Dale et al (Reference Dale, Heyman and Giovannoni2005) found high levels of anti-basal ganglia antibodies (ABGA) in the sera of children with obsessive-compulsive disorder (OCD) compared with control groups of children with streptococcal infection without OCD, paediatric autoimmune disease and neurological disorders (stroke, movement disorders and encephalitis) and concluded that central nervous system autoimmunity may play a role in a significant subgroup of children with OCD.

Recently, we found another auto-antibody, anti-phosphatidylethanolamine (aPE), which may have been associated with the sudden onset of OCD in a 5-year-old girl. Six weeks prior to showing symptoms of OCD, the girl was diagnosed with an ear infection, for which she received a full course of antibiotics. She presented at our clinic 2 months after the onset of OCD symptoms. Past medical history was significant for recurrent ear infection. Physical and neurological examinations were normal; no tics were observed. There was no family history of OCD.

At the index visit, the patient was negative for streptolysin O antibody. Throat cultures were negative for Streptococcus pyogenes and Streptococcus group A antigen. A test for deoxyribonuclease B, a marker for prior streptococcal infection, was negative.

To investigate an autoimmune diathesis, the patient was tested for IgG, IgA and aPE, anti-phosphatidylserine, anti-phosphatidylcholine and anti-cardiolipin antibodies (Reference Sokol, McIntyre and ShortSokol et al, 2000). Serial anti-phospholipid antibody testing revealed the persistence of IgG aPE antibodies; aPE antibody levels were coincident with the expression of OCD symptoms. The index and day 113 sera were also positive for IgG anti-phosphatidylserine antibodies. The patient was begun on a low dose of sertraline and her OCD improved.

We believe that this patient has a ‘paediatric autoimmune neuropsychiatric disorder associated with streptococcal infection’ (PANDAS-like condition) because the criteria, except for evidence of a group A streptococcal infection, were met. She had a history of repeated ear infections but her OCD symptoms occurred after the most recent infection. Without documenting the infectious agent, the elevated levels of aPE antibody suggest that she mounted an autoimmune reaction following another ear infection which led to the development of OCD.

We have found aPE antibodies in other neuropsychiatric conditions. An adolescent girl with a basal ganglia stroke had IgA aPE antibodies in her serum and IgG and IgA aPE antibodies in her cerebrospinal fluid; she experienced seizures and depression subsequent to the stroke (Reference Sokol, McIntyre and ShortSokol et al, 2000). Furthermore, aPE was the most frequently detected anti-phospholipid antibody in the serum of patients with psychosis (Reference O'Brien, Sokol and WagenknechtO'Brien et al, 2004). One-third of cerebrospinal fluid samples from this group contained IgG aPE antibody in the absence of this antibody in serum, suggesting intrathecal synthesis. We propose that aPE antibody may attack the basal ganglia, leading to its association with OCD and other disorders of the brain.

Although we report the finding of aPE antibodies with OCD in a single patient, we believe that aPE antibody should be considered as an additional autoimmune marker in post-infectious OCD.