|Home | About | Journals | Submit | Contact Us | Français|
Chorea has been described as an initial sign of relapse in children with herpes simplex virus encephalitis. We describe the detection of anti‐basal ganglia antibodies (ABGA) in plasma and cerebrospinal fluid (CSF) of a 2.3‐year‐old girl with severe chorea 3 weeks after acute herpes simplex virus (HSV) encephalitis. Although common neuroleptic and antidopaminergic drugs were ineffective, plasmapheresis combined with immunosuppression was followed by rapid and complete neurological recovery. These findings suggest a post‐infectious, immune‐mediated mechanism in this case of chorea after HSV encephalitis.
HSV encephalitis accounts for 10–20% of all viral encephalitis in the US.1 Occasionally, chorea has been described as an initial sign of relapse with often poor prognosis. At least three pathogenic mechanisms are possible: occurence of late‐onset symptoms of the initial viral infection, recurrence of viral replication (owing to incomplete treatment of the initial HSV encephalitis or by selection of clones of aciclovir‐resistant virus), or induction of a deleterious immunoinflammatory reaction.2 Autoimmune‐mediated brain disorders are well known after group A β haemolytic streptococcal infections: for example, Sydenham's chorea or paediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS). Here, we report on a case of severe chorea with positive ABGA 3 weeks after acute HSV encephalitis.
At 2.3 years, a previously healthy girl experienced acute HSV encephalitis with positive type 1 HSV polymerase chain reaction (PCR) in CSF and was treated with aciclovir, 30 mg/kg/day, for 15 days with rapid recovery. Follow‐up HSV PCR in CSF was negative. Cranial magnetic resonance imaging (cMRI) showed bilateral temporal oedema and cortical necroses. Owing to several symptomatic focal seizures, she was discharged on oxcarbamazepine.
Three weeks after the initial onset, she developed a compulsive behaviour, restlessness, disturbed circadian sleep–wake rhythm and, within days, developed severe chorea. Streptococcal serology and antinuclear factors (ANF) as well as HSV 1 PCR and HSV 1 immunoglobulin (Ig) M in CSF were found to be negative, whereas HSV 1 IgG was found to be positive. cMRI showed bilateral cystic temporal lesions and mild, diffuse signal changes in supratentorial white matter; the basal ganglia appeared normal. An electroencephalogram (EEG) showed moderate generalised slowing and focal bitemporal spike wave discharges. Aciclovir (60 mg/kg/day for 21 days) and intravenous immunoglobulins (0.4 g/kg/day for 5 days) were of no benefit. Common drugs were ineffective (fig 11),), and progressive swallowing difficulties necessitated nasogastric tube feeding. Five weeks after onset of chorea, nasotracheal intubation became necessary owing to severe laryngeal and pharyngeal dystonia.
ABGA, determined by immunoblotting in three pairs of samples, showed the presence of a specific, 60‐kDa protein in plasma and in CSF. The antigen used was derived from human caudate and putamen, flash frozen within hours of death. According to these findings, high‐dose intravenous methylprednisolone (20 mg/kg/day for 3 days) was given, but remained clinically ineffective. Intermittent plasmapheresis and systemic immunosuppression resulted in rapid improvement and extubation after 2 weeks. Prednisolone was slowly tapered off over the next 6 weeks and prophylactic oral application of aciclovir (30 mg/kg/day) was stopped. cMRI at that time showed bilateral temporal cysts with surrounding gliosis and moderate global brain atrophy.
Extubation was followed by rapid and full neurological recovery over the following weeks. EEG showed persistent mild slowing over both temporal lobes without epileptic discharges. ABGA could not be detected immediately after the last plasmapheresis and in further plasma samples taken at different time intervals thereafter. Eighteen months after plasmapheresis, the patient experienced no further relapse.
Our patient developed severe compulsive behaviour and chorea 3 weeks after an episode of HSV encephalitis. A relapse was excluded by negative HSV 1 PCR in CSF. Chorea and ballistic movements were resistant to common neuroleptics and antidopaminergic drugs. Nasogastric tube feeding and mechanical ventilation were necessary owing to swallowing difficulties and laryngeal stridor.
In parallel to Sydenham's chorea and PANDAS, we hypothesised that chorea after HSV encephalitis could also be an immune‐mediated process. By means of western immunoblotting, we detected ABGA in three pairs of plasma and CSF. Compared with ELISA and indirect immunofluorescence methods, ABGA western immunoblotting shows the highest sensitivity and specificity.3 In patients with PANDAS, the pathogenicity of ABGA was shown by positive and negative predictive values of 97% and 91%, respectively. This suggests that ABGA may be pathogenic.3
Molecular mimicry may play a major part in Sydenham's chorea as well as in PANDAS. Owing to antigen similarity, antibodies to streptococcal epitopes may cross react with specific basal ganglia antigens and lead to basal ganglia dysfunction or destruction.3 Another hypothesis raised in acute disseminated encephalomyelitis speculates that a disturbed blood–brain barrier may lead to systemic antigen presentation of proteins of the central nervous system and thus induce antibody formation.4 We found similar concentrations of ABGA in plasma and CSF of our patient and thus cannot distinguish between primarily systemic or autochthonous IgG production.
Plasmapheresis eliminates circulating systemic antibodies and was successfully applied in children with obsessive–compulsive disorders and tic disorders. On the contrary, systemic immunosuppression inhibits ongoing systemic as well as autochthonous ABGA production. Thus, a combination of these two approaches is reasonable. In our patient, ABGA were negative in plasma and CSF taken 10 days after initiation of this protocol.
Chorea after or associated with HSV encephalitis was primarily believed to be caused by late‐onset symptoms or relapse of HSV encephalitis.2,5 As HSV 1 PCR was negative in some of the presented cases, it may be speculated that at least some of these patients may have had immune‐mediated chorea after herpes encephalitis. Our patient showed rapid and full neurological recovery within 6 weeks after plasmapheresis, accompanied by systemic immunosuppression.
Arguing against the risk of plasmapheresis in a potentially self‐limiting process, disease duration may play an important part in the outcome. Some patients with HSV 1 PCR‐negative relapse reach complete remission, but in most, relapse leads to additional neurological damage and poor outcome or death.2,5 In our patient, the diffuse signal change in the supratentorial white matter indicated a subtle but widespread inflammatory process accompanying the onset of chorea. cMRI 3 months after the onset of chorea showed moderate brain atrophy. Limitation of the disease process may have led to an improved neurological outcome in our patient.
This, to our knowledge, is the first report on ABGA‐positive chorea after HSV 1 encephalitis. An underlying autoimmune mechanism was suggested by the finding of positive ABGA in plasma and CSF, as well as by rapid clinical improvement after plasmapheresis and systemic immunosuppression. Over an 18‐month observational period, the girl did not show any relapse of compulsive behaviour or chorea.
Competing interests: None declared.
Informed consent was obtained for publication of the patient's details described in this report.