Amino acid ratio combinations as biomarkers for discriminating patients with pyruvate dehydrogenase complex deficiency from other inborn errors of metabolism.

Abstract

Pyruvate dehydrogenase complex deficiency (PDCD) is a mitochondrial neurometabolic disorder of energy deficit, with incidence of about 1 in 42,000 live births annually in the USA. The median and mean ages of diagnosis of PDCD are about 12 and 31 months, respectively. PDCD is a major cause of primary lactic acidosis with concomitant elevation in blood alanine (Ala) and proline (Pro) concentrations depending on phenotypic severity. Alanine/Leucine (Ala/Leu) ≥4.0 and Proline/Leucine (Pro/Leu) ≥3.0 combination cutoff from dried blood spot specimens was used as a biomarker for early identification of neonates/infants with PDCD. Further investigations were needed to evaluate the sensitivity (SN), specificity (SP), and clinical utility of such amino acid (AA) ratio combination cutoffs in discriminating PDCD from other inborn errors of metabolism (IEM) for early identification of such patients. We reviewed medical records of patients seen at UPMC in the past 11 years with molecularly or enzymatically confirmed diagnosis. We collected plasma AA analysis data from samples prior to initiation of therapeutic interventions such as total parenteral nutrition and/or ketogenic diet. Conditions evaluated included organic acidemias, primary mitochondrial disorders (MtDs), fatty acid oxidation disorders (FAOD), other IEMs on current newborn screening panels, congenital cardiac great vessel anomalies, renal tubular acidosis, and non-IEMs. The utility of specific AA ratio combinations as biomarkers were evaluated using receiver operating characteristic curves, correlation analysis, principal component analysis, and cutoff SN, SP, and positive predictive value determined from 201 subjects with broad age range. Alanine/Lysine (Ala/Lys) and Ala/Leu as well as (Ala + Pro)/(Leu + Lys) and Ala/Leu ratio combinations effectively discriminated subjects with PDCD from those with other MtDs and IEMs on current newborn screening panels. Specific AA ratio combinations were significantly more sensitive in identifying PDCD than Ala alone or combinations of Ala and/or Pro in the evaluated cohort of subjects. Ala/Lys ≥3.0 and Ala/Leu ≥5.0 as well as (Ala + Pro)/(Leu + Lys) ≥2.5 and Ala/Leu ≥5.0 combination cutoffs identified patients with PDCD with 100% SN and ~85% SP. With the best predictor of survival and positive cognitive outcome in PDCD being age of diagnosis, PDCD patients would benefit from use of such highly SN and SP AA ratio combination cutoffs as biomarkers for early identification of at-risk newborns, infants, and children, for early intervention(s) with known and/or novel therapeutics for this disorder.