Researchers at the Johns Hopkins Kimmel Cancer Center have unveiled a groundbreaking artificial intelligence (AI)-driven liquid biopsy test capable of detecting early liver fibrosis and cirrhosis. Utilizing genome-wide cell-free DNA (cfDNA) fragmentation patterns, this innovative approach may also provide insights into broader chronic disease burdens. The findings, published on March 4 in Science Translational Medicine, mark the first systematic application of fragmentome technology, initially used in cancer detection, to chronic non-cancer conditions.
Supported in part by the National Institutes of Health, the study analyzed cfDNA fragmentomes from 1,576 individuals with liver disease and other comorbidities. Researchers employed whole-genome sequencing to scrutinize DNA fragments across entire genomes, focusing on fragment size and distribution, including previously uncharacterized repetitive regions. The data, encompassing roughly 40 million fragments per analysis, surpasses the scope of most existing liquid biopsy tests.
Revolutionizing Disease Detection
Machine-learning algorithms played a pivotal role in processing the extensive data, identifying disease-specific fragmentation signatures. This AI technology enabled the development of a classification system that detects early liver disease, advanced fibrosis, and cirrhosis with remarkable sensitivity. According to Victor Velculescu, M.D., Ph.D., co-director of the cancer genetics and epigenetics program at Johns Hopkins, early detection is crucial as liver fibrosis is reversible in its initial stages but can progress to cirrhosis and increase the risk of liver cancer if left undetected.
Unlike traditional liquid biopsy technologies that target cancer-related gene mutations, the fragmentome approach examines how DNA pieces are cut, packaged, and distributed across the genome. This method is applicable to a range of diseases beyond cancer, including underlying conditions that could eventually lead to cancer development. The research team, co-led by Robert Scharpf, Ph.D., and Jill Phallen, Ph.D., emphasizes the power of analyzing the entire fragmentome, which offers a wealth of information about an individual’s physiological state.
Implications for Broader Health Conditions
With an estimated 100 million Americans at high risk for liver conditions that could lead to cirrhosis and cancer, the need for more sensitive detection methods is pressing. Current blood-based markers for fibrosis have limited sensitivity, particularly in early stages, and existing imaging tools may not be accessible to all patients. Velculescu highlights the potential for early intervention to prevent the progression of fibrosis to more severe conditions.
The study’s origins trace back to a 2023 investigation into liver cancer fragmentomes. Researchers noticed that patients with fibrosis or cirrhosis exhibited subtle disease-related changes in their fragmentation profiles, prompting further exploration. The team developed a fragmentation comorbidity index that distinguished individuals with varying Charlson Comorbidity Index scores, predicting overall survival more accurately than traditional markers.
Future Directions and Broader Applications
The prototype liver fibrosis assay described in the study is not yet a clinical test. Next steps include further development and validation of the liver disease classifier, as well as exploration of fragmentome signatures in additional chronic conditions. Researchers also detected fragmentomic signals associated with cardiovascular, inflammatory, and neurodegenerative conditions, suggesting broader applicability for this technology.
In addition to Velculescu, Annapragada, Scharpf, and Phallen, the study involved a diverse team of researchers, including Zachariah Foda, Hope Orjuela, and many others. The research received support from various foundations and grants, highlighting the collaborative effort behind this promising advancement in medical science.
As the field of liquid biopsies continues to evolve, this study represents a significant step forward in leveraging AI and genomic data to enhance early disease detection and improve patient outcomes. The potential to apply this technology to a wider range of chronic diseases could transform the landscape of preventative medicine in the years to come.
