In recent years, neurology and psychiatry journals have been inundated with reports on individual symptoms of autoimmune encephalitis (AE) that are described as distinct entities such as autoimmune psychosis, obsessive-compulsive disorders, or depression. It is unquestionable that for AE the demonstration of antibodies against neuronal-surface proteins is intrinsically linked to distinct disorders (some defining new diseases) that are usually treatment-responsive and associate with comorbidities that vary according to the antigen. By contrast, for psychiatric diseases, the apparent detection of antibodies has not defined any disorder or affected the diagnosis and treatment of patients. Although these studies frequently use anti-N-methyl-D-aspartate receptor encephalitis to rationalize the findings, they rarely adopt the same rigorous investigations or address the clinical and pathogenic significance of the antibodies or discuss the limitations related to the biological sample or antibody-testing techniques. It is imperative to consider (1) some antibodies (GAD65, TPO) occur in serum of 8%–13% of healthy people; (2) VGKC antibodies are not useful unless LGI1 or CASPR2 are investigated; (3) commercial-clinical testing for Ma2, Zic4, and SOX1 antibodies causes a high number of false-positive results; (4) GlyR antibodies have unclear disease specificity when examined only in serum; and (5) the significance of antibodies against unknown antigens of endothelium, astrocytes, myelin fibers, or granule cells of hippocampus and cerebellum is questioned by the lack of disease specificity and appropriate controls. These limitations and problems are a frequent cause of neurologic consultations. Here we discuss some of these problems, emphasizing the importance of clinical judgment over antibody findings.
