Abstracts

Rationale: Metabolic testing has long been part of the routine evaluation for many children with onset of epilepsy in early life, but the yield of these studies in the modern era is not well established.Methods: We prospectively recruited children with newly diagnosed early life epilepsy (ELE), defined as onset below three years, from 17 pediatric epilepsy centers in the US (2012-2015). We extracted clinical information from the medical records and entered the data into a REDCAP© database. In this analysis we paid particular attention to the use of metabolic testing performed as part of the initial epilepsy diagnostic evaluation (in-patient or within three weeks of diagnosis).Results: Of 620 children, metabolic testing was performed in 306 (49.7%), most often involving blood (46%), urine (31%), or CSF analysis (17%). Blood for amino acids and urine for organic acids were sent in 24% of all patients. Blood tests were normal in 111/285 and there were 170/174 non-diagnostic abnormalities. Urine tests were normal in 137/193 and there were 54/56 non-diagnostic abnormalities. CSF studies were normal in 56/104 and there were 46/48 non-diagnostic results. The results of metabolic testing were diagnostic in a total of 8 individual cases (1%): three children with Leigh syndrome (two diagnosed prior to onset of epilepsy), two with non-ketotic hyperglycinemia, and one each with methylmalonic academia, Zellweger syndrome, and GM1 gangliosidosis. All of the diagnosed conditions that were found had clinical features, imaging findings, or basic metabolic derangements (e.g. blood gasses, lactate, pyruvate) that would raise suspicion of the diagnosis. One other child with pyruvate dehydrogenase deficiency was diagnosed based on genetic testing with equivocal metabolic testing results.Conclusions: The diagnostic yield of metabolic testing in children with early onset of epilepsy is low, there is a considerable rate of non-diagnostic abnormal results, and all positive results were found in children who were suspected clinically. We speculate that mandatory modern newborn screening coupled with expert neonatal evaluations may have altered the yield of standard metabolic testing in infants and very young children evaluated for new-onset seizures. Specifically, correctable metabolic diseases may be diagnosed and treated in the neonatal period thereby ameliorating or even preventing their otherwise encephalopathic courses. It may therefore be appropriate to re-examine the prioritization of routine metabolic testing in this population, particularly in those children in whom no clinical clues are present to suggest the presence of an inborn error of metabolism. Funding: The Pediatric Epilepsy Research Foundation