Identification of an immune-related gene prognostic index for predicting prognosis, immunotherapeutic efficacy, and candidate drugs in amyotrophic lateral sclerosis.

Abstract

Considering the great insufficiency in the survival prediction and therapy of amyotrophic lateral sclerosis (ALS), it is fundamental to determine an accurate survival prediction for both the clinical practices and the design of treatment trials. Therefore, there is a need for more accurate biomarkers that can be used to identify the subtype of ALS which carries a high risk of progression to guide further treatment. The transcriptome profiles and clinical parameters of a total of 561 ALS patients in this study were analyzed retrospectively by analysis of four public microarray datasets. Based on the results from a series of analyses using bioinformatics and machine learning, immune signatures are able to be used to predict overall survival (OS) and immunotherapeutic response in ALS patients. Apart from other comprehensive analyses, the decision tree and the nomogram, based on the immune signatures, were applied to guide individual risk stratification. In addition, molecular docking methodology was employed to screen potential small molecular to which the immune signatures might response. Immune was determined as a major risk factor contributing to OS among various biomarkers of ALS patients. As compared with traditional clinical features, the immune-related gene prognostic index (IRGPI) had a significantly higher capacity for survival prediction. The determination of risk stratification and assessment was optimized by integrating the decision tree and the nomogram. Moreover, the IRGPI may be used to guide preventative immunotherapy for patients at high risks for mortality. The administration of 2MIU IL2 injection in the short-term was likely to be beneficial for the prolongment of survival time, whose dosage should be reduced to 1MIU if the long-term therapy was required. Besides, a useful clinical application for the IRGPI was to screen potential compounds by the structure-based molecular docking methodology. Ultimately, the immune-derived signatures in ALS patients were favorable biomarkers for the prediction of survival probabilities and immunotherapeutic responses, and the promotion of drug development.