Abstracts

Rationale: The risk of sudden unexpected death in epilepsy (SUDEP) might be reduced by augmenting treatment of patients with refractory seizures (Ryvlin et al. 2011). Accordingly, two studies showed that SUDEP rates decreased after two years of adjunctive VNS Therapy, as compared to the first two years of treatment (Annegers et al. 2000, Granbichler et al. 2015). Over 40,000 patients with drug-resistant epilepsy (DRE) have been implanted in the United States where more extensive implant tracking is performed. The current study aimed at revisiting the SUDEP rate in patients with VNS Therapy in a much larger population than those previously reported.Methods: This retrospective study included all US patients with DRE who were implanted with VNS from 1988‐2012 with known social security numbers. Exposure to VNS was calculated from date of implant until death, device explant, known date at which the device was disabled or the last follow‐up date of December 31, 2012. Date and cause of death were obtained from the Social Security Death Index Master File and the National Death Index (NDI) which provides ICD-9 or ICD-10 codes. An algorithm was developed for SUDEP case ascertainment according to Annegers’ SUDEP classification leveraging the NDI cause-of-death codes along with any additional patient information. Both crude and age-adjusted (directly standardized using the U.S. Standard Population for year 2000) SUDEP rates were calculated per 1000 PY of exposure. Crude rates were stratified by baseline characteristics, including gender and age at implant and death, and rate ratios were calculated along with 95% confidence intervals.Results: A total of 40,443 patients (mean age at implant: 30.8 years) with a median follow-up of 7.6 years, representing 277,661 PY, were included. A total of 3,689 deaths were recorded. 953 (26%) corresponded to cause-of-death codes that potentially include SUDEP. The case ascertainment algorithm identified 660 (69%) as possible/probable/definite SUDEP, a crude rate of 2.38/1000 PY (Male: 2.67, Female: 2.07) (Table 1). The age-adjusted rate was 2.18/1000 PY (Male: 2.54, Female: 1.81) with a statistically significant 29% reduction after the first two years (95% confidence interval for rate ratio and Mann-Kendall trend test p=0.01). SUDEP determination is less reliable for patient deaths outside the 10-64 age range. A sensitivity analysis restricting to the 10-64 age range showed a crude rate of 2.49/ 1000 PY (Male: 2.81, Female: 2.16). The age-adjusted rate decreased 19% after the first 2 yearsConclusions: This represents, to our knowledge, the largest study of SUDEP in patients with drug-resistant epilepsy, with an 80-fold increase in available PY of follow-up from previous VNS Therapy studies. All three studies show a reduction in SUDEP over time, with the present study showing a statistically significant 29% reduction after the first 2 years.