Novartis has teamed up with the University of Oxford to use artificial intelligence to improve drug development. The collaboration will analyze MRI data from 35,000 multiple sclerosis patients and results from IL-17 inhibitor trials to better understand diseases important to Novartis’ prospects.
In the year since Vas Narasimhan took over as CEO of Novartis, the Swiss pharma has tied its future to digital technologies and data science to a degree that sets it apart from its peers. However, while in theory the combination of huge data sets and powerful computers could unlock drug development insights, in practice there is some way to go before the techno-optimist vision of R&D comes to pass.
Novartis’ five-year collaboration with the University of Oxford’s Big Data Institute is intended to move that vision closer to reality. The partners have ambitions to change how large data sets are merged and analyzed with a view to deriving insights that improve drug development and patient care.
“These technologies have the ability to spot disease patterns and signals much earlier in their manifestation,” Mark Toms, chief scientific officer of Novartis UK, said via email. “The ability to spot patterns within diseases, and detect commonalities across different diseases, may predict how patients will respond to new and existing medicines earlier than is currently possible.”
In the long term, Toms thinks the initiative will “be transformative for how Novartis clinical trials are designed and conducted” and could uncover “new opportunities and targets” for drug development. The early gains are expected to center on the speed and accuracy with which diseases are diagnosed.
Novartis’ efforts to realize these benefits will initially focus on two programs. For one program, the partners have gathered data on more than 35,000 MS patients that they hope will reveal patterns and timings of disease progression. Equipped with a clearer picture of the MS disease continuum, Novartis thinks it can design more efficient clinical trials.
The second flagship program will look at clinical trial data from 11,000 participants in Novartis’ IL-17 development program. Novartis won the first FDA approval for an IL-17A antagonist, Cosentyx, in 2015 and picked up the rights to IL-17C inhibitor MOR106 from Galapagos and MorphoSys last year.
Working with Oxford scientists, Novartis will apply deep learning algorithms to imaging, proteomic, genomic and other types of data generated by its development teams. Novartis thinks the data may hide insights that enable predictions of early responses and otherwise improve outcomes across a range of diseases, including psoriasis and rheumatoid arthritis.
For now, as is true with much of the AI field, it is unclear whether the approach can live up to the optimism of its advocates. But with Novartis having access to extensive data, and the university plus its neighboring collaborators possessing leading-edge data skills, the partnership is as well placed as any to push at the limits of the science.