Identification of a Blood-Based Protein Biomarker Panel for Lung Cancer Detection

Abstract

Lung cancer is the deadliest cancer worldwide, mainly due to its advanced stage at the time of diagnosis. A non-invasive method for its early detection remains mandatory to improve patients' survival. Plasma levels of 351 proteins were quantified by Liquid Chromatography-Parallel Reaction Monitoring (LC-PRM)-based mass spectrometry in 128 lung cancer patients and 93 healthy donors. Bootstrap sampling and least absolute shrinkage and selection operator (LASSO) penalization were used to find the best protein combination for outcome prediction. The PanelomiX platform was used to select the optimal biomarker thresholds. The panel was validated in 48 patients and 49 healthy volunteers. A 6-protein panel clearly distinguished lung cancer from healthy individuals. The panel displayed excellent performance: area under the receiver operating characteristic curve (AUC) = 0.999, positive predictive value (PPV) = 0.992, negative predictive value (NPV) = 0.989, specificity = 0.989 and sensitivity = 0.992. The panel detected lung cancer independently of the disease stage. The 6-protein panel and other sub-combinations displayed excellent results in the validation dataset. In conclusion, we identified a blood-based 6-protein panel as a diagnostic tool in lung cancer. Used as a routine test for high- and average-risk individuals, it may complement currently adopted techniques in lung cancer screening.

Keywords: lung cancer; molecular diagnostics; parallel reaction monitoring; plasma biomarker; targeted proteomics.

Figure 1

Pathway enrichment analysis of the differentially expressed proteins in plasma from lung cancer patients and healthy donors. The enrichment analysis was done using Pathway Commons, Kyoto encyclopedia of genes and genomes (KEGG) and gene ontology (GO) databases. The top 10 significantly enriched pathways are shown. The analysis was done based on the concentrations of the 229 differentially expressed proteins in plasma from lung cancer patients (n = 128) and healthy volunteers (n = 93).

Figure 2

Plasma levels of the 6 protein biomarkers identified as a lung cancer diagnostic panel. Scatter plots of (a) filamin-A (FLNA), (b) tubulin alpha-4A chain (TUBA4A), (c) glutathione S-transferase omega-1 (GSTO1), (d) peroxiredoxin-6 (PRDX6), (e) rho GDP-dissociation inhibitor 2 (ARHGDIB) and (f) cadherin-13 (CDH13) concentrations obtained from lung cancer patients (n = 128) and healthy volunteers (n = 93) using the LC-PRM assay targeting proteotypic peptides. Data points and their median are shown. **** Adjusted p < 0.0001 using the non-parametric Kruskal–Wallis test.