io9, a software company developing an AI platform for precision oncology that enables the correct front-line targeted therapy at diagnosis, today announced the publication of a positive validation study for OncoGaze™. OncoGaze™ is an AI-biomarker platform to identify clinically actionable biomarkers, eliminating the need for costly and time-consuming next-generation sequencing (NGS) for first-line treatment. The study, published in the Journal of Clinical Oncology, a journal of the American Society of Clinical Oncology (ASCO), reported that DeepHRD, the first biomarker test in the OncoGaze™ AI platform, outperformed FDA-approved homologous recombination deficiency (HRD) companion diagnostics in identifying patients with high-grade serous-ovarian cancer (HGSOC) and metastatic breast cancer who would likely respond to PARP-inhibitor and/or platinum therapy, the two current drug classes proven to target HRD. The study data demonstrate that OncoGaze™ outperforms NGS to both capture larger treatable patient populations and better predict therapeutic response.
io9, a software company developing an AI platform for precision oncology that enables the correct front-line targeted therapy at diagnosis, today announced the publication of a positive validation study for OncoGaze™. OncoGaze™ is an AI-biomarker platform to identify clinically actionable biomarkers, eliminating the need for costly and time-consuming next-generation sequencing (NGS) for first-line treatment. The study, published in the Journal of Clinical Oncology, a journal of the American Society of Clinical Oncology (ASCO), reported that DeepHRD, the first biomarker test in the OncoGaze™ AI platform, outperformed FDA-approved homologous recombination deficiency (HRD) companion diagnostics in identifying patients with high-grade serous-ovarian cancer (HGSOC) and metastatic breast cancer who would likely respond to PARP-inhibitor and/or platinum therapy, the two current drug classes proven to target HRD. The study data demonstrate that OncoGaze™ outperforms NGS to both capture larger treatable patient populations and better predict therapeutic response.
“Precision medicine has transformed outcomes for cancer patients, but many with advanced cancer who are eligible for precision therapy are not receiving the right first-line treatments because the journey from diagnosis to therapeutic selection is costly, time consuming, and confusing,” said senior author Ludmil Alexandrov, Ph.D., co-founder, chief scientific officer of io9 and professor of bioengineering and professor of cellular and molecular medicine at University of California San Diego. “This study illustrates how DeepHRD can predict HRD in breast and ovarian cancer from routine pathology slides at the time of initial diagnosis. Our goal is to move this deep-learning AI platform into the clinical arena as quickly as possible, and we look forward to continuing our work to make a real, immediate and immense difference in the lives of cancer patients and their families by getting them onto the treatments they need faster.” Created by a team of engineers, clinical oncologists and medical researchers, the OncoGaze™ AI platform is designed to complement a pathologist’s routine review of a hematoxylin and eosin (H&E) whole biopsy slide to identify clinically actionable biomarkers more quickly, accurately, and cost-effectively than standard molecular testing.
“OncoGaze™ is a groundbreaking technology that allows a pathologist to determine an actionable tumor biomarker at the time of cancer diagnosis, which can reduce the therapeutic selection timeline from weeks to seconds and enables more patients to receive precision treatments,” said Greg Hamilton CEO of io9. “This platform optimally integrates pathologists into first-line precision oncology workflows. We believe that OncoGaze has the potential to ultimately help oncologists and their patients win the race against time by getting them onto the correct treatment sooner which can lead to better outcomes.”
In the study, DeepHRD outperformed FDA-approved HRD companion diagnostics in identifying more HRD-positive patients. Moreover, metastatic breast cancer patients identified by DeepHRD as having HRD exhibited better complete response and progression-free survival when treated with platinum-specific therapies, while platinum-treated ovarian cancer patients exhibited better overall survival. The study evaluated multiple independent clinical cohorts with 141 ovarian cancer and 349 breast cancers patients. When compared to NGS, DeepHRD classified 1.8- to 3.1-fold more patients with HRD in ovarian and metastatic breast cancers, respectively. The published data provide proof-of-concept for io9’s OncoGaze™ platform, and underscore DeepHRD’s potential to significantly increase access to precision medicines across other cancer types, notably castrate-resistant prostate and pancreatic cancers, where HRD is common, clinically actionable and universally recommended.
“This high-impact publication is a pivotal clinical milestone for this rapidly emerging field of AI and oncology,” said co-senior author and io9 co-founder, Scott Lippman, M.D., distinguished professor of medicine at UC San Diego. “Oncologists and their patients currently wait weeks for NGS test results, required to identify potentially lifesaving first-line precision oncology therapies. Up to one-third of patients don’t get an answer at all because of an insufficient tissue sample or inconclusive test results, which limits the reach, and confounds the promise of precision oncology. This new generation of AI profiling applied to routine H&E slides can complement first-line oncologic decision making, and dramatically shorten this critical time delay from diagnosis to therapy, a primary determinant of patient survival.”
Since its founding in 2021, io9 has made significant progress in advancing the OncoGaze™ platform. The company has the exclusive license to this technology from University of California San Diego and plans to expand the OncoGaze™ platform to encompass additional biomarkers, including KRAS, EGFR and BRAF to help tailor optimal therapy decisions for patients and facilitate clinical-trial enrollment and stratification, for a broader variety of cancers including lung, colorectal, and other cancers with well-defined actionable biomarkers.
“OncoGaze™ is a robust platform that can identify biomarkers across multiple cancer types from a digital image of an H&E slide. As such, this scalable cloud-based platform has the opportunity to enable precision medicine across the globe,” said Erik Bergstrom, Ph.D., lead author of the study.
