Subscribe for email alerts

Don’t miss a single Science & Enterprise post. Sign up for our daily email alerts.

Please share Science & Enterprise

AI-Analyzed Liquid Biopsy IDs Early Lung Cancer

Lung cancer illustration

(NIH.gov)

20 Aug. 2021. A blood test analyzed with algorithms is shown to detect more than 90 percent of early-stage lung cancer cases in people at high risk of the disease. The study, conducted by a team from the company Delfi Diagnostics Inc. in Baltimore and colleagues from Johns Hopkins University, appears today in the journal Nature Communications.

Lung cancer, as noted by the authors, is the most lethal form of the disease, with a survival rate under 20 percent. Individuals at risk of lung cancer, such as smokers, can be screened for the disease with low-dose computed tomography or CT scans, but the study team indicates less than six percent of at-risk individuals use this technique, due to radiation exposure and false-positive fears. As a result, there’s an urgent need for a non-invasive technique to screen people at risk for lung cancer, and even the population at large.

Delfi Diagnostics offers a technology for cancer screening that analyzes blood samples for cell-free DNA fragments that indicate tumor formation at its earliest stage, even before detection by other means. Delfi, short for DNA evaluation of fragments for early interception, investigates DNA fragments for abnormal physical properties associated with various cancers and traced to locations in the body. Delfi Diagnostics’ analytical engine includes machine learning algorithms, a form of artificial intelligence, that the company says are trained by millions of data points.

Accurately identify 90 percent or more

In the Nature Communications paper, researchers led by Johns Hopkins oncology professor Victor Velculescu and biostatistician Robert Scharpf analyzed blood samples from 365 individuals over seven months considered at higher risk of lung cancer at a hospital in Copenhagen, Denmark. Nearly all, 323 of the 365 individuals, had respiratory or symptoms indicating lung cancer, including smokers cough. The team enlisted separate groups of 46 lung cancer patients and 385 individuals without lung cancer to validate the initial results.

Using Delfi analysis and CT scan, the researchers accurately identified 90 percent or more of participants in these groups with lung cancer in early stages 1 or 2 (91%) and in later stages 3 and 4 (96%), with 80 percent true-negative specificity. The analysis also highlights indicators of small cell and non-small cell lung cancer. Moreover, the researchers point out that Delfi screening makes it possible to reduce the use of low-dose CT or LDCT scans for lung cancer screening by half.

Velculescu and Scharpf were among the founders of Delfi Diagnostics two years ago, and also serve as the company’s CEO and head of computational biology respectively. “Because blood tests are so much easier to administer than LDCT,” says co-author and co-founder Nicholas Dracopoli, Delfi’s chief scientist, in a company statement released through Cision, “we believe a high-performing, cost effective assay could greatly increase the number of lung cancers that are detected early, when it can make a difference in care and eliminate a great deal of false positives compared to the current standard of care.”

The company is enrolling participants in a clinical trial of Delfi Diagnostics technology, enrolling 1,700 current or former smokers, or other individuals considered at high lung cancer risk. The trial is taking blood samples from participants for Delfi analysis, and then tracked for the next 12 months. The study aims to test the technology’s ability to accurately detect and characterize lung cancer, or find no signs of the disease.

More from Science & Enterprise:

*     *     *

Comments are closed.