NEW YORK – A team led by Johns Hopkins University has developed a circulating cell-free DNA (cfDNA) “fragmentome” and machine learning approach to boost the detection of hepatocellular carcinoma, suggesting the strategy can help monitor people with viral hepatitis, cirrhosis, non-alcoholic fatty liver disease (NAFLD) and other risk conditions.
“Medical societies around the world recommend screening of most-at-risk populations, currently with abdominal ultrasound with or without alpha-fetoprotein,” co-lead and co-corresponding authors Victor Velculescu, oncology researcher and co-director of The Johns Hopkins Cancer Genetics and Epigenetics Program Kimmel Cancer Center and Amy Kim of Johns Hopkins University School of Medicine and colleagues explained in Discovery of cancer Friday.
Even so, the authors noted that overall adherence to international guidelines remains low and highlighted the “great need for the development of accessible and sensitive screening approaches for HCC worldwide.”
For their latest study, the researchers relied on low-coverage whole genome sequencing to profile the characteristics of cDNA fragments in blood samples from individuals with HCC, healthy control individuals, or individuals with other liver conditions such as viral hepatitis, cirrhosis of the liver or NAFLD.
From there, they turned to an artificial intelligence approach known as DELFI (Evaluation of Fragment DNA for Early Interception) to focus on the characteristics of the cfDNA fragmentome corresponding to HCC – a approach that members of the same team used to find lung cancer. classifier in the past.
“Given the direct link between genomic and chromatin changes in liver cancer and cfDNA fragmentation, we used a machine learning approach to determine whether changes in cfDNA fragmentomes could distinguish HCC patients from those without cancer,” the authors explained.
The team started by applying the approach to blood plasma samples from more than 500 people treated in the United States or Europe, bringing available chromatin immunoprecipitation sequencing data on factor binding patterns. liver cancer transcripts to report DNA fragmentome features that were distinct in plasma samples from 75 Comparison between patients with HCC and those from 133 high-risk participants with other liver conditions or individuals healthy controls.
In the process, the researchers identified genome-binding-site, chromatin, and transcription-factor shifts that labeled DNA fragmentomas found in non-cancerous plasma samples. While HCC-related fragmentomas were more diverse, they noted, it was possible to set up a classifier score with DELFI to distinguish HCC cases – the results were then validated with blood plasma samples from 223 other Hong Kong people, including people with or without CHC. .
In particular, the team explained that DELFI scores were significantly higher for people with HCC, regardless of tumor stage, while intermediate scores tended to appear in people with hepatitis or cirrhosis. In contrast, people without HCC, hepatitis, or cirrhosis had low DELFI scores.
The researchers noted that the DELFI score could detect HCC with 88% sensitivity and 98% specificity in a group of individuals who appeared to be at average risk for HCC, while the sensitivity was 85% (with a specificity of about 80%) in high-risk patients. band.
“Increased early detection of liver cancer could save lives, but currently available screening tests are underutilized and miss many cancers,” Velculescu said in a statement.
Velculescu is a founder, board member and investor in a related spin-off company known as Defli Diagnostics.
He and his colleagues cautioned that the current findings will need further validation in other large studies. Still, the suggested, their observations that scalable, cost-effective, non-invasive cDNA fragmentoma analyzes can detect patients with liver cancer “could provide an opportunity to screen high-risk and general populations globally.” entire”.
#Liver #Cancer #Detection #Score #Developed #Machine #Learning #Analysis #Blood #Plasma #DNA #Fragments