Abstract
BACKGROUND AND PURPOSE: This study investigated the predictive value of clinical characteristics and cerebrospinal fluid (CSF) metabolites for nusinersen efficacy in children with spinal muscular atrophy (SMA). METHODS: In this study, clinical data and CSF samples were collected. We used liquid chromatography-tandem mass spectrometry to analyze CSF metabolites from 42 patients with type II and type III SMA. RESULTS: Although clinical indicators, such as age at treatment initiation and disease duration, did not predict the clinical efficacy of nusinersen, we identified 47 differentially expressed metabolites between effective- and ineffective-treatment patients with type II disease and 109 metabolites in patients with type III disease. Additionally, KEGG-enriched pathway analysis revealed differences in several pathways between the effective- and ineffective-treatment groups for both types II and III. N-myristoyl arginine and 1,1,1,2,2,2-Pentafluoro-7-phenylheptan-3-one were negatively associated with Hammersmith Functional Motor Scale Expanded changes in patients with type III and type III SMA. Furthermore, multivariate receiver operating characteristic curve analysis indicated that differential metabolites have some accuracy in predicting SMA treatment efficacy. CONCLUSION: This study identified CSF metabolites that are predictive of nusinersen efficacy. The results of this study may guide the development of adjunctive therapies for improving the efficacy of nusinersen.