles the company has been generating using a variety of tools including single-cell RNA-seq, CITE-seq, ATAQ-seq, proteomics, and spatial transcriptomics.
AMICA currently contains data on roughly 100 million cells, Solomon said. The company collects these data primarily through academic collaborations with universities and hospitals around the world, as well as through around 40 industry collaborations with pharma and biotech firms.
Another key source for AMICA is publicly available single-cell sequencing data. In 2021, Immunai acquired Swiss bioinformatics firm Nebion, whose technology enables "curating and ingesting all the public domain data on single-cell sequencing," Solomon said. "Every [single-cell sequencing] work that has been published, the data is going to be within AMICA."
The size of the AMICA resource is more than doubling each year, and the company hopes to have data on around a billion cells in the atlas within three years, including "on the order of 50,000 to 100,000 patient samples with single-cell resolution at two different time points," he said.
Beyond sheer volume, the company is also aiming to improve the quality of the data in AMICA, meaning more time points and more complete clinical metadata for the patient samples in the database. "It's not just a human cell atlas as much as a patient atlas that is measured consistently across different indications, different drugs," Solomon said.
Using the data in AMICA, Immunai develops AI-based models for addressing various questions in drug development.
To date, the company's pharma collaborations have largely focused on questions arising later in the development and clinical trial process.
"What we've been doing primarily until now has been to work on advanced or late-stage clinical programs to optimize things like dose-selection justification … optimizing the right patient stratification, the right combination agents for the drugs," he said.
The Teva deal, which Immunai said is worth in the millions of dollars, is significant in that the company will be applying its predictions and insights earlier in the development process.
"Not only Phase II and III but also Phase I and even pre-IND approval," Solomon said. He suggested that this signals increased confidence in the company's tools, given the large impact of decisions made early in the development process.
"Over time, we have built trust with our partners," he said. "We created more case studies, more success stories, and so I think the skepticism level has been reduced."
The company has deliberately pursued earlier-stage work where it could to demonstrate the validity of its platform, he noted, even though in some cases it had to provide that work at a discount.
"It was a worthwhile investment because now we have a few examples where we predicted certain things … and the results held," Solomon said. "I think that is a big demonstration of the platform."
He noted that the results the company has generated since its launch have helped demonstrate the value of approaching disease and drug development from an immune-centric perspective.
"Most companies that are [doing] drug development or discovery are studying the disease," he said. "They are trying to find mutations in the diseased tissue and find different expression compared to healthy tissue," which differs from disease to disease.
The immune system, on the other hand, remains the same system across diseases, which suggests a resource like AMICA could be useful for modeling across indications — a notion that has been borne out in the company's work.
Another key finding has been that the company is able to use the large size of the AMICA dataset to make predictions in much smaller cohorts.
"When you work with a pharma partner on a clinical cohort, usually they will give you 40, 60, 100 patient samples," Solomon said. "And they are going to ask you to make predictions or give insights on this cohort. But everyone working with AI knows that even a 100-patient cohort is a small cohort."
"The question was, could you really leverage the samples in AMICA to improve the prediction that you could make in a small cohort in one clinical trial," he said. "The answer is 'yes.' You can significantly improve those insights, even if they are not from the same drug, even if they are not from the same disease."
In an email, Eran Harary, senior VP and global head of early clinical development at Teva, noted the evidence Immunai offered that supported its platforms' capabilities.
"Through the evaluation process, our scientific leads were impressed with the data and capabilities that Immunai presented in their previous case studies and success stories," he said.
Harary added that while the collaboration only recently started and data is still preliminary, "Immunai was already able to generate rich single-cell data coming from very rare cells essential to our understanding of mechanism of action of our investigational drugs. This capability is quite sophisticated, compared to the state of the art."
Under the collaboration, Immunai and Teva will be working on several therapeutic programs in both immunology and immuno-oncology that are in or about to go into clinical development. Harary added that Immunai will be helping with questions including "understanding and refining the investigational drug's mechanism of action in patients, determining the optimal dose and treatment schedule, and identifying patient populations most likely to respond favorably."
He highlighted the uniqueness of the AMICA offering, noting that it is "something that other companies, including Teva, do not have. We believe that their platform will allow Teva to address [drug development questions] in a way we may not have been able to answer before."