Abstracts

Rationale: Autoimmune encephalitis and epilepsy remain underrecognized and undertreated; however, seizures may respond to immunotherapy rather than conventional anti-seizure drugs (ASDs) [Toledano 2015]. The goal of this study is to validate Dubey’s Antibody Prevalence in Epilepsy (APE) score and Response to Immunotherapy in Epilepsy (RITE) score in a cohort of patients with a diagnosis of autoimmune encephalitis with epilepsy [Dubey 2017]. Methods: We conducted a search through a clinic database and an electronic medical record cohort discovery tool and retrospectively identified (n=48) adult subjects at Stanford University Hospital between 2008 – 2018 with an ICD diagnosis of limbic encephalitis, lab testing of paraneoplastic antibody (Ab) panel, and procedure codes for EEG. Subjects were classified as having possible, probable or definite autoimmune encephalitis using diagnostic criteria from Graus [2016]. Subjects were excluded if they did not have a diagnosis of epilepsy (n=3), did not have an Ab panel (n=3), did not fulfill diagnostic criteria for autoimmune encephalitis (n=10), had an alternate diagnosis of Rasmussen’s encephalitis (n=2). We calculated Dubey’s APE and RITE scores via chart review for the included subjects (n=30) and the scores’ sensitivity and specificity. Responders to immunotherapy had a >50% reduction in seizures. If there was no clear response to immunotherapy on chart review, the case was classified as a non-responder. Results: Among the 30 subjects, median age of onset was 34 years old (SD ±24) and the median duration of epilepsy was 2 years (SD ±11). There were 16 females (53%). Eighteen subjects were classified as definite autoimmune encephalitis with epilepsy (60%), probable diagnosis but Ab negative (n=4, 13%) and possible diagnosis (n=8, 27%). Antibody panels from both serum and CSF were tested in 87% of subjects. Overall, 77% of the subjects had positive Ab (n=23). CNS specific Ab were positive in 57% (n=17) cases after non-specific Ab were excluded for isolated VGKC channel Ab (n=4), isolated anti-TPO Ab (n=1), or Ach R muscle Ab (n=1). In our cohort of autoimmune encephalitis with epilepsy patients, 27/30 (90%) had APE scores ≥ 4 and 21/30 (70%) had RITE scores ≥ 7. The APE score had a 94% sensitivity to predict CNS specific Ab positivity and 17% specificity; the RITE score had a 63% sensitivity and 60% specificity to predict response to immunotherapy. Conclusions: We validated the sensitivity Dubey’s APE and RITE scores in our Stanford cohort of patients with autoimmune encephalitis with epilepsy. The high sensitivity of the APE and RITE scores makes them useful tools to decide which patients to pursue Ab testing and which may potentially benefit from a trial of immunotherapy for epilepsy control. All of our subjects had an established diagnosis of autoimmune encephalitis with epilepsy which contributes to the low specificity of the scores; only 3 of our cases had an APE score < 4. Isolated VGKC Ab were positive in 3 cases tested in prior to 2014, however, it is possible that they may have CNS specific Ab if samples of their serum or CSF were available for testing today. Responder rate to immunotherapy was difficult to assess often because both ASDs and immunotherapy were given around the same time period. Because of the small numbers of cases, prospective and multi-center pooled research through national/international databases on autoimmune encephalitis with epilepsy are needed. Funding: No funding