[Limbic encephalitis--history,symptoms,and the latest classification].

The concept of limbic encephalitis has changed over time. Since the introduction of "limbic encephalitis" (LE) in 1968, LE was thought to almost always be associated with carcinoma; this belief led to the coining of the term "paraneoplastic limbic encephalitis" (PLE). In the 1990s, antineuronal antibodies, including anti-Hu and anti-Ta/Ma2, were detected; this supported the hypothesis of an autoimmune mechanism for PLE. The prognosis of patients with PLE was, however, poor. Since 2001, there have been reports of patients with LE exhibiting antibodies to the voltage-gated potassium channel; this observation is intriguing because in such cases the encephalitis was usually independent of carcinoma, and its clinical course was often reversible. Since the 1990s, cases of non-herpetic acute limbic encephalitis have been reported in Japan. In some of these cases, an autoantibody to GluRepsilon2 (NR2B) has been detected; GluRepsilon2 is a subunit of the N-methyl-D-aspartate (NMDA) glutamate receptor found in the limbic forebrain. A postulated pathophysiologic role of this antibody led to the concept of autoantibody-mediated acute reversible LE (AMED-ARLE). In 2007, some patients with ovarian teratoma developed encephalitis and exhibited antibodies to the NMDA receptor; this antibody is thought to recognize NR1/NR2 heteromers. Later, anti-NMDA receptor antibodies were also detected in some Japanese patients who had been previously diagnosed with juvenile acute non-herpetic encephalitis. Currently, limbic encephalitis is categorized into 3 groups: limbic encephalitis caused by virus infection, autoantibody-mediated limbic encephalitis (AMLE), and limbic encephalitis with autoimmune disease. In AMLE, antibodies to cytoplasmic antigens cause classical PLE (type I). In contrast, antibodies to cell membrane antigens often cause reversible limbic encephalitis in patients with (PLE type II) or without tumors (AMED-ARLE).