Recurrent spontaneous coronary dissections in a patient with a de novo fibrillin-1 mutation without Marfan syndrome.

Dear Sirs, Among the causes of myocardial infarction, spontaneous coronary artery dissection (SCAD) is rare and underdiagnosed, owing to the limited accuracy of coronary angiography. Without occlusion, SCAD can remain asymptomatic. The largest series comprised 23 cases from > 11,000 coronary angiograms at Vancouver General Hospital and 87 retrospectively identified cases that occured 1979–2011 at the Mayo clinic (1). Conditions predisposing for SCAD include vascular syndromes and fibromuscular dysplasia. Little is known, however, about underlying molecular mechanisms. We report SCAD in a 46-year-old male patient without cardiovascular risk factors or family history, who presented with acute myocardial infarction (MI) due to an occlusive dissection of the left circumflex coronary artery and four years later with spiral dissection of the left anterior descending (LAD) coronary artery in the absence of atherosclerotic plaques (▶ Figure 1 A, B). Neither event was preceded by excessive physical stress. The patient underwent comprehensive cardiovascular diagnostics, whole exome sequencing, and analysis of smooth muscle progenitor cell (SPC) abundance and function using flow cytometry and a scratch-induced migration assay. Echocardiography, computed tomography and magnetic resonance angiography excluded ectasia and fibromuscular dysplasia of medium-sized or large arteries including the aorta. Screening of candidate genes for arterial dissection revealed a de novo missense mutation in the Marfan gene FBN1 (c.4214T>G), which is presumed to be pathogenic, as well as paternally transmitted missense variants in myosin light-chain kinase (c.1609G>A) and in the SKI gene (c.2007C->G), both predicted to be benign. The patient neither displayed features of Shprintzen-Goldberg-syndrome, nor did he fulfil the criteria of the revised Ghent nosology for Marfan syndrome and related conditions (2), except displaying a minor mitral valve prolapse. The heterozygeous FBN1 mutation at the beginning of exon 34 led to a rare exchange of apolar leucin 1405 into polar arginine. Significant aberrant splicing was excluded by RNA sequencing. Increased transforming growth factor (TGF)-β levels and signalling have been implicated in aneurysm formation and arterial dissection in Marfan patients (3). As opposed to Marfan patients with FBN1 mutations but in line with the non-Marfanoid-phenotype of our patient, his TGF-β plasma levels were not elevated (▶ Figure 1 C). Arterial stiffness as determined by aortic pulse wave velocity was low (6.7 ± 0.3 m/s, 5th percentile of age-matched controls), differing from findings in Marfan patients. Serum fibrillin-1 levels were unaltered. Searching for alternative mechanisms, we found that circulating SPCs, which are involved in arterial remodeling (4), were more abundant in peripheral blood of our patient, as compared to agematched healthy controls (▶ Figure 1 D). Functionally, his SPCs showed increased proliferation and migration in vitro, thereby improving scratch-wound recovery, whereas knockdown of fibrillin-1 reduced wound recovery to comparably low levels in our patient and controls, confirming its role in SPC migration and suggesting a FBN1-dependent gain-of-function in our patient (▶ Figure 1 E). To our best knowledge, we report the first case of SCAD with a FBN1 mutation, which intriguingly is not associated with a Marfan phenotype. Software based prediction plays an important role in assessing the pathophysiological relevance of a nonsynonymous single-nucleotide polymorphism (SNP). Unfortunately, different prediction algorithms use different information, which at times results in incongruent predictions. In our case, “SIFT” and “PolyPhen2” classify the SNP as tolerated, albeit with suboptimal confidence, whereas “Mutation Taster” classifies the variant as disease-causing. Amongst others, the strong decrease in hydrophobicity and the high phylogentic conservation of the neighbouring sequence were considered, defining a need for further functional experiments. Indeed, biochemical and functional parameters corroborated that the SNP does not mediate changes seen in Marfan syndrome but rather imply distinct pathomechanisms related to SPCs. The elevated number and enhanced migratory behaviour of SPCs may give rise to structural alterations increasing the susceptibility to arterial dissection. Supportive of a role in SCAD, this FBN1 mutation was also present in a patient with thoracic aortic dissection without aneurysm or Marfan Letters to the Editor