A Blood-Based Liquid Biopsy Combined with Machine Learning for Early Lung Cancer Diagnosis

A novel liquid biopsy that collects fragments of circulating tumor DNA combined with advanced machine learning technology was able to detect over 90% of lung cancers in samples from nearly 800 individuals with and without the disease.

Investigators at Johns Hopkins Medicine (Baltimore, MD, USA) debuted the field of fragmentomics as the basis for lung cancer diagnosis. Fragmentomics studies the physical properties of circulating cell-free DNA fragments. DNA is packaged abnormally in cancer cells, resulting in abnormal fragment patterns when cancer cells die and release their DNA into the bloodstream. Cell-free DNA fragments in the blood can indicate the presence of cancer and suggest its likely location in the body.


Advanced machine learning (artificial intelligence) technology was used to compare an individual’s cell-free DNA patterns against populations with and without cancer. The technology, called DELFI (DNA evaluation of fragments for early interception), used millions of data points to identify both the presence of cancer and its tissue of origin.

For the current study, the investigators, working with colleagues in Denmark and the Netherlands, first performed genome sequencing of cell-free DNA in blood samples from 365 individuals participating in a seven-year Danish study called LUCAS. They validated the cancer detection model using the DELFI approach on an independent cohort of 385 non-cancer individuals and 46 lung cancer patients.

Results revealed that combining fragmentation features, clinical risk factors, and CEA (carcinoembryonic antigen) levels, followed by CT imaging, detected 94% of patients with cancer across stages and subtypes, including 91% of stage I/II and 96% of stage III/IV, at 80% specificity. Genome-wide fragmentation profiles across nearly 13,000 ASCL1 transcription factor binding sites distinguished individuals with small cell lung cancer from those with non-small cell lung cancer with high accuracy.

“It is clear that there is an urgent, unmet clinical need for development of alternative, noninvasive approaches to improve cancer screening for high-risk individuals and, ultimately, the general population,” said first author Dr. Dimitrios Mathios, a postdoctoral researcher at Johns Hopkins Medicine. “We believe that a blood test, or liquid biopsy, for lung cancer could be a good way to enhance screening efforts, because it would be easy to do, broadly accessible and cost-effective.”

A national clinical trial of the DELFI approach is being carried out by the biotech company Delfi Diagnostics (Baltimore, MD, USA). This trial is using the technique to evaluate 1,700 participants in the United States, including healthy participants, individuals with lung and other types of cancers.

The DELFI lung cancer study was published in the August 20, 2021, online edition of the journal Nature Communications.

Related Links:
Johns Hopkins Medicine
Delfi Diagnostics