Analysis in risk factors of postoperative recurrence of patients with refractory epilepsy and establishment of a risk prediction model

Objective To explore the risk factors for postoperative recurrence in patients with refractory epilepsy and establish a predictive model.

Methods Clinical materials of 280 refractory epilepsy patients with surgical treatment in the hospital from June 2021 to October 2022 were retrospectively collected, with a follow-up of one year after surgery. These patients were divided into non-recurrence group with 238 cases and recurrence group with 42 cases according to their recurrence status. The risk factors for postoperative recurrence in refractory epilepsy patients were analyzed by univariate and multivariate Logistic regression analyses; the receiver operating characteristic (ROC) curve was drawn to evaluate the predictive value of the model for postoperative recurrence in refractory epilepsy patients.

Results Multivariate Logistic regression analysis showed that disease duration exceeding 5 years, incomplete concordance between the preoperative localization of the lesion and thesurgical site, low serum vitamin B₆ level, and high level of serum monocyte chemoattractant protein-1 (MCP-1) were the significant risk factors for postoperative recurrence in refractory epilepsy patients (OR=2.705, 2.314, 1.790 and 2.284, P < 0.05). A regression model was built based on these findings, and an ROC curve for predicting postoperative recurrence in refractory epilepsy patients was plotted based on the predicted probability logit(P). When logit(P) exceeded 14.52, the area under the curve (AUC) was 0.850, with a sensitivity of 78.57% and a specificity of 80.67%.

Conclusion In refractory epilepsy patients with surgical treatment, disease duration exceeding 5 years, incomplete concordance between the preoperative localization of the lesion and the surgical site, low serum vitamin B₆ level, and elevated MCP-1 level are identified as risk factors for postoperative recurrence. The established regression model for predicting postoperative recurrence in refractory epilepsy patients demonstrates a high predictive value, and can be utilized to identify populations with high-risk of recurrence and guide targeted interventions to reduce the risk of recurrence.