[Show abstract][Hide abstract] ABSTRACT: Background
The interpretation of the results from genome-scale experiments is a challenging and important problem in contemporary biomedical research. Biological networks that integrate experimental results with existing knowledge from biomedical databases and published literature can provide a rich resource and powerful basis for hypothesizing about mechanistic explanations for observed gene-phenotype relationships. However, the size and density of such networks often impede their efficient exploration and understanding.
We introduce a visual analytics approach that integrates interactive filtering of dense networks based on degree-of-interest functions with attribute-based layouts of the resulting subnetworks. The comparison of multiple subnetworks representing different analysis facets is facilitated through an interactive super-network that integrates brushing-and-linking techniques for highlighting components across networks. An implementation is freely available as a Cytoscape app.
We demonstrate the utility of our approach through two case studies using a dataset that combines clinical data with high-throughput data for studying the effect of β-blocker treatment on heart failure patients. Furthermore, we discuss our team-based iterative design and development process as well as the limitations and generalizability of our approach.
Electronic supplementary material
The online version of this article (doi:10.1186/s12859-015-0550-z) contains supplementary material, which is available to authorized users.
[Show abstract][Hide abstract] ABSTRACT: In heart failure with reduced left ventricular ejection fraction (HFrEF), adrenergic activation is a key compensatory mechanism that is a major contributor to progressive ventricular remodeling and worsening of heart failure. Targeting the increased adrenergic activation with β-adrenergic receptor blocking agents has led to the development of arguably the single most effective drug therapy for HFrEF. The pressure-overloaded and ultimately remodeled/failing right ventricle (RV) in pulmonary arterial hypertension (PAH) is also adrenergically activated, which raises the issue of whether an antiadrenergic strategy could be effectively employed in this setting. Anecdotal experience suggests that it will be challenging to administer an antiadrenergic treatment such as a β-blocking agent to patients with established moderate-severe PAH. However, the same types of data and commentary were prevalent early in the development of β-blockade for HFrEF treatment. In addition, in HFrEF approaches have been developed for delivering β-blocker therapy to patients who have extremely advanced heart failure, and these general principles could be applied to RV failure in PAH. This review examines the role played by adrenergic activation in the RV faced with PAH, contrasts PAH-RV remodeling with left ventricle remodeling in settings of sustained increases in afterload, and suggests a possible approach for safely delivering an antiadrenergic treatment to patients with RV dysfunction due to moderate-severe PAH.
[Show abstract][Hide abstract] ABSTRACT: -When β-blockers produce reverse-remodeling in idiopathic dilated cardiomyopathy (IDC) they partially reverse changes in fetal-adult/contractile protein, natriuretic peptide, SR-Ca2+-ATPase gene "program" constituents. The objective of the current study was to further test the hypothesis that reverse-remodeling is associated with favorable changes in myocardial gene expression by measuring additional contractile, signaling and metabolic genes that exhibit a fetal/adult expression predominance, are thyroid hormone-responsive, and/or are regulated by β1-adrenergic receptor signaling. A secondary objective was to identify which of these putative regulatory networks is most closely associated with observed changes.
-Forty-seven IDC patients (LVEF 0.24±0.09) were randomized to the adrenergic-receptor blockers metoprolol (β1-selective), metoprolol+doxazosin (β1/α1), or carvedilol (β1/β2/α1). Serial radionuclide ventriculography and endomyocardial biopsies were performed at baseline, 3, and 12 months. Expression of 50 mRNA gene products was measured by quantitative PCR. Thirty-one patients achieved LVEF reverse-remodeling response defined as improvement by ≥0.08 at 12 months or by ≥0.05 at 3 months (ΔLVEF=0.21±0.10). Changes in gene expression in Responders vs. Nonresponders were decreases in NPPA and NPPB and increases in MYH6, ATP2A2, PLN, RYR2, ADRA1A, ADRB1, MYL3, PDFKM, PDHX, and CPT1B. All except PDHX involved increase in adult or decrease in fetal cardiac genes, but 100% were concordant with changes predicted by inhibition of β1-adrenergic signaling.
-In addition to known gene expression changes, additional calcium-handling, sarcomeric, adrenergic signaling, and metabolic genes were associated with reverse-remodeling. The pattern suggests a fetal-adult paradigm but may be due to reversal of gene expression controlled by a 'β1-adrenergic receptor gene network.' Clinical Trial Registration-www.clinicaltrials.gov; Unique Identifier: NCT01798992.
[Show abstract][Hide abstract] ABSTRACT: Increasing availability of high-dimensional clinical data, which improves the ability to define more specific phenotypes, as well as molecular data, which can elucidate disease mechanisms, is a driving force and at the same time a major challenge for translational and personalized medicine. Successful research in this field requires an approach that ties together specific disease and health expertise with understanding of molecular data through statistical methods. We present PEAX (Phenotype-Expression Association eXplorer), built upon open-source software, which integrates visual phenotype model definition with statistical testing of expression data presented concurrently in a web-browser. The integration of data and analysis tasks in a single tool allows clinical domain experts to obtain new insights directly through exploration of relationships between multivariate phenotype models and gene expression data, showing the effects of model definition and modification while also exploiting potential meaningful associations between phenotype and miRNA-mRNA regulatory relationships. We combine the web visualization capabilities of Shiny and D3 with the power and speed of R for backend statistical analysis, in order to abstract the scripting required for repetitive analysis of sub-phenotype association. We describe the motivation for PEAX, demonstrate its utility through a use case involving heart failure research, and discuss computational challenges and observations. We show that our visual web-based representations are well-suited for rapid exploration of phenotype and gene expression association, facilitating insight and discovery by domain experts.
Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing 01/2015; 20:419-30. DOI:10.1142/9789814644730_0040
[Show abstract][Hide abstract] ABSTRACT: Objectives and background:
We evaluated the ability of 23 genetic variants to provide prognostic information in patients enrolled in the Genetic Substudy of the Surgical Treatment for Ischemic Heart Failure (STICH) trials.
Patients assigned to STICH Hypothesis 1 were randomized to medical therapy with or without coronary artery bypass grafting (CABG). Those assigned to STICH Hypothesis 2 were randomized to CABG or CABG with left ventricular reconstruction.
In patients assigned to STICH Hypothesis 2 (n = 714), no genetic variant met the prespecified Bonferroni-adjusted threshold for statistical significance (p < 0.002); however, several variants met nominal prognostic significance: variants in the β2-adrenergic receptor gene (β2-AR Gln27Glu) and in the A1-adenosine receptor gene (A1-717 T/G) were associated with an increased risk of a subject dying or being hospitalized for a cardiac problem (p = 0.027 and 0.031, respectively). These relationships remained nominally significant even after multivariable adjustment for prognostic clinical variables. However, none of the 23 genetic variants influenced all-cause mortality or the combination of death or cardiovascular hospitalization in the STICH Hypothesis 1 population (n = 532) by either univariate or multivariable analysis.
We were unable to identify the predictive genotypes in optimally treated patients in these two ischemic heart failure populations.
[Show abstract][Hide abstract] ABSTRACT: Because of its comparatively recent evolution, Homo sapiens exhibit relatively little within-species genomic diversity. However, because of genome size, a proportionately small amount of variation creates ample opportunities for both rare mutations that may cause disease as well as more common genetic variations that may be important in disease modification or pharmacogenetics. Primarily because of the East African origin of modern humans, individuals of African ancestry (AA) exhibit greater degrees of genetic diversity than more recently established populations, such as those of European ancestry (EA) or Asian ancestry. Those population effects extend to differences in frequency of common gene variants that may be important in heart failure natural history or therapy. For cell-signaling mechanisms important in heart failure, we review and present new data for genetic variation between AA and EA populations. Data indicate that: 1) neurohormonal signaling mechanisms frequently (16 of the 19 investigated polymorphisms) exhibit racial differences in the allele frequencies of variants comprising key constituents; 2) some of these differences in allele frequency may differentially affect the natural history of heart failure in AA compared with EA individuals; and 3) in many cases, these differences likely play a role in observed racial differences in drug or device response.
[Show abstract][Hide abstract] ABSTRACT: Rationale: H2RA use is common and may act directly on the heart through myocardial H2 receptors or indirectly through changes in pulmonary vascular resistance. Objective: To determine the relationship between histamine H2 receptor antagonist (H2RA) use and right ventricular (RV) morphology. Methods: We studied 4,124 participants in the Multi-Ethnic Study of Atherosclerosis without overt cardiovascular disease who had magnetic resonance imaging (MRI) assessment of RV morphology and ascertainment of medication use. Multivariable linear regression estimated cross-sectional associations between H2RA use and RV morphology after adjusting for demographics, anthropometrics, smoking status, diabetes mellitus and hypertension. Further adjustment for co-medication use, left ventricular (LV) parameters, lung structure and function, renal function, or inflammatory markers were considered in separate models. Analyses in a sub-cohort restricted to H2RA or proton pump inhibitor (PPI) users accounted for confounding by the indication of gastroesophageal reflux disease. Measurements and Main Results: H2RA use was associated with lower RV mass (-0.7 g, 95% confidence interval (CI): -1.2 to -0.2 g, p=0.004) and smaller RV end-diastolic volume (EDV) (-4.2 mL, 95% CI: -7.2 to -1.2 mL, p=0.006). This relationship was unchanged with adjustment for co-medication use, lung structure and function, renal function, and inflammation. The relationship with RV mass was independent of LV mass. Results were similar in a smaller sub-cohort restricted to PPI and H2RA users. Conclusion: H2RA use was associated with lower RV mass and smaller RVEDV. This finding is novel. Additional study of histamine and H2 receptors in cardiopulmonary diseases affecting the RV may have direct clinical relevance.
[Show abstract][Hide abstract] ABSTRACT: Background:
Atrial fibrillation and heart failure with reduced left ventricular ejection fraction have interrelated pathophysiologies. New-onset atrial fibrillation in heart failure patients has been associated with increased mortality, but has not been definitively related to clinical heart failure progression.
To test the hypothesis that new-onset atrial fibrillation is related to clinical heart failure progression, in 2392 patients without atrial fibrillation at randomization in the Beta-blocker Evaluation of Survival Trial we measured clinical endpoints in patients who did (Group 1, n = 190) or did not (Group 2, n = 2202) develop new-onset atrial fibrillation. Results were also compared with the 303 patients who entered the trial in atrial fibrillation (Baseline/chronic group), and in Group 1/2 patients we conducted a multivariate analysis of covariates potentially related to time to first heart failure hospitalization.
Compared with Group 2, Group 1 patients post atrial fibrillation onset had a ∼2-fold increase in mortality (P < .0001) and a ∼4.5-fold increase in all-cause or heart failure hospitalization days/patient (hospitalization burden, both P < .0001). In Group 1, both types of hospitalization burden were 2.9-fold greater than in the Baseline/chronic group (P < .001), and hospitalization burden increased ∼6-fold (P < .0001) compared with the pre-event period. On multivariate analysis, new-onset atrial fibrillation was a highly significant (P < .00001) predictor of heart failure hospitalization.
In addition to being a discrete electrophysiologic event, in heart failure patients, new-onset atrial fibrillation is a predictor of and trigger for clinical heart failure progression.
The American Journal of Medicine 06/2014; 127(10). DOI:10.1016/j.amjmed.2014.06.006 · 5.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: -Remodeling of myocardial phospholipids has been reported in various forms of heart failure for decades, but the mechanism and pathophysiological relevance of this phenomenon have remained unclear. We examined the hypothesis that delta-6 desaturase (D6D), the rate limiting enzyme in long-chain polyunsaturated fatty acid (PUFA) biosynthesis, mediates the signature pattern of fatty acid redistribution observed in myocardial phospholipids following chronic pressure overload, and explored plausible links between this process and disease pathogenesis.
-Compositional analysis of phospholipids from hearts explanted from patients with dilated cardiomyopathy revealed elevated PUFA product/precursor ratios reflective of D6D hyperactivity, manifesting primarily as lower levels of linoleic acid with reciprocally higher levels of arachidonic and docosahexaenoic acids. This pattern of remodeling was attenuated in failing hearts chronically unloaded with a left ventricular assist device. Chronic inhibition of D6D in vivo reversed similar patterns of myocardial PUFA redistribution in rat models of pressure overload and hypertensive heart disease, and significantly attenuated cardiac hypertrophy, fibrosis and contractile dysfunction in both models. D6D inhibition also attenuated myocardial elevations in pathogenic eicosanoid species, lipid peroxidation, and ERK1/2 activation; normalized cardiolipin composition in mitochondria; reduced circulating levels of inflammatory cytokines; and elicited model-specific effects on cardiac mitochondrial respiratory efficiency, NFκB activation and caspase activities.
-These studies demonstrate a pivotal role of essential fatty acid metabolism though D6D in myocardial phospholipid remodeling induced by hemodynamic stress, and reveal novel links between this phenomenon and the propagation of multiple pathogenic systems involved in maladaptive cardiac remodeling and contractile dysfunction.
[Show abstract][Hide abstract] ABSTRACT: The complexity of standard medical treatment for heart failure is growing, and such therapy typically involves 5 or more different medications. Given these pressures, there is increasing interest in harnessing cardiovascular biomarkers for clinical application to more effectively guide diagnosis, risk stratification, and therapy. It may be possible to realize an era of personalized medicine for heart failure treatment in which therapy is optimized and costs are controlled. The direct mechanistic coupling of biologic processes and therapies achieved in cancer treatment remains elusive in heart failure. Recent clinical trials and meta-analyses of biomarkers in heart failure have produced conflicting evidence. In this article, which comprises a summary of discussions from the Global Cardiovascular Clinical Trialists Forum held in Paris, France, we offer a brief overview of the background and rationale for biomarker testing in heart failure, describe opportunities and challenges from a regulatory perspective, and summarize current positions from government agencies in the United States and European Union.
[Show abstract][Hide abstract] ABSTRACT: This study assessed the impact of bucindolol, a beta-blocker/sympatholytic agent, on the development of atrial fibrillation (AF) in advanced chronic heart failure with reduced left ventricular ejection fraction (HFREF) patients enrolled in the BEST (Beta-Blocker Evaluation of Survival Trial).
β-Blockers have modest efficacy for AF prevention in HFREF patients. Bucindolol's effects on HF and ventricular arrhythmic endpoints are genetically modulated by β1- and α2c-adrenergic receptor (AR) polymorphisms that can be used to subdivide HFREF populations into those with bucindolol effectiveness levels that are enhanced, unchanged, or lost.
BEST enrolled 2,708 New York Heart Association (NYHA) class III to IV HFREF patients. A substudy in which 1,040 patients' DNA was genotyped for the β1-AR position 389 Arg/Gly and the α2c322-325 wild type (Wt)/deletion (Del) polymorphisms, and new-onset AF was assessed from adverse event case report forms or electrocardiograms at baseline and at 3 and 12 months.
In the entire cohort, bucindolol reduced the rate of new-onset AF compared to placebo by 41% (hazard ratio [HR]: 0.59 [95% confidence interval (CI): 0.44 to 0.79], p = 0.0004). In the 493 β1389 arginine homozygotes (Arg/Arg) in the DNA substudy, bucindolol reduced new-onset AF by 74% (HR: 0.26 [95% CI: 0.12 to 0.57]), with no effect in β1389 Gly carriers (HR: 1.01 [95% CI: 0.56 to 1.84], interaction test = 0.008). When β1389 Gly carriers were subdivided by α2c Wt homozygotes (n = 413, HR: 0.94 [95% CI: 0.48 to 1.82], p = 0.84) or Del variant carriers (n = 134, HR: 1.33 [95% CI: 0.32 to 5.64], p = 0.70), there was a positive interaction test (p = 0.016) when analyzed with β1389 Arg homozygotes.
Bucindolol prevented new-onset AF; β1 and α2c polymorphisms predicted therapeutic response; and the 47% of patients who were β1389 Arg homozygotes had an enhanced effect size of 74%. (Beta-Blocker Evaluation in Survival Trial [BEST]; NCT00000560).
[Show abstract][Hide abstract] ABSTRACT: Background:
Patients with heart failure and coronary artery disease often undergo coronary artery bypass grafting, but assessment of the risk of an adverse outcome in these patients is difficult. To evaluate the ability of biomarkers to contribute independent prognostic information in these patients, we measured levels in patients enrolled in the biomarker substudies of the Surgical Treatment for Ischemic Heart Failure (STICH) trials. Patients in STICH Hypothesis 1 were randomized to medical therapy or coronary artery bypass grafting, whereas those in STICH Hypothesis 2 were randomized to coronary artery bypass grafting or coronary artery bypass grafting with left ventricular reconstruction.
Methods and results:
In substudy patients assigned to STICH Hypothesis 1 (n=606), plasma levels of soluble tumor necrosis factor-α receptor-1 (sTNFR-1) and brain natriuretic peptide (BNP) were highly predictive of the primary outcome variable of mortality by univariate analysis (BNP: χ(2)=40.6; P<0.0001 and sTNFR-1: χ(2)=38.9; P<0.0001). When considered in the context of multivariable analysis, both BNP and sTNFR-1 contributed independent prognostic information beyond the information provided by a large array of clinical factors independent of treatment assignment. Consistent results were seen when assessing the predictive value of BNP and sTNFR-1 in patients assigned to STICH Hypothesis 2 (n=626). Both plasma levels of BNP (χ(2)=30.3) and sTNFR-1 (χ(2)=45.5) were highly predictive in univariate analysis (P<0.0001) and in multivariable analysis for the primary end point of death or cardiac hospitalization. In multivariable analysis, the prognostic information contributed by BNP (χ(2)=6.0; P=0.049) and sTNFR-1 (χ(2)=8.8; P=0.003) remained statistically significant even after accounting for other clinical information. Although the biomarkers added little discriminatory improvement to the clinical factors (increase in c-index ≤0.1), net reclassification improvement for the primary end points was 0.29 for BNP and 0.21 for sTNFR-1 in the Hypothesis 1 cohort, and 0.15 for BNP and 0.30 for sTNFR-1 in the Hypothesis 2 cohort, reflecting important predictive improvement.
Elevated levels of sTNFR-1 and BNP are strongly associated with outcomes, independent of therapy, in 2 large and independent studies, thus providing important cross-validation for the prognostic importance of these 2 biomarkers.