Sodium valproate toxicity

Question 7 from the second paper of 2003 asked about the features and the management of valproate overdose. This is hard, because there are no characteristic clinical features in this overdose. There is non-specific lethargy which progresses to drowsiness and coma. Then, the LFTs come back deranged, and when you do the ammonium level it is through then roof, which makes you think. Question 9.1 from the first paper of 2017 in fact presents such a raised ammonium level, and then asks for a list of potential causes. 

Pathophysiology

• Sodium valproate is a simple branched-chain carboxylic acid, the antiepileptic properties of which were discovered quite by accident (it was the solvent used to dissolve water-insoluble bismuth salts during a series of preclinical animal experiments in 1962)
• Mitochondrial β-oxidation of valproate  involves "the carnitine shuttle", which leads to the depletion of carnitine.
• In the absence of carnitine, the liver resorts to an altertative metabolic pathway, which produces 4-en-valproic acid, a hepatotoxin.
• Hyperammonaemia develops, as valproate promotes the transport of glutamine through the mitochondrial membrane, and ammonia is released as a result of the mitochondrial metabolism of glutamine into glutamate.
• The hyperammonaemia then gives rise to cerebral oedema, some 72 hours post ingestion.

Diagnostic features

• hypotension
• hypothermia
• CNS depression
• tremor

Complications of valproate overdose

• Lactic acidosis
• hyperammonaemia and encephalopathy
• acute hepatic failure
• pancreatitis
• cerebral oedema
• hypernatremia
• Hypocalcemia
• Hypocarnitinemia, if you actually test for carnitine
• Bone marrow suppression

Drug levels

• 50-100mg/L = therapeutic
• 100-450mg/L = mild toxicity (drowsy, confused)
• 450-1000mg/L = severe toxicity (usually comatose)
• Over 1000mg/L = indication for dialysis

Treatment

• Supportive management (ventilation, vasopressors, etc)
• gastrointestinal decontamination with charcoal or whole bowel lavage
• L-Carnitine supplementation: the loading dose is 100 mg/kg IV over 30 minutes (maximum 6 g) followed by 15 mg/kg IV over 10–30 minutes every 4 hours until clinical improvement occurs. At least some of the acute ammonia-induced encephalopathy seems to be due to a carnitine deficiency, as valproate metabolism depletes the stores of carnitine. Replacement of carnitine seems to be the unquestioned dogma in valproate overdose. Lheureux et al (2005) examined the evidence behind this practice, and found that usefulness in overdose probably does not justify routine supplementation.
• Valproate is 90% protein bound and therefore poorly cleared by dialysis, but ammonia is, and therefore haemodialysis is indicated to prevent cerebral oedema. Moreover, the protein binding is saturable, and in gross overdose you can still bring the valproate levels down significantly with high intensity haemodialysis. Bellomo et al (2009) found that clinical improvement was more rapid with this strategy than with supportive care alone.

One available case report is an account of a truly massive (25g) valproate overdose, which did not require anything but supportive management, and which was not accompanied by any sort of massive organ system failure. However, the college probably wanted to talk about ammonia clearance by dialysis and EVD insertion for cerebral oedema.