A risk-prediction model that combines patients’ health history with clinical characteristics of their lung nodules may help physicians determine which will develop into cancer, according to the results of a study published in Cancer Prevention Research.

Researchers analyzed data from 2,942 patients with a lung nodule who were evaluated at a single institution between January 2002 and December 2015 and assigned them to a discovery (n = 1,469) or replication (n = 1,455) cohort. During the 13-year follow-up, 171 patients developed lung cancer. 

They created an overall risk score based on factors identified through a multivariate analysis: patient clinical health history (i.e., age, smoking-pack years, personal cancer history, and diagnosis of chronic obstructive pulmonary disease) and radiologic presentation of the nodule (i.e., size, spiculation, and ground-glass opacity). Applied through the risk-prediction model, the tool had a 73% sensitivity and 81% specificity to identify cancer risk.

The authors noted that the single-center study findings may not be generalizable and that additional research is needed to validate the results.