Evolution and significance of neuronal surface autoantibodies after Japanese encephalitis

Abstract

NMDAR-antibody encephalitis can arise as a post-infectious ‘relapse’ following HSV encephalitis. We asked whether a similar condition might occur after Japanese Encephalitis (JE). Cell-based assays for antigen-specific antibodies and IgG binding to the surface of live hippocampal neurons were performed on 13 CSFs and 65 sera, many sampled longitudinally, from 34 Vietnamese children with JE. Three month outcomes were scored according to a pediatric post-encephalitis scale; clinical features focussed on new onset symptoms during the follow-up, blinded to antibody results. Ten/34 children (29 %) had serum antibodies against known neuronal-surface antigens, 4 NMDAR, 4 CASPR2, 1 GABAAR and 1 LGI1, detected from day 23 after onset. In addition, 8/10 (80 %) of these children and 13/24 (54 %) others had antibodies that bound in a distinctive pattern to live hippocampal neurons (HN-Abs), four detected on days 9–12. A relapse was only considered possible in 5 patients, two with specific neuronal surface antibodies (NSAbs). Neither specific NSAbs (p = 1.00) nor HN-Abs (p = 1.00).were more common in patients with possible relapse than in those with unlikely relapse. There was a modest trend towards worse outcome scores in patients with known NSAbs than in the remaining patients (p = 0.089). Antibodies to neuronal surface proteins, known and unknown, are common in children after JE. Caution is urged in defining post-infectious autoimmune encephalitis on the basis of a positive neuronal antibody and new onset symptoms or deterioration following recovery from JE, or in initiating immunotherapy without confirmatory evidence of a time-dependent encephalitic illness defined by published guidelines.