To the Editor:
Ziyeh et al described a patient with valproate-induced hyperammonemia globus pallidus changes on MR imaging, and changes on 1H MR spectroscopy (1). Based on a report by Coulter and Allen (2) they conclude that the patient's encephalopathy was secondary to the valproate-induced hyperammonemia.
Based on the same report by Coulter and Allen we measured plasma ammonia concentrations in 55 assymptomatic outpatients receiving valproate (VPA), Twenty-nine (53%) of these patients had ammonia concentrations above the range of normal, the highest being 140 micromoles/liter (3). Followup showed fluctuating concentrations of ammonia, with spontaneous returns to normal, despite continuation of VPA therapy. We concluded that the patients reported by Coulter and Allen may have had alternative explanations for their encephalopathies.
The same problem arises in the report by Ziveh et al. Their patient had a plasma ammonia concentration only slightly above the highest in our series. Therefore to conclude that the valproate-induced hyperammonemia and the MR changes are related may be erroneous. It would be interesting to perform the same MR studies in asymptomatic patients with valproate-induced hyperammonemia.