Proscia and Columbia team up on pathology AI biomarkers for dementia

Proscia has signed a collaboration with Columbia University to find and validate tissue biomarkers for neurodegenerative diseases, including Alzheimer's and other dementias. The Philadelphia company announced the agreement on 27 October, and it was picked up by Tissuepathology.com the following day.

What the work involves

Columbia brings its pathology department and its digital slide repository. Proscia brings its pathology AI expertise and its Concentriq software platform, which will underpin the project. As well as looking for new biomarkers, the partners plan to check established ones. These may include semi-quantitative scores read by pathologists and composite scores generated automatically from several stains on neighbouring tissue sections that have been aligned digitally.

Get The Sequence every TuesdayThree-minute read. Free. One-click unsubscribe.

"Ultimately, we hope that the collaboration will drive new discoveries that link biomarker and therapeutic development." Kevin L. Gardner, MD, PhD, Chair of the Department of Pathology and Cell Biology, Columbia University Vagelos College of Physicians and Surgeons

Proscia sees a commercial route for what comes out of it. Its Aperture network gives it a route to further work with drug developers, and the company says it is trusted by 16 of the top 20 pharmaceutical companies.

"Columbia is world-renowned for its pathology leadership, research, and patient care," Nathan Buchbinder, Chief Strategy Officer, Proscia

The Agilent backdrop

The deal lands while Proscia is in the middle of a change of ownership. On 13 October, Agilent agreed to buy Proscia for $240 million in cash at closing, plus up to $30 million more tied to future performance milestones. Completion is expected early in Agilent's 2027 fiscal year, after which Proscia will sit in Agilent's Life Sciences and Diagnostics Markets Group. The two have worked together under a distribution agreement since September 2022.

At the time of the deal, Proscia said it served more than 4,000 laboratories, hospitals and clinics, with trailing twelve-month revenue above $18 million. Its Concentriq AP-Dx product holds FDA clearance for primary diagnosis.

The bigger picture

Most of the money and attention in pathology AI has gone into oncology, where tissue biopsies are routine and drug developers already lean on biomarkers. Neurodegeneration is harder: much of the tissue evidence comes from research archives rather than everyday clinical samples. Academic partnerships with large slide collections are one way to build the datasets that pharma will need as more dementia drugs come with biomarker questions attached.

What it means for hiring

Pathology AI companies selling into pharma need people who can translate between pathologists, data scientists and drug developers. In our view, scientific business development hires with neuroscience as well as oncology experience will be in demand.
Ownership changes such as Agilent's pending deal often trigger a rethink of sales and partnership roles. Leaders at independent digital pathology vendors may see more approaches as buyers and rivals adjust.
Academic collaborations of this kind rely on computational pathologists and image analysis scientists, a small talent pool that companies across digital pathology are competing for.

Share Ask Eric about this →