# Stratified Therapeutic Drug Monitoring Could Potentially Improve the Efficacy and Safety of Oxcarbazepine in Children with Epilepsy: Novel Insights from a Single-Center, Large-Sample, Retrospective Real-World Study

## TL;DR

This study suggests that monitoring oxcarbazepine metabolite levels in children with epilepsy could improve treatment effectiveness and safety.

## Contribution

The study identifies a critical 'risk transition point' at MHD concentrations ≥10 μg/mL and proposes a therapeutic range for Chinese children.

## Key findings

- A plasma MHD concentration ≥10 μg/mL is a critical risk transition point for treatment failure.
- Monotherapy with OXC showed a higher response rate than add-on therapy.
- Adverse events occurred in 30.5% of patients, with most risks remaining stable over time.

## Abstract

Objective: This study aimed to characterize the population exposure, efficacy, and safety profiles of oxcarbazepine (OXC) in Chinese children with epilepsy using real-world data, define its optimal therapeutic range, and inform individualized therapy. Methods: This single-center retrospective cohort study included pediatric patients (<18 years) who received OXC therapy between September 2021 and August 2024, with follow-up continuing until February 2025. The concentration of the active metabolite 10,11-dihydro-10-hydroxycarbamazepine (MHD) in plasma was monitored. A mixed-effects model identified factors influencing MHD exposure. Logistic regression and Cox proportional hazards models were used to analyze the concentration–efficacy relationship, while Kaplan–Meier and time-to-onset analyses were performed to characterize adverse events. Results: Among 824 included patients (1976 concentration samples), body weight, age, treatment duration, and epilepsy type significantly influenced MHD’s exposure levels. The 12-month overall response rate was higher in monotherapy than add-on therapy (82.9% vs. 60.4%). A plasma MHD concentration ≥ 10 μg/mL was identified as a critical “risk transition point” for treatment failure (PSM-adjusted OR = 2.42, p < 0.001). Multivariate logistic analysis confirmed higher concentrations, specific etiologies, and polytherapy as risk factors for inefficacy. Cox regression further revealed that concentrations ≥ 10 μg/mL and specific etiologies were predictors of reduced long-term treatment persistence. Adverse events occurred in 30.5% of patients; for most, the risk did not change over time. Conclusions: This study tentatively proposed a therapeutic reference range (3.0–20.0 µg/mL) of MHD for Chinese children with epilepsy and identified a concentration ≥ 10 μg/mL as a “risk transition point”. The findings provide practical, evidence-based insights for tailoring OXC therapy and managing potential risks.