Abstract
To the Editor
Valproate is now widely used in psychiatry for an extensive range of conditions. Side effects familiar to all clinicians include weight gain, gastrointestinal symptoms, sedation, tremor and mild elevation of hepatic enzymes. Many of these are dose-related and are promptly resolved with dose reduction (Haddad et al., 2009). A number of rarer side effects have also been reported, including valproate-induced hyperammonemic encephalopathy (VHE), the exact incidence of which is unknown (Carr and Shrewsbury, 2007). VHE can be a fatal condition but with prompt recognition, it can be treated with full recovery (Carr and Shrewsbury, 2007). This condition is widely recognised in the neurological literature – it is less well documented in psychiatry but has previously been described in this journal (Abreu et al., 2009).
We present here a case history of a patient who presented with cognitive changes and was diagnosed with VHE. She was a 53-year-old woman with an established history of schizoaffective disorder, maintained on valproate and quetiapine. She was admitted to a medical ward after taking an overdose of quetiapine. After 4 days she was transferred for psychiatric treatment and restarted on quetiapine, thyroxine and sodium valproate 1000 mg twice a day, which produced a therapeutic level.
On the psychiatric ward she was agitated, disorganised and disoriented. She was extensively investigated for apparent delirium over a period of 2 weeks, during which time her mental state remained unchanged. Most tests, including magnetic resonance imaging (MRI) and electroencephalography (EEG), septic screen and valproate level were normal but the serum ammonia level was found to be 123 µmol/L. This was thought to be consistent with an iatrogenic encephalopathy secondary to the rapid resumption of valproate. The valproate was ceased, followed by a rapid improvement in the patient’s mental state and a return of the ammonia level to the normal range.
After the patient’s mental state stabilised, valproate was reintroduced cautiously. However, the patient again became confused, hostile and agitated. Valproate was abandoned with a recommendation that it should not be prescribed in the future.
The pathogenesis of VHE is unclear but is thought to be due to raised ammonia from liver dysfunction. Hyperammonemia then causes decreased levels of consciousness (Gomceli et al., 2007). VHE tends to occur within days or weeks of valproate initiation but has been reported to occur months later (Abreu et al., 2009; Gomceli et al., 2007).
As in this case, retrial of valproate is likely to be unsuccessful (Abreu et al., 2009).
The typical clinical presentation of VHE includes an acute onset of impaired consciousness with increasing confusion and lethargy, neurological signs or symptoms and increased seizure frequency (Gomceli et al., 2007). Behavioural changes may occur but should be differentiated from post-seizure effects or psychiatric symptoms of the condition being treated (Gomceli et al., 2007). Laboratory tests include serum valproate level, liver function tests (LFTs), serum ammonia level and computed tomography (CT) head with EEG (Gurjar et al., 2011). For a diagnosis of VHE to be made, the serum ammonia level should be raised, LFTs may be normal but, more importantly, improvement in clinical features can be observed with valproate withdrawal or dose adjustment (Gurjar et al., 2011).
In conclusion, the diagnosis of VHE should be suspected in any patient on valproate therapy presenting with confusion. Early recognition of subtle cognitive and behavioural signs can lead to therapeutic intervention with satisfying results. Retrial of valproate is unlikely to be successful and should be carried out cautiously if there is a strong clinical need.