Development and Validation of a Five-Gene Signature to Predict Relapse-Free Survival in Multiple Sclerosis

Abstract

Background: Multiple sclerosis (MS) is an inflammatory and demyelinating disease of the central nervous system with a variable natural history of relapse and remission. Previous studies have found many differentially expressed genes (DEGs) in the peripheral blood of MS patients and healthy controls, but the value of these genes for predicting the risk of relapse remains elusive. Here we develop and validate an effective and noninvasive gene signature for predicting relapse-free survival (RFS) in MS patients.Methods: Gene expression matrices were downloaded from Gene Expression Omnibus and ArrayExpress. DEGs in MS patients and healthy controls were screened in an integrated analysis of seven data sets. Candidate genes from a combination of protein–protein interaction and weighted correlation network analysis were used to identify key genes related to RFS. An independent data set (GSE15245) was randomized into training and test groups. Univariate and least absolute shrinkage and selection operator–Cox regression analyses were used in the training group to develop a gene signature. A nomogram incorporating independent risk factors was developed via multivariate Cox regression analyses. Kaplan–Meier methods, receiver-operating characteristic (ROC) curves, and Harrell's concordance index (C-index) were used to estimate the performance of the gene signature and nomogram. The test group was used for external validation.Results: A five-gene signature comprising FTH1, GBP2, MYL6, NCOA4, and SRP9 was used to calculate risk scores to predict individual RFS. The risk score was an independent risk factor, and a nomogram incorporating clinical parameters was established. ROC curves and C-indices demonstrated great performance of these predictive tools in both the training and test groups.Conclusions: The five-gene signature may be a reliable tool for assisting physicians in predicting RFS in clinical practice. We anticipate that these findings could not only facilitate personalized treatment for MS patients but also provide insight into the complex molecular mechanism of this disease.