FBN1 coding variants and non-syndromic aortic disease

Marfan syndrome (MFS) is an autosomal dominant disorder caused by pathogenic variants in FBN1. Associated cardiovascular features include mitral valve prolapse, ascending aortic dilatation, and aortic dissection. The estimated prevalence of MFS and suspected MFS is one per 5000 individuals¹ but the frequency and significance of underlying pathogenic FBN1 variants in the general population has not been reported. We examined the clinical phenotypes of PennMedicine BioBank (PMBB) participants carrying any of 12 canonical MFS variants in FBN1: V449I, R976H, G1013R, R1170H, P1424A, M1576T, C1672Y, T1908I, I2185T, T2520M, I2585T, G2618R. These variants were chosen because they were among the first FBN1 variants reported for MFS; although the phenotypic features associated with some of these variants are not sufficient to diagnose MFS by current criteria, each variant has been associated with MFS-related phenotypes.

PMBB is a longitudinal genomic medicine cohort in which participants consent to linkage of biospecimens with electronic health record (EHR) data and is approved by the University of Pennsylvania Institutional Review Board. 70 carriers of the 12 FBN1 variants we chose to investigate were identified among 10,996 participants that underwent whole exome sequencing and passed quality control, including pruning for cryptic relatedness ($\widehat{\pi }$ ≥ 0.125; third-degree relatives)². Of the 12 variants, 6 were present in only one individual (C1672Y, G1013R, I2185T, T1908I, T2520M, I2585T). The remainder were more common, including R1170H, which was present in 40 participants (Table).

Table.

Frequency of participants in the PMBB with one of the FBN1 variants and the number with a cardiovascular phenotype

Variant	# with variant	Positive CV phenotype*,†	No CV phenotype*,†	No CV phenotype identifiable*,‡
V449I	3	1	1	1
R976H	2	0	0	2
G1013R	1	1	0	0
R1170H	42	11	14	17
P1424A	6	1	2	3
M1576T	5	0	2	3
C1672Y (?arg)	1	1	0	0
T1908I	1	1	0	0
I2185T	1	1	0	0
T2520M	1	1	0	0
I2585T	1	1	0	0
G2618R	6	2	4	0
Total	70	21	23	26

^*A cardiovascular phenotype seen in MFS

^†Based on recorded echocardiogram and /or thoracic CT or MRI

^‡Based on the absence of an echocardiogram or thoracic CT or MRI

The data that support the findings of this study are available from the corresponding author upon reasonable request. Clinical data, including office notes, diagnostic tests, and imaging studies, were reviewed as of December 1, 2018. The median age of heterozygotes was 65 years old (range 28 – 87 years old) and 56% were male. Two carriers had a clinical diagnosis of MFS. Of the remainder, 21 (30%) carriers had evidence of one or more MFS-related cardiovascular phenotypes: 4 participants had mitral valve disease; 6 had a dilated sinus of Valsalva, 5 had a dilated ascending aorta, 4 had descending thoracic or abdominal aortic aneurysms or dissections; and 4 had undergone surgical procedures involving the mitral valve or thoracic aorta. This may be an underestimate of the burden of disease given that 25 (36%) of the individuals had insufficient imaging information to fully exclude MFS-related cardiovascular traits; an even larger percentage (85%) lacked adequate physical exam data to definitively exclude non-cardiovascular manifestations.

We compared the rates of mitral valve and aortic disease between FBN1 carriers and non-carriers. Two age and sex matched controls (n=140) who did not carry known or suspected pathogenic FBN1 variants based on ClinVar annotations were selected for every carrier. The median age, sex, and years of available of EHR data were not different between FBN1 non-carriers and carriers. Only 12 (9%) participants without pathogenic FBN1 variants had evidence of mitral valve or aortic disease, although 39 (28%) lacked sufficient imaging to conclusively exclude a MFS-related cardiovascular trait. When compared to this baseline rate of disease among non-carriers, FBN1 carriers were significantly more likely to have MFS-related cardiovascular disease (OR 5.2, 95%CI 2.4 – 11.6, p=2.9x10⁻⁵); when restricted to individuals with adequate imaging to conclusively diagnose or exclude cardiovascular phenotypes this association persisted (OR 7.6, 95%CI 3.1 – 20, p=1.1x10⁻⁶).

The severity of cardiovascular involvement in MFS is known to be influenced by the specific FBN1 variant³. Our analysis of carriers of pathogenic FBN1 variants demonstrates that despite the fact that the 12 variants analyzed here are strongly associated MFS-related cardiovascular disease, the variants have varied penetrance. Although these variants have traditionally been thought to be associated with MFS, our data show they more commonly associate with isolated mitral valve or aortic pathology, and are compatible with survival into late adulthood. This suggests that the effect of many FBN1 variants requires time to manifest and that ongoing surveillance is of the utmost importance.

There are several limitations to this study. Biobanks are biased toward older individuals who have survived long enough to be recruited, likely underestimating the true association of variants with MFS and MFS-related cardiovascular disease. Additionally approximately one-third of subjects and controls lacked the necessary EHR data to exclude conclusively the traits of interest.

FBN1 is included on the ACMG list of reportable incidental genetic findings because of its association with MFS⁴. Despite the fact that most pathogenic FBN1 variant carriers in this study did not have MFS, the wide range of clinically significant mitral valve and aortic pathology seen in carriers of pathogenic FBN1 variants argues for the importance of clinical screening in this genetically predisposed population. We feel informing research participants of these findings is important and are working to return CLIA certified FBN1 results to participants.

APPENDIX

Whole exome sequencing was performed by the Regeneron Genetics Center.