Breakthrough in Cancer Research: AI Revolutionizes Epigenetic Insights

A team at the University of California, Los Angeles (UCLA), has pioneered a groundbreaking AI model that leverages epigenetic factors to precisely predict patient outcomes across various cancer types. This innovative approach surpasses conventional methods and underscores the pivotal role of epigenetics in the realm of cancer diagnosis and treatment.

Researchers from UCLA’s Jonsson Comprehensive Cancer Center have successfully developed an artificial intelligence (AI) model founded on epigenetic factors. This model exhibits remarkable accuracy in forecasting patient outcomes across multiple cancer categories.

The Significance of Epigenetic Factors in Predicting Cancer

Through meticulous examination of gene expression patterns associated with epigenetic factors—factors influencing the activation or deactivation of genes—within tumors, researchers have identified distinct clusters that offer superior predictive capabilities compared to traditional metrics such as cancer grade and stage. These findings, published on November 15th in Communications Biology, lay the groundwork for tailored therapies aimed at regulating epigenetic factors in cancer treatment, including histone acetyltransferases and SWI/SNF chromatin remodelers.

Looking Beyond Genetic Mutations in Cancer

Traditionally, cancer has been attributed primarily to genetic mutations within oncogenes or tumor suppressors. However, advanced next-generation sequencing technologies have ushered in a realization that the state of chromatin and the levels of epigenetic factors maintaining this state play a pivotal role in cancer and its progression. Various aspects of chromatin state, such as modifications to histone proteins or the presence of extra methyl groups in DNA nucleic acid bases, can significantly influence cancer outcomes. Understanding these distinctions among tumors holds the key to comprehending why some patients respond differently to treatments and experience varying outcomes, as elucidated by co-senior author Hilary Coller, a professor specializing in molecular, cell, and developmental biology.

Unveiling the Relationship Between Epigenetic Patterns and Clinical Outcomes

To explore the connection between epigenetic patterns and clinical outcomes, researchers analyzed the expression patterns of 720 epigenetic factors, categorizing tumors from 24 different cancer types into distinctive clusters. Among these adult cancer types, 10 exhibited significant disparities in patient outcomes, encompassing progression-free survival, disease-specific survival, and overall survival. This was particularly evident in adrenocortical carcinoma, kidney renal clear cell carcinoma, brain lower grade glioma, liver hepatocellular carcinoma, and lung adenocarcinoma. These disparities were associated with factors such as advanced cancer stage, larger tumor size, and more pronounced indicators of tumor spread.

Utilizing an AI Model for Predicting Patient Outcomes

Subsequently, the team harnessed gene expression levels of epigenetic factors to train and test an AI model designed to forecast outcomes for the five cancer types with noteworthy differences in survival measurements. The AI model effectively stratified patients into two groups: one with a substantially higher likelihood of favorable outcomes and the other with a heightened likelihood of unfavorable outcomes. Notably, the genes most crucial to the AI model exhibited significant overlap with the signature genes defining the clusters.

Potential for Wider Applications

The researchers anticipate the broader applicability of the pan-cancer AI model, suggesting its testing on additional independent datasets beyond the TCGA cohort. Similar epigenetic factor-based models could potentially be developed for pediatric cancers, shedding light on factors influencing decision-making processes compared to models tailored to adult cancers.

This groundbreaking research not only unveils the critical role of epigenetics in cancer but also offers a blueprint for the creation of AI models based on publicly-available lists of prognostic epigenetic factors. This roadmap holds immense potential for identifying specific targets in cancer treatment.

Funding for this study was provided in part by grants from the National Cancer Institute, Cancer Research Institute, Melanoma Research Alliance, Melanoma Research Foundation, National Institutes of Health, and the UCLA Spore in Prostate Cancer.

Reference: Communications Biology, November 16, 2023, DOI: 10.1038/s42003-023-05459-w.

Frequently Asked Questions (FAQs) about Epigenetic Cancer Prediction

More about Epigenetic Cancer Prediction

• UCLA Jonsson Comprehensive Cancer Center
• Communications Biology Article
• National Cancer Institute
• Cancer Research Institute
• Melanoma Research Alliance
• Melanoma Research Foundation
• National Institutes of Health
• UCLA Spore in Prostate Cancer