RICHARD GOLDBERG

MD

Valar Labs

Professor Emeritus and Adjunct Professor WVU School of Medicine

PLCO (Learn more about this study)

PLCOI-1794

Jan 27, 2025

Identification of Quantitative AI-histologic signatures associated with outcomes in colorectal cancer

In colorectal cancer, adjuvant systemic treatment has been demonstrated to provide a disease free survival benefit after resection. The addition of oxaliplatin to regimens containing fluoropyrimidines alone has provided further benefit, but the benefit in disease free survival 5-years after resection remains modest (MOSAIC trial: 73 vs. 67 %; NSABP C-07: 69% vs. 64%). Platinum treatment is notably also associated with clinically important toxicities, such as neuropathy. As such, biomarkers that successfully identify patients with poor prognosis and those most likely to benefit from oxaliplatin treatment could be clinically valuable in assisting in guiding patients in adjuvant therapy selection in colorectal cancer.

Valar Labs is a start-up company, founded by a team of Stanford researchers at the intersection of artificial intelligence (AI) and medicine. The company aims to construct clinically relevant biomarker signatures of morphologic features extracted from histologic slides using AI that can be useful for treatment decisions. Rather than taking a primarily deep learning approach, the Valar Labs Computation Histology AI (CHAI) Platform involves AI-powered segmentation of cell nuclei on a diagnostic H&E slide and then AI-powered extraction of hundreds of quantitative features describing qualities of the identified tumor and tumor microenvironment (such as geometric descriptors of tumor cell nucleus size/shape and spatial descriptors of immune cells in regard to tumor) that can then be associated with outcomes of interest, with subsequent machine learning used to train a signature of features capable of serving as a biomarker. Our initial work has focused on identifying such biomarkers in pancreatic cancer (PMID: 37044094) and bladder cancer (39383345). Work in bladder cancer has been the basis for a CLIA-approved test to predict benefit from treatment from BCG in non-muscle invasive bladder cancer that is now used in clinical practice.

We propose to utilize the Valar Labs CHAI platform to identify a histologic signature associated with prognosis in colorectal cancer, as well as a signature associated with benefit from oxaliplatin specifically. To do so, we propose to use PLCO data study to construct and validate a histologic signature that stratifies survival outcomes among all colorectal cancer patients with an additional exploratory aim that identifies patients who benefit from receiving adjuvant oxaliplatin therapy. Use of PLCO data would be combined with data from other sources to develop and validate an AI-histologic biomarker associated prognosis and chemotherapy benefit that can be implemented in clinical practice for the benefit of patients.