New AI tool promises rapid cancer biomarker detection from biopsy slides
Researchers at the University of California San Diego have developed an artificial intelligence system that could revolutionise precision oncology by identifying key genomic alterations directly from tumour biopsy images, potentially saving crucial time and resources in cancer treatment.
AI bypasses genomic sequencing for faster results
Scientists at the University of California San Diego have unveiled a groundbreaking artificial intelligence (AI) tool that could transform cancer care by rapidly detecting clinically actionable genomic alterations directly from tumour biopsy slides. This innovative approach, detailed in a recent paper published in the Journal of Clinical Oncology [1], promises to overcome significant barriers in precision oncology by eliminating the need for time-consuming and expensive genomic testing.
The new AI protocol, named DeepHRD, was developed by a multidisciplinary team of engineers and medical researchers led by Ludmil Alexandrov, Ph.D., professor of bioengineering and cellular and molecular medicine at UC San Diego. The technology is designed to save weeks of waiting time and thousands of dollars in clinical oncology treatment workflows, particularly for breast and ovarian cancers.
Alexander Ludmil
Addressing delays and inequalities in cancer care
Current standard practice in precision oncology often involves a prolonged wait for genomic test results following an initial tumour diagnosis. This delay can be critical, potentially impacting treatment outcomes. Moreover, the high costs associated with these tests have created significant disparities in access to precision medicine, especially in resource-constrained settings. Erik Bergstrom, Ph.D., lead author of the study and a postdoctoral researcher in Alexandrov’s lab, explained the motivation behind the project: “Unfortunately, high costs, tissue requirements and slow turnaround times have hindered the widespread use of precision oncology, leading to suboptimal – potentially detrimental – treatment for cancer patients. We wanted to see if we could develop a completely different approach to resolve this serious issue by designing AI to circumvent the need for genomic testing.”
How DeepHRD works
The AI system leverages the minimum amount of patient information available early in the diagnostic process. It focuses on the traditional tumour biopsy, a tissue sample routinely processed and examined under a light microscope – a method that has remained largely unchanged since the late 19th century.
DeepHRD applies advanced AI techniques directly to these traditional tissue slides, allowing for instantaneous and accurate detection of cancer genomic biomarkers. Specifically, the team focused on identifying homologous recombination deficiency (HRD), a condition in which cancerous cells lose a specific DNA damage repair mechanism.
This biomarker is particularly significant for patients with ovarian or breast cancers, as those harbouring HRD generally respond well to platinum and PARP (poly-ADP ribose polymerase) therapies, two common forms of chemotherapy.
Advantages over current genomic testing
The AI approach offers several advantages over current genomic testing methods. Alexandrov noted: “Oncologists can prescribe treatment immediately after initial tissue diagnosis. Remarkably, the AI test has a negligible failure rate, while current genomic tests have a failure rate of 20 to 30 percent, necessitating re-testing, or even invasive re-biopsy.” This significant reduction in failure rate, combined with the near-instantaneous results, could dramatically improve treatment timelines and patient outcomes.
Implications for global cancer care
Scott Lippman, M.D., distinguished professor of medicine and co-senior author of the study, emphasised the potential global impact of this technology. He noted that the AI tool could remove barriers of time and money, allowing for immediate, universal access to actionable genomic biomarker detection.
“The era of precision oncology took off in the late 90s, but recent U.S. studies show that the vast majority of cancer patients are not getting FDA-approved precision therapy,” Lippman said. “And the prime reason is because they’re not getting tested. As a clinical oncologist – and I’ve been doing this for nearly 40 years – there is no question that this approach is the future of precision oncology.”
Future applications and commercialisation
The AI technology behind DeepHRD is protected by provisional patents through UC San Diego and has been licensed to io9, a company aimed at bringing this AI platform into clinical use. The researchers anticipate that the same technology could be applied to most other genomic biomarkers and many forms of cancer, potentially revolutionising precision oncology across a broad spectrum of malignancies.
Reference:
1. Bergstrom, E. N., Abbasi, A., Díaz-Gay, M., et. al. (2024). Deep Learning Artificial Intelligence Predicts Homologous Recombination Deficiency and Platinum Response From Histologic Slides. Journal of Clinical Oncology.
https://doi.org/10.1200/JCO.23.02641
