Transforming Cancer Treatment with Artificial Intelligence

 UC San Diego Health has long been a force in developing new medical technologies, but if past advancements have come in waves, AI is coming on like a tsunami. AI is currently reshaping nearly every facet of cancer care, from early diagnostics to drug discovery and personalized patient treatments — and Moores Cancer Center is ready to ride this wave. 

“Adoption of AI at scale requires a culture of intention,” says Lee Stein, an industry leader in advanced technology and notable expert advisor to Moores Cancer Center Director Dr. Diane Simeone. “The entire organization must be on board to utilize AI in every single area — whether surgical, therapeutic or in predictive analytics — and be willing to get hands-on at every possible moment to actively integrate it into processes. That is the mindset it takes to be the leading health care institution in AI.” 

And this is exactly the mindset of Moores Cancer Center. Several of our innovative leaders are currently using AI-driven solutions to optimize treatments, enhance the patient experience and ultimately improve cancer outcomes.  

BUILDING A BETTER BIOPSY

 Ludmil B. Alexandrov, PhD, is leading a team of researchers to develop an AI-driven tool to improve and streamline analysis of cancer biopsies. The system, known as DeepHRD, uses machine learning to target homologous recombination deficiency (HRD), a DNA repair defect found in some breast and ovarian cancers and an indicator of responsiveness to specific therapies. 

Where traditional methods of detection require expensive and time-consuming genomic sequencing, DeepHRD can read digital images of cancer tissue slides directly, using AI to identify HRD. Such reductions in time and cost will open the door to personalized treatment approaches for more patients, especially those in less- resourced settings. 

“It is very concerning that high costs and time delays render lifesaving treatment protocols inaccessible for most patients, disproportionately impacting resource- constrained settings,” Alexandrov says. “This AI approach saves the patient critical time — oncologists can prescribe treatment immediately after initial tissue diagnosis.” 

DeepHRD has truly transformative potential. Beyond making precision medicine more accessible and efficient, early studies indicate adaptability for other cancer types, promising a future where AI-led diagnostics could revolutionize oncology on a larger scale. 

A GENETIC MATCHMAKER FOR PATIENTS AND THERAPIES

 While Alexandrov’s lab is using AI to circumvent genomic analysis, others are using AI to unlock the immense power of genomic data. Trey Ideker, PhD, and his team have developed an AI-based platform that uses genomic data to match therapies with cancers at a molecular level.

“Right now, we can’t match the right combination of drugs to the right patients in a smart way,” says Ideker. “And especially for cancer, where we can’t always predict which drugs will work best given the unique, complex inner workings of a person’s tumor cells.” 

Instead of a “one-size-fits-all” approach to cancer treatment, Ideker’s DrugCell platform analyzes the genes of individual patient tumors and determines the most effective cancer therapy for their genetic makeup. But given the size and scale of genetic information, AI is necessary to help compute it all. 

The DrugCell platform has been trained on more than 1,200 tumor cell lines and responses to nearly 700 therapeutic drugs, for a total of more than 500,000 pairings that help determine which drugs are effective for which genetic profiles — and that is just the initial version. As the platform evolves and learns more from datasets, it can further simulate cellular responses to various drugs, using AI to essentially “test” therapies outside of the body, sparing patients from unwanted side effects and leading to more effective treatments and successful outcomes.