A groundbreaking study, partly supported by the National Institutes of Health, was published on March 4 in Science Translational Medicine. This research introduces a novel application of DNA fragmentation analysis, termed fragmentome technology, specifically for identifying chronic diseases unrelated to cancer, marking a significant advancement in medical diagnostics.
Innovative DNA Fragmentation Patterns Signal Disease
Liquid biopsies that analyze circulating free DNA (cfDNA) have shown promise for cancer detection, yet their potential for diagnosing other health conditions has been underexplored. In this study, researchers conducted whole genome sequencing on cfDNA samples from 1,576 individuals suffering from liver disease and other medical issues. By investigating DNA fragments across the entire genome, they aimed to identify patterns indicative of disease.
The analysis encompassed approximately 40 million fragments from diverse genomic regions, yielding a substantial dataset compared to traditional liquid biopsy tests. Machine learning algorithms were employed to discern fragmentation patterns associated with various diseases, leading to the development of a classification system that accurately detects early liver disease, advanced fibrosis, and cirrhosis.
"This extends our previous fragmentome research in cancer, applying AI and genome-wide fragmentation profiles of cfDNA to focus on chronic diseases," stated Victor Velculescu, M.D., Ph.D., co-director of the cancer genetics and epigenetics program at Johns Hopkins Kimmel Cancer Center and co-senior author of the study. "Early detection of these diseases can significantly impact patient outcomes, especially for liver fibrosis and cirrhosis, which are reversible in their initial stages."
Distinct Approach to DNA Fragment Analysis
Unlike conventional liquid biopsy techniques that target specific gene mutations, the fragmentome method examines how DNA fragments are created, organized, and spread across the genome. This comprehensive perspective allows the approach to be applied to various conditions, including those that may elevate cancer risk. The study was co-led by Robert Scharpf, Ph.D., and Jill Phallen, Ph.D., both professors of oncology.
"Our focus on the entire fragmentome, rather than individual mutations, enhances the study's power," remarked Akshaya Annapragada, the first author and an M.D./Ph.D. student in Velculescu's lab. "This extensive dataset, combined with machine learning, facilitates the creation of specific classifiers for numerous health conditions."
Transformative Potential for Early Detection
Velculescu highlighted that around 100 million individuals in the U.S. have liver conditions that heighten their risk of cirrhosis and liver cancer. Current blood tests for fibrosis often lack sensitivity, especially in the early stages. Many individuals remain unaware of their liver disease, and earlier intervention could lead to substantial health improvements.
Study Development and Future Directions
This research evolved from a 2023 study focused on the fragmentome of liver cancer, where the team observed that patients with fibrosis or cirrhosis exhibited normal fragmentation profiles yet contained subtle disease signals. The findings suggest that this technology could have broader applications, with future investigations planned to refine the classifier for liver disease and explore its potential for detecting other chronic conditions.
Research Team and Funding Sources
The team included a diverse group of researchers, with funding from various esteemed foundations and NIH grants, ensuring a robust support system for this innovative study.