Common criteria for electroencephalographic evaluation in patients with disorders of consciousness.

Hannelore Ehrenreich, MD, DVM and Johann Steiner, MD In their Letter to the Editor, referring to our recent publication, Lancaster and colleagues defend the construct of N-methyl-Daspartate receptor (NMDAR) encephalitis. This defense is somewhat surprising, because we do not address this topic in our article. We investigated seroprevalence of a broad range of autoantibodies (AB), directed against 24 different brain antigens, among them NMDAR1-AB, in a large number (N5 4,236) of healthy and neuropsychiatrically ill subjects. We found comparable presence of all of these AB in serum of both healthy and ill subjects, with respect to immunoglobulin (Ig) class and titer. Findings about Ig class in this work are presented both as total Ig as well as separately for IgG, IgA, and IgM. Even if all 1:10 titers were completely ignored in our data set (although scientifically solid evidence for such a cutoff is lacking), the overall titer distribution among healthy and ill subjects still remained similar. In all our attempts so far to compare functionality of NMDAR1-AB of the various Ig classes in vitro and in vivo, we did not identify any fundamental differences (Hammer et al and unpublished data), although they may well exist with respect to discrete biological effects at the cellular or systems level. We are not aware of any other publication that delivered a scientifically convincing back-to-back comparison of the various Ig classes of NMDAR1-AB regarding functionality. Once again, we are not claiming in this article that NMDAR encephalitis does not occur, but we show that anti– NMDAR1-AB are not always pathogenic. Therefore, based on our data, clinicians should be highly cautious with respect to any conclusions on a causal association of serum AB with brain disease. For more detailed information, the reader is kindly referred to our original publication. As said earlier, the presence of these AB in the blood circulation does not allow any firm assumption as to whether they play a pathophysiological role in any brainrelated syndromes, and certainly does not on its own justify immunosuppressive treatment, unless these AB are also proven to be present at substantial levels in the cerebrospinal fluid. AB of all types cross into the central nervous system at all times, for example, through the circumventricular organs (which lack a blood–brain barrier), and in the case of IgG are present in normal cerebrospinal fluid at about a 1:500 dilution of their blood concentration (IgA, 1:600; IgM, 1:3,000). Rather than trying to turn back the clock, we should learn more about these serum AB that are directed against brain antigens but not necessarily associated with any disease. These AB likely modify our brain functions if the blood–brain barrier becomes temporarily or persistently compromised.