Is Schizophrenia an Autoimmune Disease?

Psychopathology and particularly psychosis has had a bit of a research dance with immunology over the past several years. For example, women with post-partum psychosis are more likely than controls to have anti-thyroid antibodies. And folks with schizophrenia and bipolar disorder are more likely to have strange anti-wheat protein antibodies than controls. In the recent, very large CATIE trial, 23.% of those with schizophrenia had IgA anti-AGA antibiodies (anti-gliadin) compared to 3.1% of a comparison group, and 5.4% had high levels of tTG antibodies compared to 0.8% of the comparison group.

Brahms Violin Concerto (very long, really famous bit begins at around minute 35)

No one is sure what these immune reactions mean. But it would be interesting to see how immune modulators might affect psychosis in a clinical trial. In evolutionary medicine, immune and inflammatory modulators could include a dietary intervention, probiotics, or even helminth therapies. To my knowledge, none of these have been applied to schizophrenia or post-partum psychosis in a clinical trial of any kind.

This week, a paper came out in the renamed Archives of General Psychiatry (Now JAMA Psychiatry) linking schizophrenia to a set of autoantibodies. The findings in this paper lend more credence to the idea that a subset of schizophrenia may be caused by an immune attack on the brain.  Blood from a group of unmedicated, hospitalized schizophrenics was compared to blood from people admitted with major depressive disorder, borderline personality disorder, and healthy controls.

9.9% of the actuely ill schizophrenics were found to have anti-NMDA receptor antibodies, compared with 2.8% of those with major depressive disorder, 0.4% of controls, and 0 of those with borderline personality disorder. The NMDA receptor (glutamate is the key neurotransmitter at this receptor) is known to be associated with psychotic symptoms. PCP and ketamine are NMDA receptor antagonists that rather famously cause agitation and psychosis.

Now there is already an illness of anti-NMDA receptors called "NMDA-R encephalitis." It affects  young women with a rare type of ovarian tumor called a teratoma, and presents with psychosis, agitation, memory problems, and seizures. It tends to progress to problems with the autonomic nervous system (which can control breathing, temperature and blood pressure regulation) and cause a catatonic state. It is treated, like many life-threatening autoimmune conditions, with high dose steroids and plasmaphoresis (or plasma exchange, which can clear the blood of the offending autoantibodies). The autoantibodies in the cases of NMDA-R encephalitis are to a different specific protein subunit of the receptor and tend to be in much higher concentrations than the folks with autoantibodies who had acute schizophrenia, so it is not exactly the same disease.  In this trial, however, two of the patients originally diagnosed with schizophrenia were re-diagnosed as NMDA-R encephalitis due to the type of antibodies they had. They also had some intriguing physical symptoms and CNS and blood  inflammatory markers that aren't typically found in schizophrenia.

But it is fascinating and needs to be studied in more populations at greater length. Is there a time coming when 10% of our first break psychosis patients might be getting plasma exchange and steroids? Would they be maintained on autoimmune dietary protocols (if effective for blood titres of antibodies) and relatively benign chronic immune modulators (again, just hypothesizing in an exciting sort of way) such as pig whipworm or killed M vaccae?

As always, more questions than answers, but getting one step closer to the bottom of the pathology of mental illness and brain diseases is always interesting, and always gives me hope. And what about the healthy control and the patients with major depressive disorder who had anti-NMDA-R antibodies? Are they more likely to have problems with psychosis or psychopathology? I suppose we will have to wait and see.