Following tie-up with Sanofi, transatlantic AI firm secures latest Big Pharma deal in nine-figure Bristol Myers pact
More than six months after securing a nine-figure deal with Sanofi for target discovery, a French AI research group has landed its next Big Pharma partnership.
Owkin, under the leadership of CEO Thomas Clozel, announced Wednesday morning that it’s signed on with Bristol Myers Squibb to optimize the Big Pharma’s clinical trials and improve the odds of clinical success, initially in cardiovascular disease. The announcement comes after a deal Owkin made with Sanofi last November that build on the pair’s previous discovery collaboration.
In Wednesday’s deal, Owkin gets $80 million in a combined upfront and equity payment and potentially further payments higher than $100 million as clinical trials move into later stages. The biotech added in a statement that it plans to use the funds to support its data generation strategy in multiple therapeutic areas, primarily focusing on single-cell omics.
The new collaboration differs from the Sanofi deal. Instead of using its AI platform to improve target discovery efforts and seek better drug combinations, Owkin is purely focused on clinical trial design and optimization.
Clozel said the company is focusing on refining the endpoints in trials, finding the right people and reducing trial “noise.” And as part of its work, there are a few things he wants to keep in mind.
First is finding the right subgroups and defining them with biomarkers. The second step, according to the CEO, is to try and introduce “synthetic control arms,” which he described as more efficient ways to optimize a study’s control arm.
“What we do is try to build a more federated access to data worldwide, and really bring a new way to match patients with multiple modalities, not only in clinical trials, but to improve the way you build for the control arms. That’s the second one,” Clozel said.
The CEO then went on to talk about how improving the control arms can help cut costs within clinical trials, affording the patients involved the opportunity to take a treatment.
Essentially, federated learning works by uploading certain prediction models to these servers where data are kept — in this case, patient data — and bringing the algorithms back to a central location, Clozel said. The models are fine-tuned as needed, and then Owkin uploads the new and improved prediction models back to the servers. This process repeats over and over, and the data remain “locked,” staying in one place.
And with this platform, it would work to identify better endpoint definitions, optimize and identify patient subgroups with distinct biomarkers, and work on estimating drug candidate efficacy.
“What we’re trying to do is really bringing the platform to help with the things that we think can be optimized,” Clozel told Endpoints News.
The third thing that Owkin hopes to accomplish is to diminish the actual rigidity of clinical trials, or the “noise,” using a system called AI-enabled covariate adjustments. Clozel said it can allow for more efficient trials and improve the p-value.
Wednesday’s deal follows three years of collaboration between Owkin and Bristol Myers, according to Clozel, including a number of projects to identify biomarkers and improve clinical trial outcomes using real-world data.
Bristol Myers SVP of global biometrics and data sciences Venkat Sethuraman said in a statement that “We look forward to collaborating with Owkin to extend our applications of AI and machine learning to enhance our discovery and development programs with high-quality, diverse data from clinical and real world sources.”
In its earlier pact with Sanofi, the Big Pharma gave Owkin $90 million just for exclusivity for the first three years of the deal, plus $180 million in an equity investment and more on the line in milestone payments. That deal was focused on four oncology targets: NSCLC, triple negative breast cancer, mesothelioma and multiple myeloma, Owkin told Endpoints at the time.