THE PATIENT, a 24-year-old man, presented with symptoms of hyperactivity, overtalkativeness, tall claims, decreased need for sleep and aggressive behavior. He had a history of two manic episodes without any depressive episode, which was diagnosed as bipolar I disorder, and responded to oxcarbazepine then to lithium along with olanzapine. Physical examinations and investigations had been unremarkable throughout. The present episode again improved with lithium 1350 mg/day (serum lithium level 0.91 mEq/L) and olanzapine 20 mg/day.
On follow up there was tremor, ataxia, jerky movements, slurring of speech, impaired finger–nose test, knee–heel test, past pointing and impaired tandem walking, and the serum lithium level was 1.46 mEq/L. Lithium was reduced to 900 mg/day, but the tremors and exacerbated manic symptoms persisted. Computed tomography (CT) of the brain showed an enlarged posterior fossa, with cyst communicating with the fourth ventricle, and hypoplasia of the cerebellar vermis, indicating Dandy–Walker syndrome.
Lithium was stopped and sodium valproate 1500 mg/day was started (serum valproate level, 142.2 µg/mL). In view of persistent manic symptoms, carbamazipine 400 mg/day was added, along with olanzapine 15 mg/day. At the time of writing the patient was well with these medications and his affective symptoms were under control.
In a previously reported case of Dandy–Walker syndrome comorbid with bipolar disorder, that case involved bipolar II disorder,1 whereas the present patient had bipolar I disorder.
Informed consent was obtained from the patient. This is the second reported case of bipolar disorder with Dandy–Walker syndrome. The appearance of cerebellar signs during treatment may be attributed primarily to lithium, because the symptoms appeared while the patient was on lithium for 20 months, and the serum lithium level was 1.46 mEq/L. Lithium is well known for cerebellar toxicity,2 whereas other possible contributory factors may be antipsychotics3 and neurodegenerative processes caused by multiplicity of episodes.4 Various regions implicated in structural neuroimaging in bipolar disorder include anterior cingulate, orbitofrontal cortex, striatum, amygdala and hippocampus.5 The brain CT in the present case did not indicate any prominent abnormality other than Dandy–Walker complex. Thus, the possibility of neurodegenerative processes caused by multiplicity of episodes is least likely.
In the present case, the cerebellar lesions of Dandy–Walker complex most likely contributed to development of recurrent manic episodes through disruption of cerebellosubcortical circuits. Functional, rather than structural neuroimaging, may be more useful in delineating the involvement.