Does epilepsy surgery really lower mortality?
Ryvlin P and Montavont A (2008) Does epilepsy surgery really lower mortality? Neurochirurgie 54:3 282–6.
Abstract: Patients with epilepsy suffer from a two to three fold increased death rate as compared to age and sex matched control population. This increased risk culminate to five fold in patients with drug resistant partial epilepsy eligible for epilepsy surgery, with the majority of deaths classified as sudden unexpected death in epilepsy (SUDEP). The pathophysiology of SUDEP remains uncertain, but all witnessed cases occurred during or immediately after a seizure. Several studies have evaluated the impact of epilepsy surgery on the risk of seizure related death and SUDEP. Four series have concentrated on operated patients, and have compared the death rates in those seizure free and non seizure free post-operatively. Three of these studies reported a significantly lower risk of SUDEP in patients cured by surgery as compared to those still seizing. Four other series have compared the mortality in surgically versus medically treated patients with refractory partial epilepsy. Three of these studies failed to show any significant difference in death or SUDEP rates between operated and and non operated patients. All the above series suffer various types of methological limitations, hampering any definite conclusion regarding the impact of epilepsy surgery on mortality. The launching of novel and large multicentric studies, which address the pitfalls of prior series, should allow to provide conclusive results within the next three years.
- Discussion of 4 studies comparing mortality between patients who responded to surgical treatment and those who did not, and 4 studies comparing medical and surgical treatment of epilepsy. The authors comment on interpretations of both the treatment modality and treatment effectiveness as predictors of SUDEP risk, raising concerns about methodology in all of the studies discussed. The methodological limitations cited by these authors were also addressed by Persson et al.; see annotation at that article for additional discussion.