Development of exploratory algorithms to aid in risk of malignancy prediction of indeterminate pulmonary nodules

Abstract

Early detection of lung cancer allows for earlier stage treatment initiation and improved patient prognosis. This report focuses on utilization of combining patient demographic information with non-invasive biomarkers and their potential ability to predict risk of malignancy of nodules. A pilot study cohort of 141 subjects with IPNs (105 stage I cancer and 36 benign nodules) were collected by RUMC. The demographic variables of gender, age, sex, race, ethnicity, nodule size (mm), and smoking pack years, as well as the plasma levels of CA-125, SCC, CEA, HE4, ProGRP, NSE, Cyfra 21-1, hs-CRP, Ferritin, IgG, IgG1, IgG2, IgG3, IgG4, IgE, IgM, IgA, KFLC, and LFLC, were assessed for this cohort. Multivariable analyses of the previously aforementioned biomarkers and demographic variables yielded a reduced algorithm consisting of CA-125, total IgG, IgA, IgM, IgE, LFLC, nodule size, and smoking pack years with improved performance (AUC 0.82, 95 %CI 0.74–0.90) over the same analysis of the demographic variables (age, nodule size, and smoking pack years) alone (AUC 0.70, 95 %CI 0.61–0.78). This reduced algorithm of biomarkers and demographic variables may aid in assessing the risk of IPN malignancy which could be a useful stratification tool in early detection of lung cancer in high-risk subjects.