Within the past several months, researchers at the Johns Hopkins Kimmel Cancer Center have made great advancements in the development of a new blood test that can be used to successfully screen for cancer. The test, the DNA Evaluation of Fragments for Early Interception (DELFI), is named after the process by which it identifies cancer in a patient; an AI system is trained to detect abnormal patterns of DNA fragments that are released by dying cancer cells. So far, it has been tested in several human trials with impressive results. In a trial composed of 208 patients from the US, Denmark, and the Netherlands, DELFI was successful in identifying seven types of cancer (breast, colorectal, lung, ovarian, pancreatic, gastric, and bile duct) at different stages in up to 99% of the blood samples. Similarly, in a trial composed of 215 healthy patients, DELFI gave only four false-positive cancer diagnoses. Most existing blood tests for cancer attempt to identify the mutations within or methylation of DNA that is present in a sample, but this can lead to cancer being overlooked as not all patients show this phenotype. Rather, the DELFI test is rooted in the way that DNA is packaged within the cell nucleus. Under normal conditions, DNA is very organized; in cancer cells, it is packaged chaotically and often contains fragments of differing sizes in different regions. When the DNA contents of a cancer cell are released upon death, these cell-free DNA (cfDNA) fragment patterns can be recognized, analyzed, and referenced against control patterns to determine if their nature is cancerous. Also, the cfDNA can be used to backtrack and identify the tissue of origin with up to 75% accuracy. Since the test is simple to perform and requires inexpensive laboratory procedures, it could end up being more effective in both accuracy and cost than current screening tests. More information on this development can be found here.