Establishment of a decision tree model for diagnosis of early rheumatoid arthritis by proteomic fingerprinting

Abstract

Aim: The objective of this study was to identify proteomic biomarkers specific for rheumatoid arthritis (RA) by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) in combination with weak cationic exchange (WCX) magnetic beads.

Methods: Serum samples from 50 patients with RA and 110 disease controls (50 SLE and 60 SS) and 51 healthy individuals were analyzed. The samples were randomly divided into a training set or test set to develop a diagnostic model for RA.

Results: A total of 83 protein peaks were identified to be related with RA, in which four of the peaks with mass-charge ratio (m/z) at 8133.85, 5844.60, 13 541.3 and 14 029.0 were selected to establish a model for diagnosis of RA. This classification model could separate patients with RA from diseased and healthy controls with sensitivity of 84.0% and specificity of 92.5%, and its accuracy was confirmed in the blind testing set with high sensitivity and specificity of 80.0% and 93.3%, respectively.

Conclusions: This study suggested that potential serum biomarkers for RA diagnosis could be discovered by MALDI-TOF-MS. The classification tree model set up in this study might be used as a novel diagnostic tool for RA.

Keywords: mass spectrometry; matrix-assisted laser desorption/ionization time-of-flight; proteomics; rheumatoid arthritis.