in this gene have been shown to be present in the majority of patients with a clinical diagnosis of pyridoxinedependent seizures 3-6 . These mutations cause a deficien cy in α-aminoadipic semialdehyde (α-AASA) dehydrogenase, indirectly leading to a secondary deficiency in pyri doxal-5-phosphate (P5P), which causes seizures. Administration of pyridoxine restores the P5P pool, thus controlling seizure activity 3 . Another consequence of α-AASA dehydrogenase deficiency is an accumulation of α-AASA within the body, leading to increased plasma and cerebrospinal fluid lev els 3 . α-AASA is excreted in the urine, thus leading to highly increased urinary levels in pyridoxine dependent patients with ALDH7A1 mutations. We and others have shown that urinary α-AASA should be used as the biomarker for pyridoxine dependent seizures 3-5,7 . In conclusion, in contrast to what Lin and co-authors have stated, the underlying pathophysiology and genetic mutations can be determined in the vast majority of patients with pyridoxine-dependent seizures. The diagnosis is no longer merely dependent on clinical criteria, but can be confirmed at the metabolite level by measuring αAASA in the urine of suspected patients. Ideally, urinary α-AASA quantification and, if the biomarker is increased, subsequent DNA analysis should be performed in any child with suspected pyridoxine-dependent seizures. A trial of withdrawal in these patients should be omitted 7 .