Abstracts

The multicenter epilepsy surgery study has prospectively tracked patients with localization-related epilepsy treated either medically or surgically for many years. Consequently, an opportunity exists to assess mortality and influences on death rates in medically refractory patients spanning different referral populations in the U.S. We assessed risk of death in refractory epilepsy and whether mortality differs in medical and surgically treated patients. Seven epilepsy centers enrolled patients in a study of outcome after resective epilepsy surgery. While most patients had surgery, some individuals did not and were followed on medical therapy. Patients were prospectively observed for up to 7.7 years after enrolling in the study. Seizure outcome was assessed by office visits and telephone calls. Mortality and cause of death were determined by direct contact and searching the Social Security index. Survival analyses were performed with the product-limit method and mortality rates per 1000 person-years (p-y) were calculated along with confidence intervals (Wald method). 532 patients had adequate data for analysis. 388 had surgery and 144 had medical treatment. 282 were female, 250 were male, and mean age at enrollment was 37 yrs, (range 13.8-66.2 yrs). Mean follow-up duration was 4.5 yrs (range 0.27 to 7.7 yrs). Mortality was higher in medical than surgically treated patients (p=.002, Cox-Mantel). In the medical group, 11 patients (7.6%, 95% CI = 4.2-13.3%) died, a death rate of 20.3 per 1000 p-y. In the surgical group, 11 patients (2.8%, 95% CI = 1.5-5.1%) died during follow-up, a death rate of 6.0 per 1000 p-y. Most had recurrent seizures. Mortality in medical patients exceeded that of surgical patients who recurred postoperatively (p [lt] .01). Of deceased medical patients with known follow-up, all had recurrent seizures. Causes of death were known in 13 patients and included probable SUDEP in 5, seizure-related in 4, suicide in 3, and cancer in 1 patient. As expected, refractory epilepsy is associated with high death rates. Mortality appears to be lower in surgically treated patients than medically treated patients. This is most likely related both to elimination of seizures and inherent disease severity. Further exploration of specific risk factors is needed. (Supported by National Institutes of Health)