There is no reliable biomarker for early identification of acute lymphoblastic leukemia (ALL), and the pathophysiology of central nervous system involvement (CNSI) in individuals with ALL was unknown. For a study, researchers sought to find independent risk indicators that might detect CNSI early, and an untargeted cerebrospinal fluid (CSF) metabolomics investigation was carried out.

A CNSI evaluation score (CES) was created to assess the likelihood of CNSI based on three independent risk variables after finding 33 substantially changed metabolites between ALL patients with and without CNSI (8-hydroxyguanosine, L-phenylalanine, and hypoxanthine). With positive prediction scores of 95.9% and 85.6% in the training and validation sets, respectively, this predictive model could identify CNSI. Additionally, the higher the extent of central nervous system involvement (CNSI), the higher the CES score.

In addition, they verified the model by monitoring the changes in CES before, during, and after CNSI as well as in remission. The CES showed a good propensity to forecast CNSI development. When compared to observed likelihood, the nomogram we created to forecast the risk of CNSI in clinical practice functioned well.

They may be able to comprehend the pathophysiology of CNSI, identify CNSI at an early stage, and subsequently accomplish tailored precision therapy thanks to the innovative CSF metabolomics study.