Peter Bach, MD, Chief Medical Officer of DELFI Diagnostics discusses blood test for lung cancer detection

November is Lung Cancer Awareness Month. Lung cancer is the leading cause of cancer deaths in America. Fortunately, if lung cancer is caught early through screening, it is more likely to be curable, according to the American Lung Association. Based on early-stage lung cancer statistics (stage 1 versus stage 4), the likelihood of surviving 5 years or more improves to 63 percent if lung cancer is caught before it spreads.

MLO: What is DELFI’s approach to cancer screening?

Peter Bach, MD: DELFI Diagnostics develops accessible blood-based liquid biopsy tests, which offer a new way to enhance early cancer detection. The DELFI Diagnostics FirstLook test is a simple blood test for lung cancer. It can be done in the doctor's office with routine blood work, making it convenient for more patients to get the first level of screening and any subsequent need for additional testing. The goal is to increase screening rates and thus find more lung cancers when they are at their early stage, ultimately decreasing mortality rates.

MLO: Who is DELFI aimed at supporting?

Peter Bach, MD: DELFI tests are built to solve the health issues of the highest-burden population, including those in historically underserved demographics. Lung cancer remains the biggest cancer killer, accounting for 1 in 5 cancer-related deaths in the United States; however, it is also treatable when caught early. Right now, 94% of eligible U.S. adults aren’t being screened for lung cancer partly due to low-dose CT (LDCT) scan availability, LDCT scheduling delays, and lack of same-day or weekend testing. 

MLO: What makes the fragmentomics technology so powerful in DELFI’s approach to liquid biopsy?

Peter Bach, MD: The DELFI FirstLook Lung test evaluates patterns of DNA fragments in the blood that are correlated to the presence of lung cancer. The test is based on fragmentomics, the discovery that the DNA in cancer cells is packaged in a more chaotic manner than DNA in normal cells. When the cancer cells die, they release cell-free DNA (cfDNA) fragments in the blood that have these tell-tale patterns and characteristics of tumor DNA. In an independent validation, FirstLook Lung was shown to have 80% sensitivity in a screening population. The test also demonstrated a negative predictive value (NPV) of 99.8%, which measures how unlikely it is that lung cancer will be detected by LDCT if the FirstLook Lung test returns a ‘Not Elevated’ result.

MLO: How are AI and machine learning guiding the science and lab workflows at DELFI?

Peter Bach, MD: The DELFI platform applies advanced machine-learning technology to whole-genome sequencing data to assess individuals’ cfDNA fragmentation patterns against populations with and without cancer. FirstLook Lung uses these millions of data points to help identify individuals who may have cancer detected through low-dose CT, including early-stage disease, with a high negative predictive value. AI is a powerful step forward for how data and information can support new ways to make technology and solutions accessible and meaningful to those who need them most. 

MLO: What roles do data science and bioinformatics have in the platform DELFI has built?

Peter Bach, MD: Our platform uses machine-learning algorithms, AI, and millions of data points to reliably identify individuals who may have lung cancer, including early-stage disease, and the cancer’s tissue of origin. It can also reveal other DNA changes, including somatic genomic alterations such as chromosomal gains and losses.

MLO: What are other applications for your technology?

Peter Bach, MD: The DELFI-Tumor Fraction, or “DELFI-TF,” is a Research-Use Only (RUO) assay and is being used by biotech and pharmaceutical companies in their search for innovative methods of treating cancer. Because DELFI-TF predicts the fraction of circulating tumor DNA (ctDNA) in blood, companies can gather evidence of what cancer treatments are effective and can prioritize, fund, and accelerate those based on the results. A new paper published in Nature Communications examines the effectiveness of this new approach in a preeminent, peer-reviewed journal on scientific breakthroughs.

Peter Bach, MD, Chief Medical Officer of DELFI Diagnostics is a pulmonary physician and lung cancer epidemiologist. He has devoted his career to repairing defects in the healthcare delivery system that impede access to high-quality cancer care and working to ameliorate healthcare’s cost crisis. His work spans seminal studies including that identification of racial gaps in lung cancer care, the development of the first lung cancer risk prediction model (the “Bach model”), lead authorship on multiple lung screening guidelines, and definitional work on pharmaceutical pricing and value. He has been elected to the National Academy of Medicine, American Society for Clinical Investigation, and the Johns Hopkins Society of Scholars.