A narrative review of risk prediction models for lung cancer screening

Abstract

Lung cancer is the leading cause of cancer-death worldwide. The U.S. Preventative Services Task Force (USPTSF) approved screening for current or former smokers aged 55–80 based on the results of the National Lung Screening trial (NLST). Current guidelines use rigid inclusion criteria, therefore new attention has turned to use of risk-prediction models for lung cancer to reduce the number needed to screen as well as identify high-risk patients who don’t meet current screening guidelines. Our paper serves as an expert narrative review of new literature pertaining to lung cancer risk prediction models for screening based on review of articles from PubMed and Cochrane database from date of inception through June 11, 2020. We used the MeSH search terms: “lung cancer”; “screening”; “low dose CT”, and “risk prediction model” to identify any new relevant articles for inclusion in our review. We reviewed multiple risk-prediction models including recent updates and systematic reviews. Our results suggest risk projection models may reduce false positive rates and identify high risk patients not currently eligible for screening. However, most studies were heterogenous in both their variables and risk threshold cutoffs for screening. Furthermore, a lack of prospective validation continues to limit the generalizability. Therefore, we acknowledge the need for further prospective data collection regarding use of risk-prediction modeling to refine lung cancer screening.