Bielinski, Suzette J.

Abstract Background Heart failure (HF) is a complex clinical syndrome that affects over 60 million persons worldwide.1 The 5-year survival rate of HF is less than 50% irrespective of ejection fraction (EF) and the limitations of EF in sufficiently phenotyping HF are recognized.2,3 High throughput proteomics assays are an attractive option for precision phenotyping.4 Purpose To explore the proteomic profiles and biomarker pathways associated with death in an HF community cohort. Methods We measured proteomics in plasma specimens collected from an HF cohort (2003–2012) of patients predominantly of European Ancestry (92%). The relative concentrations of circulating plasma proteins were measured using an aptamer-based assay containing 7335 human protein targets. Proteins significantly associated with mortality were selected while adjusting for age, sex, and eGFR. These proteins were employed in clustering analysis and subjected to 10-fold cross-validation. We quantified and corrected the clustering bias via bootstrapping and subsequently validated the results via semi-supervised clustering. Multivariable Cox regression was implemented for the cluster variable while adjusting for the Meta-analysis Global Group in Chronic Heart Failure (MAGGIC) risk score. Additionally, we performed enrichment and pathway analysis of the outcome-associated proteins considering significance thresholds of −log10(P-value) >2, FDR corrected P<0.05 (regulator analysis) and Z-scores of >2 and <−2 as thresholds for activation and inhibition of proteins/pathways. Results From a total of 1,388 patients (mean age 75±13; 52% male), 2 distinct clusters, (HR: 2.04; 95% CI [1.78–2.34]; p-value: <0.0001 independent of MAGGIC) were identified, cluster 1; n=722, mean age 72±14; NYHA class III/IV: 41%/26%; EF: 47%±17 and cluster 2; n=666, mean age 79±11; NYHA class III/IV: 41%/31%, EF: 48%±16 based on the 447 proteins [354 positively and 93 negatively] significantly associated with mortality (1,158 events). Gene-set enrichment analysis of all 447 proteins revealed cell adhesion, extracellular matrix organization, enzyme-linked protein/tyrosine kinase signaling pathways, humoral immune response, tissue development, growth factor signaling, post-translational protein modification, and platelet degranulation as the dominant biological themes. Finally, we found 18 upstream regulators predicted to be driving the expressions of the proteins in the dataset including a predicted outcome-associated protein-Transforming Growth Factor Beta 1 – TGFB1 (p=7.46E-09), Interleukin 1B, IL1B (p=9.40E-08), and oncostatin M, OSM (p=6.61E-05) as the top 3 regulators (predicted activation) based on FDR-corrected statistical significance of P<0.05 (Figure 1) Conclusion We identified distinct clusters based on proteomic profiles associated with HF mortality independently of clinical prognostic makers. Further, we delineated protein regulators driving the major pathways associated with HF survival. Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): Department of Intramural Research, NIH/NHLBI

Abstract Background Risk factors for the pathogenesis of chronic rhinosinusitis (CRS) remain largely undetermined, which is likely due to the heterogeneity of the disease. White blood cell counts have been largely unexplored as a risk factor for CRS even though different types of white blood cells are involved in the inflammatory process of CRS. Objective To investigate causal associations between different types of white blood cells on risk of CRS utilizing a Mendelian randomization (MR) analysis. Methods A two-sample MR analysis was performed using respective GWAS summary statistics for the exposure traits (neutrophil count, eosinophil count, basophil count, lymphocyte count, and monocyte count) and outcome trait (CRS). For the exposure traits, the European Bioinformatics Institute database of complete GWAS summary data was used. For the outcome trait, summary statistics for CRS GWAS were obtained from FinnGen. Primary analysis for MR was performed using inverse-variance weighted two-sample MR. Sensitivity analyses included weighted median, MR-Egger, and MR-PRESSO (raw and outlier-corrected). Results Eosinophils were associated with CRS (OR = 1.55 [95%CI 1.38,1.73]; p = 4.3E-14). Eosinophil results were similar across additional MR methods. MR results did not demonstrate significant causal relationships between neutrophils, lymphocytes, monocytes, or basophils with CRS. No significant pleiotropic bias was observed. Conclusions In a two-sample MR analysis, a potential causal link between blood eosinophil counts and CRS has been demonstrated. In addition, causal relationships between blood counts among other white blood cell types and CRS were not found. Further studies involving genetic variation in CRS are needed to corroborate genetic causal effects for CRS.

Background: Preeclampsia (PE) is a pregnancy-specific hypertensive disorder which affects up to 7% of all pregnancies. It is associated with increased risk for cardiovascular disease later in life. We have recently reported increased senescence and accelerated aging in women with PE. We hypothesize that senescence burden remains after PE. The aim of this study was to compare senescence profiles, and history of individual chronic conditions and multimorbidity, a marker of accelerated aging, in women with and without a history of PE. Methods: Forty women with histories of PE and 40 with histories of normotensive pregnancy confirmed by manual chart review were age- and parity-matched. None of the women had a history of CVD at the time of recruitment. Markers of senescence (p16, p21 and Klotho) and metabolic syndrome (leptin, adiponectin) were measured and compared between women with and without histories of preeclampsia using conditional logistic regression. The rates of chronic conditions and multimorbidity between the groups were also compared. Results: Urinary Klotho/creatinine levels were significantly lower (median [IQR]: 117 [87.4, 224] vs. 205 [101, 359], respectively, p=0.042), while leptin/adiponectin ratio was significantly higher (median [IQR]: 0.25 [0.13, 0.45] vs. 0.05 [0.02, 0.12], p=0.027) among women with histories of PE compared to controls. There were no significant differences observed in plasma p21 or p16 concentrations. Women with histories of PE experienced a higher prevalence of two or more co-morbidities compared to women with normotensive pregnancies [67.5% (27/40) vs. 22.5% (9/40), p<0.05]. Results were similar after removing hypertension as a comorbidity [45% (18/40) vs. 20.0% (8/40), p<0.05]. Conclusion: Women with a history of PE had altered markers of senescence, metabolic syndrome and multimorbidity decades after PE.

Background Hepatocyte growth factor (HGF) is a biomarker with potential for use in the diagnosis, treatment and prognostication of cardiovascular disease (CVD). Elevated HGF is associated with calcification in the coronary arteries. However, knowledge is limited on the role HGF may play in extracoronary calcification (ECC). This study examined whether HGF is associated with ECC in the aortic valve (AVC), mitral annulus (MAC), ascending thoracic aorta and descending thoracic aortic (DTAC). Methods At baseline, adults aged 45–84 years, free of CVD, in the Multi-Ethnic Study of Atherosclerosis had HGF and ECC measured by ELISA and cardiac CT scan, respectively. ECC measurements were repeated after an average of 2.4 years of follow-up. Prevalent ECC was defined as Agatston score >0 at baseline. Incident ECC was defined as Agatston score >0 at follow-up among participants with Agatston score=0 at baseline. We used Poisson and linear mixed-effects regression models to estimate the association between HGF and ECC, adjusted for sociodemographic and CVD risk factors. Results Of 6648 participants, 53% were women. Mean (SD) age was 62 (10) years. Median (IQR) of HGF was 905 (757-1087) pg/mL. After adjustment for CVD risk factors, the highest HGF levels (tertile 3) were associated with greater prevalence and extent of AVC, MAC and DTAC at baseline compared with the lowest tertile (tertile 1). Additionally, the risk of incident AVC and MAC increased by 62% and 45%, respectively, in demographic-adjusted models. However, the associations were not statistically significant in fully adjusted models. The highest HGF levels were also associated with 10% and 13% increase in MAC and DTAC progression, respectively, even after adjustment for CVD risk factors. Conclusion Higher HGF levels were significantly associated with a greater risk of calcification at some extracoronary sites, suggesting an alternate biological pathway that could be targeted to reduce CVD risk.

Introduction: The Framingham Heart Study Dementia Risk Score (FDRS) was developed in a general population of elderly individuals. It is unknown how the FDRS variables predict dementia in heart failure (HF) and atrial fibrillation (AF) populations. Thus, we aimed to evaluate the predictive ability of the FDRS variables in population-based cohorts of individuals with HF and AF and to determine whether the addition of other comorbidities and risk factors improves risk prediction for dementia. Methods: Residents aged ≥50 from 7 southeast MN counties with a first diagnosis of HF (ICD-9 428/ICD-10 I50) or AF (ICD-9 427.3/ICD-10 I48 with AF on electrocardiogram/Holter monitor) between 1/1/2013 and 12/31/2017 were identified. Patients with dementia (defined by ICD codes) prior to or within 6 months after their index event and patients who died within 6 months after index were excluded. For both cohorts, models were constructed to predict dementia ≥ 6 months post index including the variables in the FDRS (age, marital status, BMI, stroke/TIA, diabetes and cancer). Additional comorbidities and risk factors were added to the models. For all models c-statistics were calculated. Results: Among 3,291 patients with HF (mean age 76 years, 53% male), 626 cases of dementia occurred and among 4,107 patients with AF (mean age 74 years, 57% male), 736 cases of dementia occurred. Among patients with HF, the FDRS variables predicted dementia with a c-statistic of 0.68 (Table). Adding additonal comorbidities and risk factors impoved the c-statistic slightly to 0.69 (p=0.01). Among patients with AF, the FDRS variables also performed well (c-stastic: 0.73; Table) and adding other comorbidities and risk factors improved the performance (c-statistic: 0.75; p<0.01). Conclusions: The variables from the FDRS predict dementia well in both HF and AF populations. The addition of comorbidities and other CVD risk factors only modestly improved prediction, thus indicating that the FDRS variables are appropriate to predict dementia in patients with HF and AF.

Introduction: Heart failure (HF) is a clinical syndrome with high mortality. Left ventricular ejection fraction (EF) determines clinical phenotyping and HF is classified most often as preserved (HFpEF) or reduced (HFrEF) EF. The importance of expanding phenotyping beyond EF is recognized. We hypothesized that machine learning applied to echocardiography will provide information important for deep phenotyping. Methods: We collected clinical and echo data in a HF community cohort between 2013 and 2017. Unsupervised clustering was applied to 23 echo variables using partitioning around medoids (PAM), with Gower dissimilarity used to address the mixed variable types. Proportional hazards regression examined the association between cluster assignment and death. Results: We identified 4978 persons with HF [mean age 73.9 years (SD 13.9), 47% women] with echocardiography within 6 months of diagnosis. We identified 2 primary clusters using echo variables, as depicted using t-Distributed Stochastic Neighbor Embedding (t-SNE) applied to the dissimilarity matrix (Figure). Diastolic function variables exhibited the largest differences by clusters as shown by absolute standardized differences (Table). After 3.2 + 1.9 years of follow up, 2001 deaths occurred. Assignment to cluster 2 was associated with a near 20% excess risk of death independent of age, sex, comorbidity, and EF [HR 1.17 (1.07, 1.28) p<0.001]. Conclusions: Unsupervised clustering applied to echocardiography in a large community HF cohort identified 2 distinct clusters, which defined HF subgroups with different characteristics and were independently associated with risk of death.

Introduction: Hepatocyte growth factor (HGF) is released in response to endothelial injury, and higher levels of circulating HGF are associated with atherosclerosis. However, the association of HGF with atherosclerosis progression is unknown. Hypothesis: Circulating HGF is positively associated with atherosclerosis progression, measured as progression of 1) carotid artery plaque and 2) coronary artery calcium (CAC). Methods: The Multi-Ethnic Study of Atherosclerosis is a large, population-based cohort study of adults aged 45-84 and free of clinical cardiovascular disease. Participants had HGF measured in serum at baseline (2000-02), and were followed for progression of carotid plaque and CAC through 2012. We defined CAC and carotid plaque progression as dichotomous variables (progression, no progression). Carotid plaque progression is an increase in the number of carotid plaques. We used the Hokanson method to define CAC progression as an increase of ≥ 2.5 mm 3 in the square root-transformed calcium volume score. Relative risk regression quantified the association between HGF and carotid plaque progression. Cox regression models using inverse-probability-of-attrition weights estimated the association of HGF with progression of CAC. All effect estimates are adjusted for potential confounding variables. Results: The cohort included 6714 Americans with HGF measurements: 2593 non-Hispanic white, 797 Chinese, 1848 black, and 1476 Hispanic Americans. In 3363 participants with measurements for both carotid plaque and HGF, each standard deviation (259 pg/ml) increase in HGF was associated with a 4% greater risk (95% CI: 1.01-1.07) of carotid plaque progression (n=1875), which did not differ by race/ethnicity (interaction p = 0.3). Out of 6714 participants with measurements for both CAC and HGF, 2994 had CAC progression during follow-up. The association of HGF with CAC progression differed by race/ethnicity (interaction p = 0.01). Each standard deviation increase of HGF was associated with a 10% greater risk of CAC progression in non-Hispanic white (95% CI: 1.05-1.16) and black Americans (95% CI: 1.03-1.17), while HGF was not associated with CAC progression in Chinese (p = 0.5) nor Hispanic (p = 0.7) American participants. Conclusions: Circulating HGF was associated with a modestly increased risk of atherosclerosis progression in this large and diverse population, with the strongest associations between HGF and CAC progression in non-Hispanic white and black Americans. Our results suggest that HGF may have utility as a marker of atherosclerosis progression in some populations.

Abstract Background Cognitive function is essential to effective self‐management of heart failure (HF). Alzheimer's disease and Alzheimer's disease‐related dementias (AD/ADRD) can coexist with HF, but its exact prevalence and impact on health care utilization and death are not well defined. Methods Residents from 7 southeast Minnesota counties with a first‐ever diagnosis code for HF between January 1, 2013 and December 31, 2018 were identified. Clinically diagnosed AD/ADRD was ascertained using the Centers for Medicare and Medicaid (CMS) Chronic Conditions Data Warehouse algorithm. Patients were followed through March 31, 2020. Cox and Andersen‐Gill models were used to examine associations between AD/ADRD (before and after HF) and death and hospitalizations, respectively. Results Among 6336 patients with HF (mean age [SD] 75 years [14], 48% female), 644 (10%) carried a diagnosis of AD/ADRD at index HF diagnosis. The 3‐year cumulative incidence of AD/ADRD after HF diagnosis was 17%. During follow‐up (mean [SD] 3.2 [1.9] years), 2618 deaths and 15,475 hospitalizations occurred. After adjustment, patients with AD/ADRD before HF had nearly a 2.7 times increased risk of death, but no increased risk of hospitalization compared to those without AD/ADRD. When AD/ADRD was diagnosed after the index HF date, patients experienced a 3.7 times increased risk of death and a 73% increased risk of hospitalization compared to those who remain free of AD/ADRD. Conclusions In a large, community cohort of patients with incident HF, the burden of AD/ADRD is quite high as more than one‐fourth of patients with HF received a diagnosis of AD/ADRD either before or after HF diagnosis. AD/ADRD markedly increases the risk of adverse outcomes in HF underscoring the need for future studies focused on holistic approaches to improve outcomes.

P-selectin is a cellular adhesion molecule that enhances procoagulant activity and activation of leukocyte integrins. P-selectin has been shown to be crucial in the development of atherosclerosis. Using a stratified random sample of 2406 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) at Exam 2, we investigated the association of soluble P-selectin and cardiovascular disease. The Tobit model was used to investigate the association of P-selectin levels and coronary artery calcium (CAC) and Cox proportional hazards regression was used to assess the association between protein level and incident cardiovascular disease (CVD) events (myocardial infarction, stroke, angina, or other atherosclerotic death). Mean levels of soluble P-selectin differed significantly by ethnicity; 46±18 ng/mL African Americans, 58±20 ng/mL Caucasians, 43±15 ng/mL Chinese, and 61±20 ng/mL Hispanics (p<0.0001). CAC was present in 46% of African Americans, 61% of Caucasians, 52% of Chinese, and 53% of Hispanics. P-selectin levels were significantly associated with CAC in Caucasians (p < 0.001) and African Americans (p = 0.025), see Table. This association was robust to adjustment for traditional risk factors and a significant interaction of P-selectin and ethnicity was observed (p= 0.02). Over an average of 6.5 years of follow-up there were 111 incident CVD events after MESA exam 2, 26 in African Americans, 33 in Caucasians, 16 in Chinese, and 36 in Hispanics. The risk of CVD events increased 34% per standard deviation increase in P-selectin levels only in Caucasians (p = 0.007) and this finding was consistent after adjustment for CVD risk factors (HR = 1.30, p = 0.017). We have shown in this large cohort that mean circulating P-selectin levels differed by ethnicity. Despite similar percentages of subjects with CAC in each ethnic group, P-selectin levels were significantly associated with CAC only in African Americans and Caucasians. Furthermore, in Caucasians only, P-selectin levels were associated with incident CVD events. Association of P-selectin, per SD (20 ng/mL) for the population and cardiovascular disease Outcome African Americans n = 579 Caucasians n = 620 Chinese n = 602 Hispanics n = 605 CAC, Agatston Score Beta (S.E.) P-value Beta (S.E.) P-value Beta (S.E.) P-value Beta (S.E.) P-value Model 1 85.9 (38.4) 0.025 99.4 (28.7) <0.001 8.2 (33.3) 0.81 32.1 (27.3) 0.24 Model 2 79.3 (39.8) 0.046 94.7 (29.1) 0.001 −19.5 (33.6) 0.56 27.8 (27.6) 0.31 CVD Events HR (CI) HR (CI) HR (CI) HR (CI) Model 1 1.08 (0.75, 1.57) 0.67 1.34 (1.08, 1.65) 0.007 1.09 (0.57, 2.08) 0.80 0.99 (0.71, 1.39) 0.97 Model 2 1.10 (0.76, 1.58) 0.61 1.30 (1.05, 1.62) 0.017 0.82 (0.38, 1.78) 0.62 1.00 (0.71, 1.43) 0.98 Model 1 = age and sex ; Model 2 = age, sex, BMI, smoking status, LDL, HDL, triglycerides, hypertension, diabetes