Determining the genetic contribution in patients with non-syndromic ascending thoracic aortic aneurysms: Correlation with findings from computational pathology

Abstract

ObjectivesThis study aims to identify the clinical utility of targeted-genetic sequencing in a cohort of patients with TAA and establish a new method for regional histological characterisation of TAA disease. MethodsFifty-four patients undergoing surgery for proximal TAA were recruited. Exclusions: connective tissue disease, bicuspid aortic valves, redo surgery. All patients underwent next generation sequencing (NGS) using a custom gene panel containing 63 genes previously associated with TAA on Illumina MiSeqor NextSeq550 platforms. Explanted TAA tissue was obtained en-bloc from 34/54 patients, and complete circumferential strips of TAA tissue processed into whole slides which were subsequently digitalised. Computational pathology methods were employed to quantify elastin, cellularity and collagen in six equally divided regions across the whole aneurysm circumference. ResultsOf 54 patients, clearly pathogenic or potentially pathogenic variants were found in 7.4%: namely LOX, PRKG1, TGFBR1 and SMAD3 genes. 55% had at least one variant of unknown significance (VUS) and seven of the VUSs were in genes with a strong disease association (category A) genes, whilst 15 were from moderate risk (category B) genes. Elastin and collagen abundance displayed high regional variation throughout the aneurysm circumference. In patients with <60% total elastin, the loss of elastin was more significant on the outer curve (38.0% vs 47.4%, p = 0.0094). The presence of VUS, higher pulse wave velocity and advancing age were predictors of elastin loss (regression analysis: p < 0.05). ConclusionsThese findings demonstrate the heterogeneity of TAA disease microstructure and the potential link between histological appearance and clinical factors, including genetic variation.