(192ae) Pharmacometabonomics Approach for Early Prediction of Neuropathy

Vincristine (VCR) is a core chemotherapeutic drug administered to pediatric Acute Lymphoblastic Leukemia (ALL) patients. In a subgroup of population, it leads to Vincristine Induced Peripheral Neuropathy (VIPN), which is the dose-limiting toxicity. For few patients, VIPN is severe with long-term effects. Even though VCR has been used in chemotherapy for more than 50 years now, predictors and mechanism of VIPN induction are unclear. Historically, an empirical cap of 2 mg has been kept for dosage. Due to this, a population not susceptible to VIPN receives sub-therapeutic treatment, while another population may experience severe neuropathy. It is of interest to find predictors/biomarkers which can discriminate populations based on VIPN severity, before the treatment starts. A predictive model with the selected set of biomarkers will aid clinicians in identifying patients susceptible to neuropathy, and in adjusting dosage accordingly.

We are using pharmacometabonomics approach to find metabolites as biomarkers that can predict VIPN severity in pediatric ALL patients. In a retrospective pilot study done on 12 patients, metabolites were found to be differentially expressed in high and low VIPN patients. Presently, retrospective non-fasting plasma samples of 36 pediatric ALL patients are being used. These samples were collected at three time points: day 8, day 29 of induction period, and after around 6 months from the start of the therapy. A preliminary analysis (feature selection using elastic net logistic regression) showed that a set of 17 metabolites in +6 months samples could accurately identify patients with high and low VIPN. A logistic regression model built with these 17 metabolites had an area under the Receiver Operating Characteristics curve of 0.99. Currently, extensive analysis of metabolomics data at all the time points is being done. We aim to find metabolites that are becoming differentially expressed over time. Longitudinal feature selection algorithms are being explored for this purpose. To further explore the biomarkers found, we intend to identify the structure of these metabolites and find their relevant pathways.