How do we bring lung cancer diagnostic biomarkers to the clinic?

Lung cancer is the #1 cancer killer Worldwide and accounts for 38% people dying of respiratory diseases. It is a big healthcare burden with costs of approximately $28 Billion per year. Early diagnosis is crucial for improvement of survival rate. Screening lung cancer with low dose chest CT screening effectively could save approximately 8% of the population dying yearly from cancer. Nodules detected by chest CT are categorized proportionally by risk level based on size from low to high. The biggest problem represents the intermediate zone, the largest category, which needs improved diagnostics in a noninvasive way. Imaging and biomarkers may be helpful in stratifying the risk. Biomarkers have multiple layers of evidence and add value, cost effective and eventually save lives. In collaboration with a group in the UK, 15,000 benign lung nodules and 932 cancers from NLST were analyzed. The artificial intelligent model LCP-CNN reclassifies about one third of nodules into a lower or higher risk group, decreasing number of unnecessary procedures and decreases time to diagnosis. Biomarker research of early detection should reach ‘evidence’ based randomized controlled clinical trials observational studies clinical expertise. Additionally, we need to make sure patients are involved in the design of the studies, need informed consent and we need to work with Statistical models who help assess the strength and robustness of biomarkers. We need to look at the continuous risk factors and risk prediction models, molecular and imaging, to address the shortcomings of the clinical prediction models to find a solution of our current evaluation of lung nodule.

Learning Objectives:

• Understand the difference between LDCT vs Multi Analyte Assay Analysis
• Identify risk factors and risk predictions models to address the lung cancer screening gap