Identification and validation of novel circulating biomarkers for early diagnosis of lung cancer

Abstract

Objectives: This study aimed to identify novel circulating biomarkers in lung cancer.

Materials and methods: Expression of 1000 secreted proteins in 15 early lung cancer patients and 10 healthy controls' plasma were examined by RayBiotech antibody array. Candidate biomarkers were identified by SPSS analysis (p < 0.05 between lung cancer and healthy controls was considered significant) and validated by ELISA in 371 lung cancer patients and 388 healthy controls. Receive operating characteristic curve and binary logistic regression were performed to evaluate the diagnosis efficacy and to establish diagnostic models.

Results and conclusion: Twenty-two molecules expressed aberrantly in early lung cancer patients compared to healthy controls. Large sample validation showed significantly up-regulated levels of brain angiogenesis inhibitor 1, E-Cadherin, integrin-binding sialoprotein, and down-regulated expression of thrombospondin-1 (all p < 0.0001) in lung cancer patients. Receive operating characteristic curve analysis indicated E-Cadherin, brain angiogenesis inhibitor 1 and thrombospondin-1 had higher sensitivity and specificity than classical biomarkers carcinoenbryonic antigen, carbohydrate antibody 19-9 and cytokeratin 19 fragments. The high sensitivity and specificity of E-Cadherin brain angiogenesis inhibitor 1 and thrombospondin-1 were also confirmed in early lung cancer analysis. Combination analysis showed brain angiogenesis inhibitor 1, E-Cadherin and thrombospondin-1 had better diagnostic efficacy than classic lung cancer biomarkers. Our findings demonstrated the potential status of BAI-1, E-Cadherin and TSP-1 in diagnosis of lung cancer.

Keywords: Antibody array; Combination analysis; Early diagnosis; Lung cancer; Novel biomarkers.