Paige launches pathology AI for detecting breast cancer metastases in lymph nodes

Digital pathology company Paige has launched a new artificial-intelligence-powered tool for researchers designed to help catch cases of breast cancer that have metastasized to nearby lymph nodes.

The system works alongside a pathologist’s own interpretation of a sample. But working to spot small clumps of interloping tumor cells over a comparatively large area of tissue can be time-consuming without assistance, the former Fierce 15 winner said.

“Accurate detection of breast cancer metastases is paramount for physicians and their patients, but it can be a laborious, manual task for pathologists,” David Klimstra, M.D., Paige’s founder and chief medical officer, said in a statement. 

Ascertaining whether cancer cells from the breast have spread to the lymphatic system can change the course of a patient’s diagnosis and treatment. Paige’s AI program aims to increase the accuracy and efficiency of spotting potential tumors of any size, including micrometastases that might otherwise go unnoticed.

The research-use-only software will become available immediately as part of the Paige Breast Suite, the company said. The lymph node module uses the same underlying AI as the company’s FDA-approved Paige Prostate, which can analyze a broad range of pathology data gathered from different slide-staining techniques and imaging scanners.

Positive indications of metastatic breast cancer are displayed at the individual lymph node level by highlighting all suspicious areas on a slide. Those slides and cases are prioritized for review by the pathologist.

“This product launch is an important step in our overall commercialization strategy as we bring the power of our AI platform to new disease areas,” said CEO Andy Moye, who was named chief this past January after serving for nearly a year as Paige’s chief commercial officer. 

Meanwhile, Paige last month announced a partnership with pathology AI developer Mindpeak, bringing in the German company’s BreastIHC software for distribution through Paige’s platform to help detect, classify and count breast cancer cells.

February also saw a similar distribution partnership with Colorado-based Flagship Biosciences in immuno-oncology for the latter’s software that quantifies PD-L1 expression in immunohistochemistry tissue samples.

Another company in the growing AI pathology field, and one also specifically developing programs for spotting PD-L1 cancer cases, is PathAI. The fellow Fierce 15 winner, which raised $165 million for its efforts last year, previously showed its software could help identify up to 28% additional PD-L1 positive patients that had first been classified as negative by manual scoring, through a retrospective analysis of a clinical trial in patients with melanoma.

PathAI is also developing algorithms for CD8 expression as a cancer biomarker along with programs to aid in diagnosis of nonalcoholic steatohepatitis.

Meanwhile, digital pathology developers Proscia and Visiopharm announced this month that they would team up to provide a single solution combining Visiopharm’s CE marked AI applications for spotting breast cancer biomarkers and metastasis with Proscia’s Concentriq Dx platform for image and workflow management. 

Visiopharm’s AI diagnostics are only available in Europe, while Proscia’s Concentriq Dx can be used in the U.S. for primary diagnosis under the regulatory leeway provided by the FDA during the COVID-19 pandemic for remote pathology programs.