Welcome to the EAS 2023 Interactive Program

Background and Aims

Chronic inflammation and immune dysregulation drive atherosclerosis development in rheumatoid arthritis (RA). Recent evidence suggests a link via humoral responses against high-density lipoproteins (HDL). However, their specificity, clinical relevance and emergence along disease course are unknown, especially during the earliest phases of RA.

Methods

IgG and IgM serum levels of antibodies against HDL (anti-HDL) and Apolipoprotein A1 (anti-ApoA1) were measured in 82 early RA patients, 14 arthralgia individuals and 96 controls. Established RA patients (n=42) were included for validation. Atherosclerosis and vascular stiffness were measured by Doppler-ultrasound. Lipoprotein content, particle numbers and size were measured by H-NMR. Cytokines were measured by immunoassays. A cardiometabolic-related protein panel was evaluated using high-throughput targeted proteomics.

Results

Anti-HDL and anti-ApoA1 responses were increased in early RA compared to controls and were comparable to established disease. Only anti-ApoA1 antibodies were increased in arthralgia. IgG anti-HDL and anti-ApoA1 were associated with unfavourable lipoprotein traits in RA and arthralgia, respectively. A similar picture was observed for inflammatory mediators. No associations with clinical features or risk factors were found. IgG anti-HDL were independently associated with atherosclerosis occurrence in early RA, and outperformed patient stratification over conventional algorithms and their anti-ApoA1 counterparts. Anti-HDL antibodies correlated with proteins involved in immune activation, remodelling, and lipid metabolism pathways in early RA.

Conclusions

Humoral responses against HDL particles may be a missing link between inflammation and atherosclerosis along arthritis course, although quantitative and qualitative differences can be noticed. These differences informed distinct capacities as biomarkers and underlying pathogenic circuits.

Background and Aims

Overweight and obesity are modifiable risk factors for coronary artery disease (CAD) in the general population, but their prevalence in patients with familial hypercholesterolaemia (FH) and whether they confer additional risk of CAD independent of LDL-cholesterol in FH is largely unknown.

Methods

Cross-sectional analysis on 35,919 patients across 46 countries, included in the EAS FH Studies Collaboration (FHSC) Registry. We assessed body mass index distribution (adults: underweight:<18.5, normal weight:18.5-<25.0, overweight:25.0-<30.0, obesity:≥30 kg/m²; children: z-scores) in patients with homozygous (HoFH; n=57 children/adolescents [<18 years], n=325 adults) or heterozygous FH (HeFH; n=6275 children/adolescents, n=29,262 adults), overall and by geographic region, and whether obesity was independently associated with CAD independent of LDL-cholesterol levels.

Results

Among patients with HoFH, 55% of adults and 25% of children were overweight or obese; corresponding proportions for HeFH were 52% and 27%; Figure-top. By region, prevalence of overweight/obesity was highest in Northern-Africa/Western-Asia. In HeFH and HoFH, CAD was more prevalent in patients with vs. without obesity; Figure-middle. After adjusting for age, sex, LDL-cholesterol and lipid-lowering medication, being obese vs. non-obese was associated with higher odds of CAD in patients with HeFH (OR, adults: 2.16 [95%CI: 1.97,2.36]; children/adolescents: 6.87 [1.55,30.46]), but not in patients with HoFH; Figure-bottom. Results remained similar after further adjustment for diabetes.

Conclusions

Obesity is independently associated with the presence of CAD in adults and children with HeFH, with similar trends in HoFH. In addition to LDL-cholesterol control, preventing and treating obesity are required to reduce the risk of CAD among those with FH.

Background and Aims

Background: Lipoprotein(a) may be implicated in peripheral arterial disease (PAD) and abdominal aortic aneurysms (AAA), yet data from general population studies are limited, and large studies are needed to determine risk in individuals with the highest levels and associated risk of major adverse limb events (MALE).

Aims: To test whether high lipoprotein(a) levels associate with increased risk of PAD, AAA, and MALE, and to provide genetic evidence of causality using LPA genotypes.

Methods

Methods: We included 108,146 individuals from the contemporary Copenhagen General Population Study (CGPS). During follow-up, 2,450 developed PAD, 1,251 AAA, and 489 MALE. We used the historic Copenhagen City Heart Study (CCHS, N=10,960) to replicate findings for MALE (N=116).

Results

Results: High lipoprotein(a) levels and corresponding LPA risk genotypes associated with increased risk of PAD and AAA in the CGPS (Figure 1). For individuals with lipoprotein(a) levels >99^th percentile (≥143mg/dL,≥306nmol/L), multivariable adjusted hazard ratios were 2.99(95% confidence interval:2.09-4.30) for PAD and 2.22(1.21-4.07) for AAA compared to individuals with levels <50^th percentile (≤9mg/dL,≤17nmol/L). Corresponding hazard ratios for MALE were 2.70(1.67-4.37) in the CGPS and 8.12(3.56-18.55) in the CCHS. Absolute 10-year risks of PAD and/or AAA were 11% and 29% in smoking women aged 70-79 years with lipoprotein(a) <50^th vs. >99^th percentile. Equivalent values in men were 19% and 47%. The percentage of events attributable to lipoprotein(a) >50^th percentile was 9.5% and 12.3% for PAD and AAA.

Conclusions

Conclusion: High lipoprotein(a) levels increased risk of PAD, AAA, and MALE in the general population, opening opportunities for prevention given future lipoprotein(a) lowering therapies.

Background and Aims

Obesity promotes diverse pathologies, including atherosclerosis and dementia, which frequently involve vascular defects and endothelial cell (EC) dysfunction. Each organ has distinct EC subtypes, but whether ECs are differentially affected by obesity is unknown. Here, we aim to identify organ-specific vulnerabilities of endothelial cells in mouse obesity model.

Methods

We utilize scRNA-seq to analyse transcriptomes of ~375,000 ECs from seven organs in male mice at progressive stages of obesity induced by Western diet. We confirm our findings with orthogonal approaches including immunofluorescent stainings, tracing experiments, and in situ RNA hybridization.

Results

We reveal unique organ- and EC-subtype specific molecular changes driven by obesity, including lipid handling, metabolic pathways, AP1 transcription factor and inflammatory signaling. The transcriptomic aberrations worsen with sustained obesity and are only partially mitigated by dietary intervention and weight loss. By switching obese animals onto a chow diet, we uncover distinct obesity-driven changes in ECs that are responsive or resistant to weight loss. For example, dietary intervention substantially attenuates dysregulation of liver and adipose tissue, but not of kidney EC transcriptomes. Through integration with human GWAS data, we further identify a subset of vascular disease risk genes that are induced by obesity.

Conclusions

Our work catalogues the impact of obesity on the endothelium, and provides a useful resource for further investigation of potential organ-specific vascular therapeutic targets. The data generated in this study are available through an interactive website - https://obesity-ecatlas.helmholtz-muenchen.de.

Background and Aims

The 2017 Japan Atherosclerosis Society (JAS) familial hypercholesterolemia (FH) criteria adopts a cut off value of ≥ 9 mm of Achilles tendon thickness (ATT) detected by X-ray as one of the three key items. This threshold was determined based on data assessing ATT of 36 non-FH individuals in 1977. Although the specificity of this clinical criteria is extremely high, there are substantial number of patients with FH whose ATT < 9 mm. We aimed to determine a cut off value of ATT using X-ray FH based on genetic diagnosis.

Methods

The individuals (male/female = 486/501) with full assessments of genetic analyses for FH-genes (LDLR, and PCSK9), serum lipids, and ATT detected by X-ray at Kanazawa University Hospital and National Cerebral and Cardiovascular Center Research Institute were included in this study. Receiver operating characteristic (ROC) analyses were performed to determine a better cut off point of ATT predicting a pathogenic mutation of FH.

Results

ROC analyses revealed the best cut off values of ATT as 7.6 mm for male, and 7.0 mm for female with the sensitivities and specificities of 0.83 and 0.83 for male and 0.86 and 0.85 for female, respectively. If the thresholds of ATT of 8.0/7.5 mm and 7.5/7.0 mm were applied to diagnose of male/female FH, the sensitivities/specificities predicting a pathogenic mutation of FH by the 2017 JAS FH clinical criteria would be 0.82/0.90 and 0.85/0.88, respectively.

Conclusions

The optimal cut-off value of ATT detected by X-ray is obviously lower than 9.0 mm adopted by the 2017 JAS FH clinical criteria.

Background and Aims

RNA-binding protein HuR may control the fate of pro-inflammatory and pro-thrombotic genes in human atherosclerosis.

Methods

Single-cell RNA-sequencing (scRNA)-data were analyzed to compare cell-specific expression profiles of HuR and its target genes in murine and human atherosclerotic vascular disease. HuR individual-nucleotide cross-linking and immunoprecipitation (iCLiP), small interfering RNA-mediated HuR silencing, HuR overexpression and RNA stability assays were used to determine HuR-mediated regulation of pro-atherosclerotic genes. Expression levels of HuR were measured in peripheral blood mononuclear cells derived from 718 individuals at risk for cardiometabolic disease or with established chronic coronary syndromes (CCS) and followed for major adverse cardiovascular events (MACE).

Results

scRNA-data revealed increased HuR expression levels in vascular endothelium and infiltrated immune cells. HuR iCLiP experiments revealed HuR binding sites in the 3’ untranslated region of several genes in endothelial cells. Silencing of HuR reduces, overexpression of HuR induces the expression of targeted genes (P<0.05 for all). Inhibition of transcription by actinomycin D showed that endothelial HuR controls the mRNA stability of CCL2, CTSS, CXCL1, CXCL2 and SERPINE1 (P<0.05 for all). In humans, HuR mRNA expression was increased in CCS compared to non-coronary artery disease (CAD) participants (p<0.001) and independently associated with the presence of CAD (OR=2.67 for highest vs. lower HuR tertiles) after adjustment for traditional risk factors. High HuR concentrations independently associated with MACE incidence across a median follow-up period of 48 months (higher versus lowest tertile: 7.4% vs. 1.27%, log rank test P=0.009).

Conclusions

HuR controls the stability of pro-atherosclerotic targets, predicts the progression of atherosclerosis and MACE in humans.

Background and Aims

After diagnosis, patients with familial hypercholesterolemia (FH) are advised lifelong cholesterol-lowering treatment. Although without trial evidence, we always propose to combine a healthy lifestyle with statins as the first line of therapy. The geographic variation of life expectancy in the Netherlands has enabled a trial by nature of the same untreated FH mutation in a beneficial and general environment.

Methods

We performed a nationwide cascade screening and obtained a large pedigree with the S306L mutation in the low-density lipoprotein receptor (LDLR ) gene. Before the statins became available in 1990, we determined the all-cause standardized mortality ratio (SMR). Patients and spouses were compared with Poisson regression (relative risk).

Results

Between 1800 and 1989, 147 deaths occurred in 14,152 person-years. Overall, the SMR was 1.25 (95% CI 1.05-1.46; P = 0.0056). The spouses in a beneficial environment had an SMR of 0.75 (95% CI 0.55-0.99; P =0.0547). The mortality in this beneficial environment was not increased (SMR 1.16; 95% CI 0.85-1.54 ; P =0.1794), while other parts confirmed the known overall excess mortality from untreated FH (SMR 1.29; 95% CI 1.05-1.57; P = 0.0076). However, the FH patients in the beneficial environment had 1.5 times more deaths relative to their spouses (RR 1.46; 95% CI 0.96-2.22; P =0.07388).

Conclusions

A favorable local environment decreased excess mortality significantly. However, the S306L mutation in the LDLR gene still caused the expected excess mortality locally. Our findings show that a beneficial lifestyle is an add-on but not an alternative for the immediate start of cholesterol-lowering treatment.

Background and Aims

Patients with homozygous familial hypercholesterolaemia (HoFH) have a premature, high risk for ASCVD, and treatment should be initiated as early as possible. We evaluated paediatric HoFH patients from three pooled trials to describe safety and efficacy of evolocumab (a proprotein convertase subtilisin/kexin type 9 [PCSK9] inhibitor monoclonal antibody) and the relationship between LDL-C lowering and genetic mutations.

Methods

HoFH patients (aged 10-17 years) participated in the TAUSSIG, RAMAN, or HAUSER-OLE studies and received open-label subcutaneous evolocumab (420 mg QM or Q2W) with background lipid-lowering therapy for 12 to 260 weeks. Patients receiving lipoprotein apheresis (n=4) were analyzed separately. The primary endpoint for this analysis was TEAEs, analyzed per 100 patient-years. Efficacy endpoints were percent change in lipids at week 12 and LDL-C change by LDLR mutation type.

Results

Of 39 patients in the pooled analysis set, 69% were male, mean age was 13.5 (SD=1.8) years, and 82% had genotyped HoFH (Table). Overall, median (Q1, Q3) evolocumab exposure was 18.2 months (3.0, 18.5). TEAEs with exposure-adjusted incidence rate of ≥5% were upper respiratory tract infection (6.6%), influenza (5.2%), and acne (5.0%). The rate for serious TEAEs was 13.3%. At week 12, mean percent change from baseline LDL-C was –10.0% (SD=21.2) in non-apheresis patients, with high variability depending on residual LDLR activity (Figure).

Conclusions

In paediatric patients with HoFH, LDL-C reduction achieved during 12 weeks of PCSK9 inhibitor therapy was variable, depending on residual LDLR activity. However, several patients achieved a clinically meaningful reduction in LDL-C. Safety findings were consistent with previous studies of evolocumab.

Background and Aims

Low Cholesterol efflux capacity (CEC) and elevated levels of IL-1Beta (IL-1β) are strongly and independently associated with cardiovascular outcomes after Myocardial Infarction (MI). Both pathways are currently evaluated as new therapeutic targets in ongoing clinical trials. Animal studies demonstrated that alteration of cholesterol efflux leads to IL-1ß production, suggesting that defective CEC triggers inflammation.

The present study aims to evaluate (1) potential synergetic effect of CEC and IL-1β on recurrent major cardiovascular events (MACE) at one-year after a first MI and (2) the capacity of HDL particles to activate inflammation according to the risk of recurrent MACE.

Methods

CEC and IL-1B levels were measured among 2012 ST-segment elevation MI patients. The association of these biomarkers with the primary outcome of MACE was evaluated at one-year follow-up using a multivariate-cox analysis. In-vitro analyses of HDL functions including anti-inflammatory properties and CEC was evaluated in human THP-1 macrophages.

Results

Patients exhibited both reduced CEC and an elevated level of IL-1β displayed an increased risk of recurrent MACE at one year compared to patients with both high CEC and low IL-1β. HDL particles isolated from highest risk patients exhibited a reduced CEC from macrophages, as well as higher pro-inflammatory properties, as compared to patients at lower risk of recurrent MACE.

Conclusions

Defective CEC and elevated IL-1B increase synergistically the residual cardiovascular risk in MI patients. HDL particles from patients at highest risk of recurrent MACE after MI exhibit defective biological activities. These latter patients may benefit from a dual therapeutic approach targeting both CEC and inflammation.

Background and Aims

GWAS identified GPR146 as a novel lipid gene. Subsequent studies in mice validated Gpr146 deletion reduces circulating lipids and is atheroprotective, independent of LDL-R. Whereas, Mendelian randomisation studies highlight that GPR146 is causally linked with plasma levels of ALT, AST, and CRP. Several lines of evidence point to a role for ANGPTL3 mediating the effects of GPR146.

Methods

We used somatic gene editing to generate liver-specific deletions of Gpr146, Angptl3, and Gpr146/Angptl3. The experimental groups were fed with a regular chow and 60% kcal HFD, independently. Plasma levels of total cholesterol (TC), triglycerides (TG), ALT and AST and hepatic lipids were measured. Hepatic and aortic histology, RNAseq, qRT-PCR, western blotting and targeted proteomics were used to characterise the mouse model.

Results

Somatic gene editing showed 85% reductions in the targeted genes at the mRNA and protein levels. Compared to chow-fed controls, plasma TG was reduced by 28%, 58% and 75%, whereas plasma TC was reduced by 21%, 37% and 58% and hepatic lipid levels were reduced by 21%, 26%, and 46% in Gpr146, Angptl3 and Gpr146/Angptl3 deleted mice, respectively. Body and liver weights of the experimental groups remained unchanged.

Conclusions

This study shows that function of Gpr146 is independent of Angptl3, and combined deletion of both genes has a favourable outcome by reducing circulating and hepatic lipids in an additive manner. These findings indicate the therapeutic potential of targeting Gpr146 and Angptl3 to protect against atherosclerosis and fatty liver disease.

Background and Aims

PCSK9 is a serine protease that increases LDL-cholesterol levels due to endosomal and lysosomal degradation of the LDL-receptor. Studies have shown that PCSK9 is attached in all plasma lipoproteins. We investigated the effect of LDL and HDL-associated PCSK9 on platelet aggregation and on endothelial cell activation.

Methods

Washed platelets (WP), isolated from healthy volunteers, were incubated with 0.1mg/ml LDL or HDL in the presence or absence of 2.5μg/ml polyclonal antibody against PCSK9 (anti-PCSK9pAb) and activated with 0.1 units/ml thrombin or 0.25mM AA and monitored using Light Transmittance Aggregometry. Human Umbilical Vein Endothelial Cells (HUVECs) were cultured and pre-incubated with 0.1mg/ml LDL or HDL in the presence or absence of anti-PCSK9pAb for 5min followed by activation with 0.25ng/ml Tumor Necrosis Factor-α (TNF-α) for 6h. Cell activation was studied by flow cytometry for the membrane expression of ICAM-1 (Intercellular Adhesion Molecule-1).

Results

Thrombin and AΑ-induced platelet aggregation was inhibited by 50±19% and 38±11%, respectively in the presence of HDL and by 28±15 and 39±12%, respectively in the presence of LDL. These effects were not significantly influenced in the presence of anti-PCSK9pAb. ICAM-1 membrane expression was inhibited by 49±7.5% and 20±1.7% in the presence of HDL and LDL respectively, whereas these inhibitions were significantly reduced (p<0.05) in the presence of anti-PCSK9pAb by 22±6.9% and 9.2±0.8%, respectively.

Conclusions

HDL and LDL-associated PCSK9 does not affect the inhibitory effect of these lipoproteins on platelet activation whereas it significantly increases their inhibitory effect on inflammatory stimulation of endothelial cells.

Background and Aims

The outcomes of lipid-lowering therapy (LLT) vary between individuals. Previously we showed that for FH patients cellular lipid trafficking profiles are associated with on-treatment LDL-c levels (Pfisterer et al. CellRep Methods, doi:10.1016/j.crmeth.2022.100166). Here, we investigated how cellular profiles link with LDL-c levels, lipoprotein profiles, polygenic risk scores, and cardiovascular outcomes in the general population.

Methods

We used peripheral blood mononuclear cells (PBMCs) collected by THL Biobank as part of the Finnish population-based FINRISK 2012 study. We obtained samples for 400 subjects, including 200 recipients of LLT with available genetic, drug reimbursement, NMR metabolomic and longitudinal CVD outcome data. Using a multiplexed high-content imaging platform we obtained over 26 readouts for lipid uptake and storage in leukocytes for each sample.

Results

Cellular lipid uptake and storage were highly variable in our population-based study sample, and, in patients receiving either medium- or high-intensity statin monotherapy (n=39), were negatively correlated with serum LDL-cholesterol. Combining lipid uptake and storage readouts into cellular lipid trafficking scores (LTSs) improved this relationship (R=-0.45, p=0.0036). Subsequent combination of LTS with LDL-PRS produced even stronger correlation (R=-0.54, p=0.0004). Subjects in the lowest quintile of the LTS were at lower odds to be at their LDL target level (OR=0.068 CI [0.007,0.63], p=0.013) and at higher odds to experience MI or stroke (OR=30, 95% CI [2.64,339.75] p=0.004) in the 8-year follow-up period as compared to the rest of the group.

Conclusions

Cellular readouts provide novel insight into interindividual variation of LLT outcomes, providing new opportunities for treatment optimization and risk assessment for CVD prevention.