Anti-NMDA-receptor encephalitis "mimicking an autistic regression"

Anti-NMDA-receptor encephalitis is not something that I ever envisaged talking about so much on this blog primarily concerned with autism research. Describing an often severe form of encephalitis where the body mounts an immune response against self ("the NR1 subunit of the NMDA receptor"), this condition is more readily associated with the symptoms of psychosis than anything specifically autism-related.

But yet again (see here and see here) I'm talking about peer-reviewed research suggesting that when one talks about the appearance of 'autistic regression' one might consider testing for anti-NMDA-receptor encephalitis just to rule it out as a possible cause. This time around the findings are those reported by Yael Hacohen and colleagues [1] and their description of "two toddlers who presented with developmental regression, particularly of their social communication skills, mimicking an autistic regression, who were found to have NMDAR-Ab [N-methyl-d-aspartate receptor antibodies] in the serum and cerebrospinal fluid." Aside from various "other neurological features" hampering timely diagnosis of anti-NMDA-receptor encephalitis, the authors report that subsequent "immunotherapy was beneficial in both patients, with significant improvement of their language skills and behaviour."

A few more details are worthy of discussion: Case 1 talks about a 2-year old girl who following a typical developmental profile experienced an "acute onset of behavioural change, disrupted sleep, and loss of motor, language and social communication skills." Even after some investigation for "an organic brain syndrome" everything came out as within typical ranges despite her continuing to regress after initial discharge. Following a second admission, during which time "she was mute and had orofacial and right hand sterotypies" she was found to have positive NMDAR-Ab and treatment was initiated in the form of methylprednisolone. We are told: "At follow-up, aged 4 years, 2 years from symptom onset, she is well, has age-appropriate language acquisition, no behaviour and psychiatric concerns, and remains relapse-free."

Case 2 was another young girl also aged 2 years who following a similar pattern to case 1 followed a typical pattern of development that was quite suddenly replaced with "abnormal behaviour" including a loss of speech, biting, pinching and "banging her head." Again, most parameters were in the typical zone when it came to assessments, although her sleep EEG "revealed occasional focal and generalized epileptiform discharges." Those NMDAR-Ab were detected and treatment initiated. Interestingly, authors reported that methlyprednisolone only provided short-term improvements for this child and so more aggressive treatment (rituximab and then mycophenolate mofetil) was put in place. The longer-term prospect for this child also did not seem to mirror case 1 as we are told that despite improvements in speech for example, "she continues to suffer from ongoing behavioural concerns and sleep disruption."

Combined with those other reports that dot the autism research literature, anti-NMDA-receptor encephalitis does seem to be at least one probable 'cause' of autistic regression. The important idea that one can screen for the presence of those antibodies (although looking in cerebrospinal fluid is quite invasive) also opens up a role for medicine and investigation when such a clinical outcome presents. 

And then there is the idea that when detected alongside certain autistic or autistic-like symptoms this type of encephalitis might be 'treatable'. No matter what your views are on autism or which 'kingdom' you subscribe to, I don't think anyone would truly suggest that a medical diagnosis of encephalitis should not be treated in an appropriate and timely manner. Indeed, various interventions could be indicated for such encephalitic cases [2] including the use of IVIg and methylprednisolone, all under appropriate medical supervision (I say this without any medical or clinical advice given or intended). Corticosteroid 'therapy' for regressive autism for example, has been previously discussed in the peer-reviewed literature (see here) so there may be quite a bit more research to do in this area. I might also add that none of the treatments discussed in the Hacohen paper are without potential side-effects so caution and medical input is a must.

Regression occurring in cases of autism is, at last, moving out of the shadows and becoming more readily accepted these days. I would hope that reports such as the one from Hacohen and colleagues are finally moving the conversation on in terms of what might cause instances of regression (see here for another example) and what interventions might be possible, all set within the context of the very plural and very heterogeneous autisms.

I will leave you with a few final sentences from Hacohen et al: "NMDRAR-Ab should be tested in cases of regression of social and communication skills with additional neurological symptoms such as movement disorder or seizures. Unlike autism, early diagnosis and treatement of NMDAR-Ab encephalitis is associated with much improved outcome." Make of that what you will...

----------

[1] Hacohen Y. et al. N-methyl-d-aspartate (NMDA) receptor antibodies encephalitis mimicking an autistic regression. Developmental Medicine & Child Neurology. 2016. June 3.

[2] Luca N. et al. Anti-N-Methyl-D-Aspartate Receptor Encephalitis: A Newly Recognized Inflammatory Brain Disease in Children. Arthritis and rheumatism. 2011;63(8):2516-2522. 

----------

Hacohen Y, Wright S, Gadian J, Vincent A, Lim M, Wassmer E, & Lin JP (2016). N-methyl-d-aspartate (NMDA) receptor antibodies encephalitis mimicking an autistic regression. Developmental medicine and child neurology PMID: 27255282