A new consortium, the Alliance for Global Implementation of Lung and Cardiac Early Disease Detection and Treatment (AGILE), has proposed a public health program of longitudinal chest computed tomography (CT) screenings among a global high-risk population of tobacco-exposed individuals utilizing artificial intelligence (AI) to ease the workflow burdens of the early detection of tobacco-related diseases. A perspective on this topic by Mulshine et al was published recently in the European Journal of Cancer.
The paper by James L. Mulshine, MD, of Rush University in Chicago, and colleagues suggests that chest CT screening can be used to identify lung cancer, emphysema, and coronary artery disease in earlier stages when chest CT images are assessed over time at screening intervals. Multi-chest disease assessment could collectively and significantly improve cure rates and patient outcomes, as outlined in a companion article in the European Journal of Cancer by Mulshine et al. Use of AI tools could assist in reducing the workload involved in reviewing the quantity of serial annual CT scans and potentially identifying other useful early disease markers.
There are a number of vital infrastructure challenges laid out in the article for establishing such a public health program that could identify these diseases among a large high-risk population, including radiologic workflow and capacity, validation of screening and image data quality, costs and investment, cloud and storage capacity, other potential uses of the data, constantly evolving AI tools, and more. The consortium recommended steps to take to implement global standards for low-dose CT screening for lung cancer to manage challenges of international standardization in CT image acquisition and interpretation, as well as system management.
Additionally, the screening registry requires participants to voluntarily provide their scans to the database, necessitating an avenue of reaching and communicating to possible study participants. “The need for effective approaches to communicating about multiple disease detection with multiple potential audiences is apparent,” Dr. Mulshine and colleagues wrote. The consortium believes, however, that AI could help overcome this issue as well by providing tools for outreach, quality control, and other communication needs.
The potential benefits of the program are also described, including vast data available for clinical research aimed at identifying targeted drugs for treating or preempting the lethality of screen-detected lung cancer, emphysema, and coronary artery disease, or a combination of these diseases. In addition, the data could be used in other research studies with a financial investment for those wishing to access the screening database for the benefit of public health.
“Through regular analysis of the images, cases, and data deposited in these registries, a more precise understanding of the natural history of a variety of tobacco-related diseases may emerge,” the authors stated.
Although the efforts to establish this comprehensive chest CT scanning program are challenging, various studies to date have paved the way to success. For instance, the European 4-IN-THE-LUNG-RUN screening study compared the false-negative rate of an AI tool for first reads of lung cancer screening CT images compared with that of experienced radiologists. Long-term screening research was also conducted in the International Early Lung Cancer Action Project, which recently reported 25 years of follow-up of screen-detected lung cancer cases and demonstrated long-term cancer-free survival in over 80% of patients.
With this proposed public health project, the AGILE consortium seeks to bring forward attention in both the medical community and the public about possible detection and management of earlier and treatable stages of serious tobacco-associated diseases. International cooperation to define harmonized global chest screening workflows and to evolve computational imaging tools may advance the prospects to make such an ambitious shared endeavor feasible.
Disclosure: For full disclosures of all study authors, visit ejcancer.com.
