Surprising clinical lessons from targeted next generation sequencing of thoracic aortic aneurysmal genes. B. Loeys1,2, D. Proost1, G. Vandeweyer1, S. Salemink2, M. Kempers2, G. Oswald3, H. Dietz3, G. Mortier1, L. Van Laer1 1) Center for Medical Genetics, University of Antwerp/ Antwerp University Hospital, Antwerp, Belgium; 2) Department of Genetics, Radboud University Medical Center, Nijmegen, The Netherlands; 3) Mc Kusick Nathans Institute for Genetic Medicine, Johns Hopkins University Hospital, Baltimore, USA.

   Thoracic aortic aneurysm/dissection (TAA), an important cause of death in the industrialized world, is genetically heterogeneous and at least 14 causative genes have been identified, accounting for both syndromic and non-syndromic forms. The diagnosis is not always straightforward because a considerable clinical overlap exists between patients with mutations in different genes, and mutations in the same gene cause a wide phenotypic variability. Molecular confirmation of the diagnosis is becoming increasingly important for gene-tailored patient management but consecutive, conventional molecular TAA gene screening is expensive and labor-intensive. To shorten the turn-around-time, to increase mutation-uptake and to reduce the overall cost of molecular testing, we developed a TAA gene panel for next generation sequencing (NGS) of 14 TAA genes (ACTA2, COL3A1, EFEMP2, FBN1, FLNA, MYH11, MYLK, NOTCH1, SKI, SLC2A10, SMAD3, TGFB2, TGFBR1 and TGFBR2). We obtained enrichment with Haloplex technology and performed 2x150 bp paired-end runs on a Miseq sequencer in a series of 57 consecutive TAA patients, both syndromic and non-syndromic. The sensitivity and false positive rate were previously shown to be 100% and 3%, respectively. Applying our NGS approach, we identified a causal mutation in 16 patients (28%). This uptake is really high as on average one molecular study per patient (range 0-6) was performed prior to inclusion in this study. One mutation was found in each of the 6 following genes: ACTA2, COL3A1, TGFBR1, MYLK, SMAD3, SLC2A10 (homozygous); two mutations in NOTCH1 and eight in FBN1. An additional 6 variants of unknown significance were identified: 2 in FLNA, 2 in NOTCH1, 1 in FBN1 and 1 heterozygous in EFEMP2. All variants were confirmed by Sanger sequencing. Remarkably, from the eight FBN1 positive patients, three patients had previously been tested FBN1 negative by certified labs, indicating that the sensitivity of Sanger sequencing is not 100%. Interestingly, in two FBN1 mutation positive patients the clinical diagnosis of Marfan syndrome was unsuspected. Similarly, the clinical diagnosis of vascular Ehlers-Danlos syndrome (COL3A1) had not been made. Finally, the ACTA2 mutation was identified postmortem from paraffin-embedded extracted DNA. We conclude that our NGS approach for TAA genetic testing overcomes the intrinsic hurdles of Sanger sequencing and becomes a powerful tool in the elaboration of clinical phenotypes assigned to different genes.

You may contact the first author (during and after the meeting) at