Lung Cancer Screening Challenges and Solutions

One only needs to look back to when Breast Cancer Screening was taking off (jumping from 30% to 60% between 1987 and 1993¹) to see how mammography was the driving force in a transformation of the practice of radiology. The flood of mammography reports going back to primary care physicians shined a light on the “unintelligible descriptions and ambiguous recommendations” radiologists were often using.² In response, the American College of Radiology created a standardized lexicon and reporting system called the Breast Imaging-Reporting and Data System (BI-RADS). The success of BI-RADS has inspired the creation of many different standardized reporting systems in use today³, and has influenced radiology reporting in general, driving it to be more structured.⁴ The introduction of a screening test changed the field of radiology for the better and gave rise to products such as Volpara’s Patient Hub.

Today, low dose CT Lung Cancer Screening (LCS) is taking off. Large, multi-site studies such as NLST⁵, NELSON⁶, and UKLS⁷ are consistently reporting reductions in lung cancer mortality rates of approximately 20%. Screening rates, with an expected COVID-19 hiccup in 2020, are consistently climbing. This time, Radiology departments are better prepared to clinically deploy a new screening exam. The BI-RADS inspired Lung-RADS and Fleischner Society⁸ reporting guidelines provide LCS with structured reporting. However, with lung cancer, structured reporting is not sufficient to guide appropriate follow up. Lung nodules, many of them benign, are found in 30% or more of these screening exams⁹, and the downstream cost can be an order of magnitude or more than the screening itself.¹⁰ In addition, interventional procedures (biopsy and thoracic surgery) have low positive predictive values, and a reported 22% complication rate that costs the US $1.4 Billion annually.¹¹ From this, whilst breast and lung screening present many similarities, false positives in lung screening could be overwhelming cost-wise.

We founded RevealDx because we believe that for LCS to be most effective, radiologists need tools that can help them improve on the performance of using Lung-RADS guidelines alone. We have invested years of research and development, as well as years of multi-site validation and testing, to build a radiomics and AI based software application, RevealAI-Lung, that can help radiologists perform lung screening better. Better clinical decision making can result in lower cost, fewer complications and better outcomes. RevealAI-LUNG (which already has CE Mark) will be submitted to the FDA in early 2022.

Adoption of RevealAI-LUNG into LCS programs will not only reduce cost and improve clinical decision making, it will represent a proof point of how radiomics can work with guidelines and radiologists to provide a more valuable report for referring physicians. Our hope is that our success will inspire similar solutions in other clinical areas. A key aspect is the close interaction between our RevealAI-LUNG and Volpara’s Patient Hub.

Lung cancer is responsible for more cancer deaths than breast, prostate and colon cancer combined, and 2020 was the first year that smoking increased in the US in several years. Wildfires and pollution are increasing, with an as yet unknown impact on lung cancer risk. We are driven to make LCS programs successful and cost effective today and in the future. Together, with Volpara, RevealDx hopes to help realize this incredibly worthwhile goal.