Pathways linking atherosclerosis to aortic stenosis

Abstract

Cardiovascular disease is the most common cause of death world-wide where atherosclerosis is the main culprit and aortic valve disease accounts about two percent of all CVD deaths. Atherosclerosis is a lipid and inflammation driven disease that share many features with aortic valve stenosis (AVS). Globally, the prevalence of AVS has been estimated to over 10 million patients and the incidence to over 12 500 new cases annually which is likely increasing due to increased longevity, yet no medical treatment is available. A link between atherosclerosis and AVS has previously been established by overlapping prevalence and common pathobiological hallmarks including lipid infiltration, inflammation, and calcification. Recent genetic studies have demonstrated several loci in which single nucleotide polymorphisms are associated with both diseases. However, there is also evidence pointing to separate etiologies including disease specific genetic risk factors, histopathological differences, and isolated clinical presentation. The aim of this thesis was to establish the interplay between atherosclerosis and AVS. A physiologic part was covered in Article I, specific mechanisms in Article II-IV and molecular epidemiology in Article IV.

In Article I, arterial stiffness was determined in a cohort with ascending aortic dilatation and/or aortic valve disease before and after cardiac surgery. Arterial stiffness correlates with atherosclerotic cardiovascular disease and aggravates the increased left ventricular stress in AVS. Cardio-ankle vascular index (CAVI) measures arterial stiffness from the heart to the ankle and was lower in subjects with AVS compared with aortic regurgitation and ascending aortic dilation, before surgery, despite being older. In contrast, aortic stiffness assessed by carotid femoral pulse wave velocity (cfPWV) was not different between the groups. After surgery, CAVI but not cfPWV increased in patients with AVS but remained unchanged in patients undergoing aortic surgery. Age, diabetes, lower body mass index, decreased ejection time and lower preoperative CAVI was associated with an increased CAVI after surgery. The results suggest that AVS may mask an increased arterial stiffness if peripheral arteries are included in the measurement. Also, ejection time emerged as an important variable to account for when measuring arterial stiffness in aortic valve disease patients. Future work should aim to establish if arterial stiffness may be used to risk-stratify AVS patients.

In Article II, the impact of a single nucleotide polymorphism (SNP) within FADS1 on aortic valve gene expression and fatty acid composition was identified. Fatty acid desaturase (FADS)1 and FADS2 encode rate limiting enzymes in the metabolism of omega-3 and omega-6 polyunsaturated fatty acids (PUFAs) and the SNP within this locus is associated with lower risk of both AVS and CAD. The SNP rs17547 was associated with FADS2 mRNA expression in calcified aortic valve tissue and the enzymatic activity of both FADS1 and FADS2. In addition, the aortic valve omega-3 PUFA docosahexaenoic acid proportion was higher in non-calcified compared with calcified tissue and positively correlated with the SNP. The results indicate that the protective effects of the SNP might be mediated via an increased DHA proportion in the aortic valve and/or possibly via downstream mediators from DHA such as specialized pro-resolving mediators which have been shown to dampen inflammation.

Further pathophysiological evidence of shared pathways between CAD and AVS was obtained in Article III. The presence of antiphospholipid antibodies (aPL) in the general population is higher in patients with a recent myocardial infarction. Positivity for antibodies against β2-glycoprotein I and/or cardiolipin of IgG isotype was identified to be 8-fold higher in AVS patients compared with matched controls. In aortic valve tissue, aPL positivity was associated with downregulated interferon pathways and upregulated pathways related to mechanosensory signaling. Importantly, the differentially expressed genes could predict resilient (healthy), thickened (fibrotic) and calcified aortic valve tissue with high accuracy using supervised machine learning models suggesting a tight relationship between aPL related genes and local disease progression. The overall results imply that aPL IgG in the general population (without rheumatic disease) could be a risk factor for AVS and may potentially be used guide AVS precision medicine.

In Article IV, CAD associated gene expression in aortic valve tissue was identified. First, the prevalence of CAD in a contemporary surgical tricuspid AVS cohort was established at 49% and was associated with claudication, smoking, male sex, and diabetes. An exploratory analysis of aortic valve transcriptomic data from 74 patients revealed that severe CAD, affecting 2 or 3 vessel territories, was associated with the most prominent difference in gene expression. The differentially expressed genes were primarily found in non-calcified tissue and were enriched in pathways related to oxidative stress, inflammation, and lipids. Furthermore, a supervised machine learning model could predict if aortic valve tissue stemmed from patients with severe CAD, at high accuracy. The most important gene predictors of severe CAD could further be used to predict atherosclerotic or macroscopically normal carotid artery tissue. The results suggest that AVS patients with concomitant severe CAD exhibit more atherosclerosis related mechanisms in non-calcified tissue, ultimately leading to a common end-stage disease with severe AVS.

In summary, the results in this thesis demonstrate that AVS may be a cause of masked systemic arterial stiffness. Furthermore, pathways related to fatty acid metabolism and aPL are implicated in the pathophysiology of AVS and patients with severe CAD exhibit upregulated pathways related to atherosclerosis in the aortic valve. Collectively, pathways linking and differentiating aortic valve and vascular atherosclerotic disease were unraveled which open up for novel precision treatment regiments to halt AVS.