Search

Introduction:Canagliflozin (CAN) vs. placebo (PBO) reduced the risk of hospitalization for heart failure (HHF) in type 2 diabetes. Whether this benefit is uniform across diabetes-specific HF risk scores (WATCH-DM and TRS-HFDM) is not known.Methods:Using data from the pooled CANVAS and CANVAS-R trials, we stratified participants without prevalent HF by the integer WATCH-DM score derived quintiles (low [≤14] = quintiles 1-2, intermediate [15-19] = quintiles 3-4, high [≥20] = quintile 5) and by the TRS-HFDMscore (low [0-1], intermediate [2], high [3-6]). Discrimination and calibration were assessed by Harrell’s C-index and Hosmer-Lemeshow test, respectively. Cox regression models evaluated the effect of CAN on risk of HHF across risk score categories.Results:Among participants without prevalent HF (n = 8,691), CAN vs. PBO reduced the risk of HHF (HR 0.80, 95% CI 0.62-1.03). The WATCH-DM score demonstrated a C-index of 0.70 (95% CI, 0.66-0.73) and no evidence of miscalibration (χ2< 20). CAN consistently reduced risk of HHF across WATCH-DM strata (P-intxn = 0.55) with the greatest absolute risk reduction and lowest NNT observed in the highest vs. lowest risk cohort (ARR 4.6% vs. 0.3% and NNT 22 vs. 333) (Fig. A). Comparatively, the TRS-HFDMdemonstrated a C-index of 0.67 (95% CI, 0.63-0.71) and no evidence of miscalibration (χ2< 20). Similar to WATCH-DM, patients in the highest TRS-HFDMrisk group derived the greatest absolute risk reduction and lowest NNT (ARR 3.1% vs. -0.3%% and NNT 32 vs. -333) with no differences across strata (P-intxn = 0.17) (Fig. B).Conclusions:Both the WATCH-DM and TRS-HFDMcan accurately stratify HHF risk in patients with type 2 diabetes and free of HF. Greater absolute risk reductions with CAN vs PBO were observed with higher risk scores.

Introduction:Trimethylamine N-oxide (TMAO) is a gut microbiota-derived metabolite of dietary phosphatidylcholine and carnitine. Experimentally, TMAO causes renal tubulointerstitial fibrosis and kidney injury. Yet, little is known about prospective associations between TMAO and renal outcomes.Hypothesis:Higher plasma TMAO levels are prospectively associated with higher risk of incident chronic kidney disease (CKD) and greater renal function decline.Methods:We included 4,131 US adults from the Cardiovascular Health Study, a community-based cohort, with normal baseline renal function (estimated glomerular filtration rate [eGFR] ≥ 60 mL/min/1.73m2). TMAO was measured using liquid chromatography-mass spectrometry at baseline and year 7. Creatinine and Cystatin C were measured 4 times during serial follow-up and used to compute eGFR. Incident CKD was defined by eGFR (decline ≥ 30% plus eGFR<60 at a follow-up visit) or presence of CKD hospitalization ICD codes. Time-varying Cox models assessed how serial TMAO measures related to incident CKD, adjusting for sociodemographic, lifestyle, diet, and CVD risk factors. Linear regression models assessed how baseline TMAO related to annualized eGFR change in 3,997 participants with at least one follow-up eGFR measure.Results:During a median follow-up of 14 years, 1,136 participants developed CKD. There was a non-linear dose-response relationship of TMAO with CKD incidence: higher TMAO levels were associated with higher risk of incident CKD with a plateauing of risk above the ~75thpercentile of TMAO levels (top Figure). Higher baseline TMAO levels were also associated with greater annualized eGFR decline (P-linearity=0.004. bottom Figure) after adjusting for baseline eGFR and other covariates.Conclusions:In this community-based cohort of older US adults, higher serial measures of plasma TMAO were associated with a higher risk of incident CKD and a greater rate of annualized renal function decline.

Introduction:Older age is independently associated with greater NT-proBNP levels. As such, the aim of this study was to establish the influence of age on serial NT-proBNP levels over time and the impact of age on response to natriuretic peptide-guided therapy (NPGT) versus usual care for heart failure with reduced ejection fraction (HFrEF).Hypothesis:Age will modify the effect of NPGT on serial NT-proBNP in patients with HFrEF.Methods:Using data from GUIDE-IT, the impact of age on the association of NPGT with serial NT-proBNP and long-term outcomes was assessed. Time-to-event analyses and generalized linear mixed modeling were used to evaluate the effect of age, by quartile, on long-term outcomes and serial NT-proBNP by randomized treatment (NPGT vs. usual care), respectively.Results:In this cohort (N=893), the median age was 63 [range 21-90, IQR 53-71] years and was 32% female and 36% black race. Age was modestly correlated with NT-proBNP (Spearman rho 0.34, p<0.001). At baseline, older age was associated with lower doses of ACEI/ARB (p-trend 0.019) and aldosterone antagonist use (p-trend<0.001), but not beta-blocker use (p=0.06). Older age was associated with a greater risk of all-cause death (log-rank, p<0.001) but not associated with a greater risk of combined CV death or HF hospitalization (log-rank, P=0.56). Although older age was associated with greater NT-proBNP levels over time (Q4 vs. Q1 beta 2.1 x pg/mL/days, p<0.001), it did not modify the association of NPGT on serial NT-proBNP (Figure,Q4, Q3, and Q2 vs. Q1 by treatment arm, p-interaction>0.09 for all). Age also did not modify the impact of NPGT on either mortality or the composite, CV death or HF hospitalization (P-interaction>0.05 for both).Conclusions:Older age was associated with greater NT-proBNP levels. However, age did not modify the association of NPGT on clinical outcomes or serial NT-proBNP. These data do not support the hypothesis that age modifies the NT-proBNP response to HFrEF treatments.

Introduction:Sarcopenia is associated with worse outcomes in various clinical situations. Traditional markers of strength and frailty have been used for pre-operative risk stratification in transcatheter aortic valve replacement (TAVR). However, the availability of computed tomography (CT) scans provides an opportunity to obtain direct skeletal muscle measurements.Hypothesis:We hypothesized that sarcopenia would lead to worse outcomes in patients following TAVR.Methods:Patients undergoing TAVR between January to July 2018 with pre-procedural chest CT were included. Semi-automatic measurements of skeletal muscle area (SMA) were made at the twelfth thoracic vertebra. SMA was normalized by height to obtain skeletal muscle index (SMI, cm2/m2). Sarcopenia was defined as the lowest sex-stratified SMI tertile. Strength and functional testing data had been collected as part of the routine pre-TAVR evaluation. The primary outcome of interest was all-cause mortality.Results:A total of 76 patients were included, 26 sarcopenic based on SMI. Table 1 shows comparisons between the groups. During a median follow-up of 1496 (1401-1562) days, 10 (38.5%) deaths occurred in the sarcopenic group and 9 (18.0%) in the non-sarcopenic group. Figure 1 demonstrates this significant difference by Log-Rank testing (p=0.042). The secondary outcomes of length of stay and 30-day readmission did not differ between the groups on unadjusted comparison.Conclusions:Sarcopenia was associated with increased mortality in patients who underwent TAVR. A larger study is underway to assess the potential of muscle measurements serving as an additional pre-operative risk stratification tool.

Introduction:The hemodynamic gain index (HGI) is a novel simple parameter calculated from resting and peak systolic BP (SBP) and heart rate (HR) during exercise testing. It is a predictor of mortality in population-based cohorts. However, the prognostic significance of the HGI in chronic HFrEF has not been well studied.Hypothesis:The HGI is independently associated with adverse outcomes, and is comparable to other cardiopulmonary exercise testing (CPET) parameters.Methods:Medical records of 1,067 consecutive HFrEF patients (EF ≤ 40%) undergoing CPET for symptom evaluation from 12/2012 to 9/2020 were reviewed. HGI was calculated using the formula, [(SBPpeakx HRpeak)-(SBPrestx HRrest)]/(SBPrestx HRrest). Patients with missing vital signs, hypotensive or bradycardic response to exercise were excluded. Primary outcome was the composite endpoint of all-cause mortality, LVAD implantation, and heart transplantation. Multivariable Cox proportional hazard models were used with subgroup analyses based on median age, sex, BMI of 35 kg/m2, respiratory exchange ratio (RER) of 1.05, and beta-blocker use. ROC curves with AUCs were used to compare HGI, peak VO2, VE/VCO2slope and peak end-tidal pressure of CO2(PEtCO2).Results:We included 954 HFrEF patients (mean age was 56.3 ± 12.0 years, 72% men, 17% with BMI ≥ 35 kg/m2, 86% on beta-blockers, and 73% with RER > 1.05). During a median follow up time of 946 days, the incidence of the composite outcome was 331 (34.7%). After adjustment for age, sex, BMI, comorbidities, and EF, higher HGI was independently associated with lower risk of the adverse outcomes in the main cohort (hazard ratio per unit increase 0.46, 95%CI 0.35 to 0.59, p< 0.001), and all subgroups, except for patients with BMI ≥ 35 kg/m2(Figure 1A). The HGI was also comparable to peak VO2, and outperformed VE/VCO2slope and PEtCO2(Figure 1B).Conclusions:The HGI is an independent predictor of adverse outcomes, and comparable to peak VO2in patients with chronic HFrEF.

Introduction:Low serum albumin is a marker of protein malnutrition and is commonly associated with worse outcomes in various clinical settings. Furthermore, significant overlap exists between malnutrition and sarcopenia, which can be an independent predictor of worse outcomes.Hypothesis:We assessed the hypothesis that the presence of sarcopenia with low albumin (SLA) would lead to synergistically worse outcomes in patients with acute decompensated heart failure (ADHF).Methods:Patients hospitalized for ADHF from 2017 to 2019 with computed tomography of the abdomen/pelvis within 30 days and albumin level within 24 hours before discharge were studied (n=181). Given the high prevalence of hypoalbuminemia, low albumin was defined as the lower fiftieth percentile. Semi-automatic measurements of skeletal muscle area were made at L3 (Figure 1A) and adjusted using height squared to obtain skeletal muscle index (SMI). Sarcopenia was defined as the lowest sex-stratified SMI tertile.Results:The prevalence of sarcopenia alone was 11.6%, low albumin alone 28.7%, and SLA 20.4%. The groups had similar demographics but differed in BMI (lowest in sarcopenia alone, p<0.001) and admission NT-proBNP (highest in SLA, p=0.029). Patients with SLA were discharged to facilities most often (55.2%, p<0.001). All-cause mortality differed among the groups (p<0.001) during a median follow-up of 23.4 (4.2-45.0) months (Figure 1B). Compared to controls, those with SLA were at the highest risk (HR 2.46, 95%CI 1.50-4.02, p<0.001), followed by patients with low albumin alone (HR 1.77, 95%CI 1.11-2.83, p=0.016) or sarcopenia alone (HR 1.73, 95%CI 0.93-3.20, p=0.082), the latter not reaching significance.Conclusions:In conclusion, sarcopenia with low serum albumin is associated with higher mortality. Available resources and interventions should be utilized during ADHF hospitalization to optimize nutrition for these patients.

Introduction:While implicated in both cardiac and renal dysfunction, it remains unclear if myeloperoxidase (MPO) plays a role in the development of acute cardiorenal syndrome (CRS). We developed an assay to quantify MPO inhibitory capacity (MIC) to assess a plasma sample’s capacity to inhibit MPO activity.Hypothesis:We hypothesized that unopposed MPO activity (less inhibition) contributes to development of acute CRS and greater MPO inhibition leads to improved clinical outcomes.Methods:90 paired samples of acute heart failure (AHF) patients were collected at enrollment and between 24-96 hours follow-up. Diluted samples were supplemented with 100 ng/mL exogenous MPO and allowed to equilibrate for 1 hour at room temperature. Equal parts sample and Amplex UltraRed Reagent were incubated for 30 minutes, and fluorescence was read. 59 patients had long-term follow-up data for time to readmission survival analysis, censored for loss to follow up, death, or transplantation.Results:Patients with CRS have higher rather than lower MIC compared to that in uncomplicated patients (50% vs 34%, p=0.048). Serial MIC measurements during hospitalization revealed that patients with persistently low MIC experienced better outcomes (median time to readmission 575 days vs 104 days, p=0.023, Figure).Conclusions:Contrary to our original hypothesis, we observed acute CRS patients had higher rather than lower MIC, and that persistently low MIC experienced better long-term outcomes. These results imply that in patients experiencing acute CRS, there is a circulating MPO inhibitory component that has not been previously accounted for, and that unopposed MPO activity was not associated with CRS or poor outcomes.

Background:Cardiac magnetic resonance imaging (CMR) is the reference standard for right ventricle (RV) quantification, however predictors and prognosis of RV dysfunction in cardiomyopathy patients, including differences between ischemic (ICM) and non-ischemic (NICM) cardiomyopathies, remain not well understood. We evaluated the factors and outcomes associated with CMR-derived reduced RV systolic function, and compared by type of cardiomyopathy.Methods:Adult cardiomyopathy patients undergoing CMR with RV quantified during 2002-2017 were retrospectively studied. Multivariable linear and Cox regression were used to identify factors associated with reduced RV ejection fraction (RVEF) and the primary endpoint (all-cause death, heart transplant or LVAD) during follow-up.Results:Amongst 771 ICM and 624 NICM patients (mean ages 62.3±11.2 and 52.8±15.9 years, 190 (24.6%) and 253 (40.5%) were female respectively), mean RVEFs were 43±14% and 43±13% respectively. The primary endpoint occurred in 455 (59.0%) ICM and 113 (18.1%) NICM patients over mean follow-up of 5.3±4.5 years. Predictors of lower RVEF in ICM were younger age, lower eGFR, lower left ventricular ejection fraction, higher right ventricular end-diastolic volume indexed and higher mitral regurgitant fraction; and in NICM were male sex, NYHA class III-IV, right bundle branch block, lower left ventricular ejection fraction, higher right ventricular end-diastolic volume and higher mitral regurgitation fraction. Reduced RVEF was associated with the primary endpoint in univariable analysis in both ICM and NICM, but only remained so for NICM in multivariable analysis of the primary endpoint as shown in the table.Conclusion:Reduced RVEF is associated with adverse prognosis, however was only independently associated with the primary endpoint in NCIM but not ICM. Key predictors of RV dysfunction in both ICM and NICM were identified, with some similar and some different predictors between them.

Background:Deep learning (DL) models typically interpret images without prior knowledge of anatomic significance. However, pathophysiology is highly classified by human definitions according to underlying affected anatomy. Therefore, we examine the impact of introducing explicit knowledge of anatomy through cardiac contours on cardiac magnetic resonance images (CMR) to DL models. The DL models were then trained to differentiate between ischemic cardiomyopathy (ICM) and non-ischemic cardiomyopathy (NICM) using late gadolinium enhanced (LGE) CMR.Method:We evaluated 301 CMR studies; ICM (n=176) and NICM (n=125). Manual contouring of LGE images was performed using CVI42 (Circle, Ontario). We compared a radiomic and end-to-end DL approach to identify cardiomyopathy (CM) etiology from short axis LGE images. Patients are randomly assigned to training (70%) and testing (30%). Model performance was assessed with area under curve (AUC).Result:Table 1 shows the results of radiomic and deep learning approaches to differentiate NICM and ICM. Segmented/manually contoured images with removal of non-cardiac structures greatly improved classification accuracy across all deep learning models. The average improvement in AUC was 0.163 when using segmented images compared to the full images. Furthermore, the deep learning models outperformed the radiomics models. The best radiomic model and deep learning model achieved AUCs of 0.914 and 0.947, respectively. Both radiomic based models achieved AUCs above 0.874 while all 2D deep learning models with segmented images achieved AUCs above 0.875. The two 3D deep learning models which utilize 3D convolutions provided lower AUCs ranging between 0.743 and 0.900.Conclusion:Manual segmentation of LGE images improved the ability to train DL models with fairly small volumes of labeled data, resulting in higher classification accuracy.Table1 AUCs of various models.

Introduction:Asking patients about sodium intake is standard practice in heart failure (HF).Hypothesis:Self-reported sodium intake in HF with preserved ejection fraction (HFpEF) reflects a vulnerability to sodium intake driven by aldosterone.Methods:The TOPCAT trial registered self-reported sodium intake at baseline. Cox-regression and linear-mixed models were used to assess the relationship between self-reported sodium intake and outcome. Interaction analysis between self-reported sodium intake and the treatment effect of spironolactone on HF-outcome, blood pressure (BP), dyspnea and edema at follow-up were used to assess sodium vulnerability mediated by an aldosterone effect.Results:Self-reported sodium intake of 1748 HFpEF patients included in TOPCAT were divided according to tertiles of sodium intake (47% low, 35% moderate and 18% reported high sodium intake). After covariate adjustment lower self-reported sodium intake associated with higher risk for HF-admission (p=0.009). Patients with a lower sodium intake demonstrated higher E-wave and LV-end diastolic volume and higher estimated plasma volume (p<0.001). Lower sodium intake was associated with a larger treatment effect of spironolactone on HF-admission (HR=0.69, 95%CI 0.53-0.91 vs highest tertile HR=1.37 95%CI 0.79-2.38, p-interaction=0.030). Additionally, linear-mixed models indicated larger BP-reduction (Figure) and larger reduction in dyspnea and edema (p-interaction all<0.001) in patients with lower sodium intake receiving spironolactone.Conclusions:Low self-reported sodium in HFpEF is associated with volume status and higher risk of HFH. The more pronounced treatment effect with spironolactone on HF-outcome, BP and volume status in patients with the lowest self-reported sodium intake potentially suggests that low self-reported sodium intake is an indicator of a sodium vulnerable, potentially aldosterone driven, state.

Introduction:Individuals presenting with acute decompensated heart failure (ADHF) have varying response to diuretic therapy and short- and long-term prognosis.Hypothesis:If machine learning can risk stratify patients with ADHF and identify subgroups at risk for diuretic resistance.Methods:Participants with ADHF from the ROSE-AHF and CARRESS-HF clinical trials were included (n=451) and clustered using multivariable finite-mixture models based on diuretic efficiency (fluid output over first 72 hours per total intravenous loop diuretic dose). Differences in diuretic efficiency, in-hospital length of stay, and in-hospital mortality were assessed using linear and logistic regression models. Phenogroups were externally validated in trial (DOSE/ESCAPE, ATHENA-HF) and real-world (GWTG-HF) cohorts.Results:Clustering identified 3 phenogroups. Participants in phenogroup 1 (n=271, 60%) had worse diuretic efficiency [median(IQR) = 11.6(6.6-17.9) mL/mg) compared with phenogroups 2 (n=145, 32%) and 3 (n=35, 8%) [median(IQR) = 16.3(11.2-23.9) and 20.2(12.3-49.9) mL/mg, respectively; p<0.001]. An integer-based risk score to predict phenogroup 1 (lowest diuretic efficiency) was created: BAN-ADHF (Fig.). Net urine output was 2600 vs. 1090mL per 24 hours in patients with scores of 5 and 15, respectively (Fig). In the external validation cohorts, participants with scores ≥11 vs. <11 had significantly lower global well-being, higher natriuretic peptide levels on discharge, longer length of stay, and higher risk of in-hospital mortality.Conclusions:We developed and validated a phenomapping strategy and risk score for individuals with ADHF and differential response to diuretic therapy, which was associated with length of stay and mortality.

Background:The prognostic significance of cardiac magnetic resonance (CMR) based left atrial ejection fraction (LAEF) is not well defined in the ischemic cardiomyopathy (ICM) cohort.We assess the additive effect of LAEF in a model to predict outcomes in patients with ICM.Methods:Patients with ICM, who underwent CMR between April 2001 and March 2019 were retrospectively included.Clinical characteristics and CMR parameters were collected and analyzed and LAEF was calculated in addition to myocardial infarct size (MIS) using late gadolinium enhancement (LGE) and CMR based mitral regurgitant fraction (MRF). The primary clinical endpoint was a composite of all-cause mortality and cardiac transplant. A multiple-variable Cox proportional hazards regression model which included established predictors of outcome was constructed, followed by the addition of LAEF. Four pre-specified interactions of LAEF with left ventricular end systolic volume index (LVESVi), MIS and mitral regurgitant fraction (MRF) were tested at a significance level of 0.05.RESULTS:LA functional data was measured in 718 patients. There were 416 deaths and/or transplants, with a median duration of follow up of 1763 days (4.8 years) The mean LA EF was 36 [2 ,74]. In univariate analysis, lower LAEF and higher LAVI were significant predictors of worse outcome, HR= 0.12, 95% CI [0.06, 0.25], p<0.001; HR 1.008, (1.003-1.014) p = 0.003. After addition to the multivariable model, a normal LAEF was highly predictive of reduced risk, HR 0.24, (0.12, 0.48), p<0.001. In the interaction between LAEF and MIS with MR Fraction, the highest risk was observed in patients with an LAEF < 20% and an MIS of > 30%, with MR Fraction of >35%, HR of 3.2 (1.73-5.93), p= 0.009. The lowest risk was in patients with an LAEF of >50% with an MIS of <15 and MR Fraction of <35%, HR 1.07 (0.81-1.42), p=0.009. Figure 1.CONCLUSION:Reduced LAEF is an important predictor of increased mortality in patients with ICM and additive to MIS and MR Fraction.

Introduction:Machine learning methodology can provide means for the identification of diabetic cardiomyopathy (DbCM), a severe and evolving complication of diabetes that leads to high morbidity and mortality. Phenotypic characterization of patient subgroups may support clinically relevant risk stratification in the population with DbCM.Methods:Among individuals with diabetes from the ARIC study cohort (training, n=953), unsupervised hierarchical clustering was performed with 24 candidate variables incorporating echocardiographic parameters, NT-proBNP, and hs-cTnT. The cluster with highest risk of HF was identified as DbCM. A deep learning (DL) classifier was developed to predict DbCM in the ARIC training cohort and validated in a pooled community-based cohort (ARIC testing plus CHS; n=1,050) and an electronic health record (EHR) cohort (n=3,139).Results:Clustering identified 3 phenogroups. Participants in group 3 (vs. 1 and 2) were more commonly men, had higher levels of creatinine, hs-cTnT, and NT-proBNP, higher LA size and LVMi, and increased prevalence of diastolic dysfunction and hypertension. The 5-year risk of HF was significantly higher in phenogroup 3 and thus identified DbCM (17.8% vs. 2.0% [phenogroup 2] vs. 3.5% [phenogroup 1]) (Figure 1A). The key predictors of DbCM were NTproBNP, LVMi, LA size, and diastolic dysfunction parameters (Figure 1B). The DL classifier demonstrated high model performance in identifying DbCM (AUROC = 0.96, accuracy = 0.93, and precision = 0.75). In the validation cohort (community-based), the DL classifier identified 16% of participants with DbCM with a two-fold higher risk of HF (HR [95% CI], 1.99 [1.47-2.67]; ref = no DbCM). A similar pattern of findings was observed in the EHR cohort (37% with DbCM; DbCM vs. no DbCM: HR [95% CI], 1.58 [1.17-2.12]).Conclusion:Machine learning-based techniques can be used to define and identify DbCM which is associated with higher risk of overt HF.

Introduction:Health systems have had limited success in improving on the suboptimal use of medical therapy for heart failure (HF).Hypothesis:We sought to understand if Guideline-Directed Medical Therapy (GDMT) scoring in combination with EHR-based HF registry data could be utilized to delineate system factors associated with higher or lower GDMT scores.Methods:We performed a cross-sectional analysis of patients with ≥ 2 outpatient encounter for heart failure (HF) in the prior 24 months. Total GDMT score was based on the absence, presence, and dosage of pertinent medications (Fig.1). Multivariable logistic regression analysis of patient, provider, and acute utilization factors associated with GDMT score above the median was performed.Results:Among 6,454 HF patients who had a documented echocardiographic LVEF ≤40 in in the prior 24 months, we observed that having a GDMT score ≥ 4 was associated with having a HF physician encounter (OR:2.02, 95%CI: 1.77-2.32), having two cardiology visits (any sub-specialty) within the prior 12 months (OR: 1.44,95%CI: 1.26-1.65), fewer all-cause (OR: 0.91, 95%CI: 0.87-0.94) and HF-specific hospitalizations (OR:0.64, 95%CI: 0.53-0.75), and higher eGFR (OR: 1.18, 95%CI: 1.12-1.25) (Fig. 2).Conclusions:GDMT scoring that is embedded in an EHR-Based Heart Failure Registry can help to define system-specific system factors associated with, the use of optimal HF medical therapy. Connectivity to cardiology, particularly HF specialists, was independently associated with higher GDMT score in our health system.

Background:Functional mitral regurgitation (FMR) is associated with poor prognosis, however the determinants of FMR progression are not well understood. We aimed to determine clinical and cardiac magnetic resonance imaging (CMR) factors associated with FMR progression in patients with non-ischemic cardiomyopathy (NICM) who underwent baseline and follow-up echocardiography.Methods:NICM patients undergoing CMR between 12/2002-12/2017 with baseline (within 90 days of CMR) and follow-up echocardiography were evaluated. Progressive FMR was assessed by echocardiography based on reported FMR severity (none, mild, moderate, moderate-severe, severe). Associations between clinical and CMR parameters (left ventricular and left atrial (LA) size and function, late gadolinium enhancement, and mitral valve (MV) geometry quantification) and progressive FMR were assessed by univariable and stepwise multivariable linear regression.Results:Amongst 311 NICM patients (age 53±15.7 years, female 121 (38.9%)). A total of 17 patients (5.5%) had at least 1-grade of deterioration in FMR, while 66 patients (21.2%) had at least 1-grade of improvement. Univariable and multivariable analyses results are listed in the table. Mean baseline mitral regurgitant fraction by CMR was 14% ± 13%, while the mean of mitral regurgitation severity by echo was in the mild range (1.28 ±1). Baseline FMR grade by TTE (P<0.001), CMR-LA volume indexed (P=0.003), sphericity ratio (P=0.005), MV annular mean indexed (P=0.027), and BMI (P=0.044) were independently associated with significant change in FMR.Conclusion:CMR-derived remodeling, LA and MV geometry parameters predicted progressive vs regression in FMR. Further studies are needed to determine if CMR predictors of FMR progression vs regression can improve selection criteria for procedural intervention vs guideline-directed medical therapy.

Introduction:Obesity is a major risk factor for HFpEF. Metabolic surgery has proven to be effective in reducing heart failure incidence.Hypothesis:This study examined the impact of metabolic surgery on LV structure and function in patients with high risk for HFpEF.Methods:Patients undergoing metabolic surgery between March 2005 and February 2019 with pre- and post-operative TTE imaging, a preserved EF (>=50%) and a H2FPEF score >=5 were recruited for this study. A H2FPEF score of ≥5 is considered a high possibility of HFpEF diagnosis.Results:Table 1 shows baseline characteristics of the 126 patients that met the entry requirements. The mean age was 65.5 years (SD 8.9) and 64.3% were female. Table 2 outlines the clinical and echocardiographic variables before and after metabolic surgery. There was a significant decrease in mean body mass index (BMI) post operatively, from 47.7 to 36.6 (P<0.001), along with a number of statistically significant improvements in cardiovascular risk factors including a decrease in mean systolic blood pressure (BP) from 138 to 131 mmHg (P=0.005) and mean diastolic BP from 75 to 71 mmHg (P=0.014). There were improvements seen in the LV structure and function with the LV mass decreasing from 247.8 g to 220.7 g (P<0.001), and both the mean septal wall (1.4 cm to 1.3cm, P<0.001) and posterior wall (1.2 cm to 1.1 cm, P=0.007) thicknesses decreasing. There was also an improvement in the average LV global longitudinal strain (GLS) from -15.6 to -17.7% (P<0.001).Conclusions:This study demonstrates that for patients at high risk of HFPEF, metabolic surgery was associated with improvements in LV structure and function, including LVGLS.

Introduction:Cardiopulmonary exercise stress test (CPET) is an essential prognostic tool in patients with heart failure with reduced ejection fraction (HFrEF) for advanced therapy evaluation. We studied the predictive utility of the metabolic gain index (MGI), which reflects the relative gain of the CPET-estimated stroke volume by the relative change in the product of VO2and heart rate (HR) from resting to peak exercise. This novel MGI aims to reflect cardiac reserve function, and, overall prognosis in HFrEF patients.Hypothesis:The MGI is an independent predictor of adverse outcomes in patients with HFrEF.Methods:We reviewed electronic medical records of 843 HFrEF patients (LVEF ≤40%) undergoing CPET from 12/2012 to 9/2020. The MGI was calculated using the formula, [(Peak VO2x Peak HR)-(Resting VO2x Resting HR)]/(Resting VO2x Resting HR). Patients with hypotensive or bradycardia response during exercise were excluded. Primary outcome was the composite of all-cause mortality, LVAD implantation, and heart transplantation. Multivariable Cox proportional hazard models and log rank test were used. For subgroup analysis, patients were stratified by the RER, age, sex, BMI, and beta-blocker use.Results:In our study cohort (mean age 56.2±12.2 years, 73% men, 85% with beta-blockers, mean LVEF 24.8±8.0 %, 16% with BMI ≥35 kg/m2, and 75% with RER >1.05), there were 279 (33.1%) patients developed adverse events during median follow up time of 897 days. Lower MGI was independently associated with increased risk of the adverse outcomes with adjustment for age, sex, comorbidities, LVEF, and peak VO2(quartile 1 vs 4, hazard ratio 1.83, 95% CI 1.03-3.25, p=0.040, Figure 1A), especially in patients with RER > 1.05, BMI < 35 kg/m2, and without beta-blocker use (Figure 1B).Conclusions:Lower MGI is associated with poor prognosis independently of peak VO2. The prognostic value of MGI needs to be further corroborated in other multicenter HFrEF cohorts.

Background:Echocardiography is the primary imaging modality for assessment of patients with non-ischemic cardiomyopathy (NICM), but it is known to underestimate cardiac magnetic resonance (CMR) measurements of LV volumes. We sought to assess the impact of increasing left ventricular end-diastolic volume index (LVEDVi) and left atrial volume index (LAVi) on the accuracy of echocardiographic measurements, when compared to CMR.Methods:NICM patients undergoing CMR between 4/4/2001-12/29/2017 with echocardiography measurements that were obtained within 90 days were retrospectively recorded. Bland Altman analysis and Pearson’s correlation were performed to compare echocardiographic and CMR assessment. Multivariable logistic regression analysis was also performed to predict Echo bias > 50cc/m2 for LVEDVi.Results:Echocardiographic and CMR measurements of LVEDVi and LAVI were obtained in 390 and 400 patients, respectively. Figure 1 illustrates the Bland Altman and Pearson’s correlation of echo vs CMR measurements of LVEDVi and LAVi. While echo and CMR LVEDVi and LAVi were significantly correlated (r=0.762 and 0.605, respectively), there was significant increase in echocardiographic underestimation as the LVEDVi and LAVi as the left ventricle and left atrium increased in size. Multivariable logistic regression analysis is presented in Figure 1. There was a trend toward increased error in women (p=0.076), but age and BME were not significantly associated with increased echo underestimation. Increasing LVEDVi was the strongest predictor of echo underestimation (p<0.0001).Conclusion:Echo underestimation of LV and LA size in patients with NICM increases as the LV and LA size increases. Increased error in echocardiographic measurements may have significant clinical impact in regards to prognosis and procedural therapeutic management strategies.

Introduction:Familial Hypercholesterolemia (FH), an autosomal dominant genetic disorder, confers increased risk for premature cardiovascular disease (CVD). Unfortunately, FH is often only diagnosed after a CVD event. The electronic health record (EHR) holds potential to identify FH patients before an event occurs, but little is known about utilizing algorithms within the EHR to classify at-risk FH patients.Methods:We retrospectively reviewed the EHR of 2,745,615 20-49 year olds seen at the Cleveland Clinic to identify 329 patients with premature CVD and an in-house laboratory LDL ? 190 mg/dL resulted between 2014 and 2019. Patients were identified using laboratory results and ICD-10 codes for premature CVD, physical exam findings, pertinent family history, and genetic variant testing.Results:Using the Dutch Lipid Network Clinic (DLNC) criteria, we identified 21 patients with definite FH (DLNC >8), 102 patients with probable FH (DLNC score 6-8), and 206 patients with possible FH (DLNC 3-5). In all groups, there were disparities in appropriate diagnosis, treatment, and specialist referrals. Of the 21 subjects identified as definite FH, only 5 were actually diagnosed with FH by ICD-10 codes (23.8%), with 15 (71.4%) on high-intensity statin therapy and 4 (19.0%) on PSCK9 inhibitors. In the probable FH subset, 78 (76.4%) were on high-intensity statins yet only 8 (7.8%) had a formal FH diagnosis by ICD codes. Of the total 329 patients, only 24 (7.3%) were referred to medical genetics, 59 (17.9%) to preventative cardiology, and 191 (58.0%) were referred to any cardiology specialist.Conclusions:We developed an algorithm to utilize EHR data to identify potentially undiagnosed cases of FH and facilitate a referral method for appropriate CVD risk reduction and cascade testing. Through EHR data advancements, there is opportunity for more proactive classification and treatment of FH patients, and a way to improve current disparities.

Introduction:Increased synthesis of pro-fibrotic proteins is a common feature of cardiac fibrosis. However, critical factors and mechanisms for translational control of cardiac fibrosis remain unclear. In mammals, glutamyl-prolyl-tRNA synthetase (EPRS) catalyzes the attachment of glutamic acid (E) and proline (P) to their tRNAs for protein synthesis.Hypothesis:Cardiac stress-induced EPRS promotes cardiac fibrosis via increased Pro-tRNAProand enhanced stabilization and translation of proline codon rich (PRR) pro-fibrotic mRNAs in cardiac fibroblasts.Methods and Results:By overlapping of aminoacyl-tRNA synthetases (ARSs) induced in TGF-? treated human cardiac fibroblasts, human ARSs with genetic mutations in the congenital heart disease, mouse ARSs associated with isoproterenol (ISO)-induced cardiomyopathy by GWAS, and ISO-induced ARSs in mouse failing hearts, we identified EPRS as a major ARS involved in cardiac pathogenesis. We found that EPRS is induced in failing human and mouse hearts compared to non-failing hearts. EPRS functions as an integrated node downstream of multiple hypertrophic and fibrotic stimuli. Low-dose halofuginone (Halo), an (E)PRS-specific inhibitor, and genetic knockout of one allele of EPRS in the mouse genome, reduces cardiac hypertrophy and fibrosis in mouse heart failure models. Using RNA-seq and polysome profiling-seq in Halo-treated mouse fibroblasts overlapped with genome-wide PRR gene search and available quantitative proteomic data, we have identified novel PRR genes, such as Sulf1, Furin, Fn1, and Ltbp1/2 besides collagens, which are translationally regulated by EPRS. Inactivation of SULF1 and FURIN by RNAi or inhibitors attenuates cardiac fibroblast activation in vitro. We also found that inhibition of EPRS enzymatic activity reduced translational efficiency and enhanced decay of PRR mRNAs that is dependent on elongation factor 5 and mRNA decapping enzyme 2.Conclusions:Our results indicate that EPRS preferentially controls the translational activation of Pro-rich pro-fibrotic genes in cardiac fibroblasts and augments pathological cardiac remodeling. The integrated omics-based screen from our studies has identified multiple novel anti-fibrosis pharmacological target genes.

Introduction:Angiotensin converting enzyme inhibitors (ACEI) decrease major atherosclerotic events in coronary heart disease with reduced ejection fraction (EF<40%) but not in coronary artery disease with preserved ejection fraction (EF>50%) on standard lipid lowering therapy. Recently, mid-range ejection fraction (mrEF, EF 40-49%) has emerged as a separate entity in patients with heart failure. We investigate the role of ACEI in stable coronary heart disease patients with mrEF using the Prevention of events with angiotensin converting enzyme inhibition (PEACE) dataset.Hypothesis:ACEI use will not decrease major atherosclerotic events in stable coronary disease patients with mrEF on standard medical therapy.Methods:Patients with mrEF were selected from the PEACE dataset using EF cutoffs 40-49%. Patients were followed for minimum of 4 years and maximum of 7 years. Cardiovascular outcomes were compared between groups (Trandolapril use) using KM curves and log-rank test. P<0.05 was considered significant. R was used for statistical analysis.Results:In this post-hoc analysis, among 1,236 patients [age (64+/-8 years), 88% males], there was no statistically significant difference among the two groups in terms of MACE (figure 1).Conclusions:Addition of ACEI may not impart additional protective effect to standard medical therapy in patients with stable coronary disease and mrEF.Figure 1: Cardiovascular outcomes comparison among the study population

Introduction:The growing awareness of cardiac dysfunction by cancer treatment has led to the emerging field of cardio-oncology. This is a pressing need to improve early prevention of cardiac dysfunction resulted from cancer treatment.Hypothesis:Could we stratify cardiac risk subgroups from clinical variables-derived patient-patient networks?Methods:This study performs an unbiased network analysis of 5,194 cancer patients with both cardiovascular imaging data (10 echocardiogram [Echo] variables) and 53 other clinical variables from the Cleveland Clinic Epic database. We built topology-based patient-patient networks by quantifying Pearson correlation coefficient (PCC) of clinical variables across individuals: