Using Artificial Intelligence Blood Testing to Detect Lung Cancer

Johns Hopkins Kimmel Cancer Center researchers developed novel artificial intelligence blood testing technology to detect lung cancer in patients.

The test approach, called DELFI (DNA evaluation of fragments for early interception), finds unique patterns in the fragmentation of DNA shed from cancer cells within the bloodstream.

Using artificial intelligence technology on blood samples from 796 individuals in Denmark, the Netherlands, and the United States, researchers found that the DELFI approach could accurately differentiate between patients with and without lung cancer.

By combining the test with clinical risk factor analysis, a protein biomarker, and computer tomography imaging, DELFI detected 94 percent of patients with cancer across different stages and subtypes. The detection included 91 percent of patients with earlier or less invasive stage I/II cancers and 96 percent with more advanced stage III/IV cancers.

While lung cancer is the most common cause of cancer death, claiming almost 2 million lives each year, fewer than 6 percent of Americans at risk of the illness undergo the recommended low-dose computed tomography screenings. According to senior study author Victor E. Velculescu, MD, PhD, projections show the screenings could prevent tens of thousands of deaths.

Low screening rates could be due to concerns of potential harm from investigating false positive imaging results, radiation exposure, or worries about potential complications from invasive procedures.

“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,” lead author Dimitrios Mathios said in a press release.

“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.”

The DELFI technology uses a blood test to indirectly measure the packaging of DNA inside the nucleus of a cell by examining the size and amount of cell-free DNA present in the circulation from different regions across the genome. When cancer cells die, they then release DNA into the bloodstream.

DELFI identifies the presence of cancer using machine learning to study millions of cell-free DNA fragments for abnormal patterns. The approach provides a view of cell-free DNA called the “fragmentome” and only requires low-coverage sequencing of the genome, allowing technology to be cost-effective in a screening setting.

For the study, Johns Hopkins researchers worked with fellow researchers from Denmark and the Netherlands. The team first performed genome sequencing of cell-free DNA in blood samples from 365 individuals participating in a seven-year Danish study called LUCAS. Most patients were at high risk for lung cancer and had smoking-related symptoms, including cough and difficulty breathing.

The DELFI approach discovers that patients who were later diagnosed with cancer had wide variation in their fragmentome profiles, while those who didn’t have cancer had consistent fragmentome profiles.

Overall, the approach detected over 90 percent of patients with lung cancer (including those with early and advanced stages) and with different subtypes.

“DNA fragmentation patterns provide a remarkable fingerprint for early detection of cancer that we believe could be the basis of a widely available liquid biopsy test for patients with lung cancer,” said author Rob Scharpf, PhD.