Metabolomics profiling in dried blood spots differentiates clinical phenotypes in VLCADD
Authors: Suzan Knottnerus1,3, Jeannette Bleeker1,3, Maria van der Ham2, Mia Pras-Raves1,3, Riekelt Houtkooper1, Gepke Visser1,3, Monique de Sain-van der Velden2
Affiliations:
1Laboratory Genetic Metabolic Diseases, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands;
2Department of Medical Genetics, Section Metabolic Diagnostics, Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands;
3Department of Metabolic Diseases, Wilhelmina Children’s Hospital, University Medical Center
Utrecht, Utrecht, The Netherlands;
Corresponding author’s information
Corresponding author: Suzan J.G. Knottnerus Corresponding author’s e-mail: s.j.knottnerus@amc.uva.nl
Background
The inclusion of very long chain acyl-CoA dehydrogenase deficiency (VLCADD) (OMIM 201475) in many newborn screening (NBS) programs worldwide has led to identification of more patients including pre-symptomatic individuals. It is not yet clear which patients will develop severe symptoms and which not, so the need for early phenotype prediction is high. We previously reported a strong correlation between long-chain fatty acid oxidation (lc-FAO) flux in fibroblasts and clinical outcome. Disadvantages of this method is the need for skin biopsies and a long turnaround time. In this study, we developed a rapid, as less invasive as possible, prognosis prediction method for VLCADD.
Method
DBS from VLCADD patients were available from regular visits to the outpatient clinic (n=25) and from the Dutch NBS (n=13). A metabolic fingerprint (>20.000 features) was generated by direct infusion high-resolution mass-spectrometry. Before metabolomics profiling patients were divided in phenotypic groups based on residual lc-FAO flux and clinical features. Data sets were used for hierarchical clustering using MetaboAnalyst (http://metaboanalyst.ca/). Data analysis included supervised multivariate statistics (Partial Least Squares Discriminant Analysis; PLS-DA).
Results
As expected the biomarker for NBS, C14:1-carnitine, was significantly elevated in all. In addition , C16:2-carnitine was significantly elevated in all VLCADD patients. Based on untargeted metabolite profiles by PLS-DA patients with a severe phenotype could be discriminated from patients with a mild phenotype. The results were confirmed by blinded NBS bloodspot analysis of 13 patients and 35 age matched NBS bloodspot controls. Of those 2 had a metabolic profile fitting the severe phenotype.
One of the patients was deceased, the other had severe VLCADD related symptoms. Both had the lowest lc-FAO flux of the VLCADD patients.
Discussion
State-of-the-art DBS metabolomics showed different metabolic profiles between VLCADD patients with the severe and mild phenotypes. This method may accelerate prognosis prediction and improve rapid initiation of individualized therapy.