Risk factors and a nomogram model for drug-resistant epilepsy (DRE) in Asian population

Yangju Chen; Rong Tang; Zhihai Shao; Wei Li; Guoqiang Fei; Xin Wang

doi:10.1016/j.clineuro.2025.109009

Risk factors and a nomogram model for drug-resistant epilepsy (DRE) in Asian population

Clin Neurol Neurosurg. 2025 Jun 5:256:109009. doi: 10.1016/j.clineuro.2025.109009. Online ahead of print.

Authors

Yangju Chen¹, Rong Tang¹, Zhihai Shao¹, Wei Li¹, Guoqiang Fei², Xin Wang³

Affiliations

¹ Department of Neurology, Zhongshan Hospital (Xiamen), Fudan University, Xiamen, China.
² Department of Neurology, Zhongshan Hospital (Xiamen), Fudan University, Xiamen, China; Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai, China. Electronic address: fei.guoqiang@zsxmhospital.com.
³ Department of Neurology, Zhongshan Hospital (Xiamen), Fudan University, Xiamen, China; Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai, China. Electronic address: wang.xin@zs-hospital.sh.cn.

PMID: 40570754
DOI: 10.1016/j.clineuro.2025.109009

Abstract

Objective: Epilepsy is one of the most widespread neurological diseases, affecting millions globally. Despite the effectiveness of antiepileptic drugs (AEDs), 20-30 % of patients develop into drug-resistant epilepsy (DRE). Identifying early risk factors for DRE could guide timely interventions. This study targeted to build a nomogram model for predicting DRE in Asian epilepsy patients based on clinical and imaging risk factors.

Methods: A retrospective cohort study was carried out on 452 epilepsy patients treated at Zhongshan Hospital (Xiamen) from Aug 2018 to Aug 2023. Patients were classified into drug-sensitive epilepsy (DSE) and DRE groups. Data on clinical characteristics, VEEG, and MRI findings(a total of 20 parameters) were analyzed. Univariate and multivariate logistic regression identified significant risk factors for DRE, which were incorporated into a nomogram. The model's function was validated using receiver operating characteristic (ROC) curves, concordance index (C-index), and calibration plots.

Results: Of the 452 patients, 122 (27 %) were classified as having DRE. Multivariate analysis identified five independent risk factors: age of onset, abnormal MRI findings, seizure duration, seizure frequency, and multiple seizure types. The nomogram demonstrated good predictive fidelity, with an AUC of 0.88 in the training set and 0.83 in the validation set, and calibration plots testify strong agreement between predicted and inspected outcomes.

Conclusion: The developed nomogram provides a worthwhile tool for predicting DRE in Asian patients, aiding early intervention strategies. Moreover, the model reveals highly predictive accuracy. Hence, the nomogram prediction model can serve well in the promptly distinction of DRE in Asian.

Keywords: Asian population; Drug-resistant epilepsy (DRE); Drug-sensitive epilepsy (DSE); Nomogram; Risk factors.