Drug-resistant epilepsy and mortality-Why and when do neuromodulation and epilepsy surgery reduce overall mortality

Patients with drug-resistant epilepsy have an increased mortality rate, with the majority of deaths being epilepsy related and 40% due to sudden unexpected death in epilepsy (SUDEP). The impact of epilepsy surgery on mortality has been investigated since the 1970s, with increased interest in this field during the past 15 years. We systematically reviewed studies investigating mortality rate in patients undergoing epilepsy surgery or neuromodulation therapies. The quality of available evidence proved heterogenous and often limited by significant methodological issues. Perioperative mortality following epilepsy surgery was found to be <1%. Meta-analysis of studies that directly compared patients who underwent surgery to those not operated following presurgical evaluation showed that the former have a two-fold lower risk of death and a three-fold lower risk of SUDEP compared to the latter (odds ratio [OR] 0.40, 95% confidence interval [CI]: 0.29-0.56; p < .0001 for overall mortality and OR 0.32, 95% CI: 0.18-0.57; p < .001 for SUDEP). Limited data are available regarding the risk of death and SUDEP in patients undergoing neuromodulation therapies, although some evidence indicates that vagus nerve stimulation might be associated with a lower risk of SUDEP. Several key questions remain to be addressed in future studies, considering the need to better inform patients about the long-term benefit-risk ratio of epilepsy surgery. Dedicated long-term prospective studies will thus be required to provide more personalized information on the impact of surgery and/or neuromodulation on the risk of death and SUDEP.

Automated summary

Patients with drug-resistant epilepsy have an increased mortality rate, with the majority of deaths being epilepsy related and 40% due to sudden unexpected death in epilepsy (SUDEP). Epilepsy surgery has been found to reduce the risk of death and SUDEP, although the quality of available evidence is variable.