Assessing the use of GEE methods for analyzing binary outcomes in family studies: the Strong Heart Family Study.

The generalized estimating equations method (GEE) is commonly applied to analyze data obtained from family studies. GEE is well known for its robustness on misspecification of correlation structure. However, the unbalanced distribution of family sizes and complicated genetic relatedness structure within each family may challenge GEE performance. We focused our research on binary outcomes. To evaluate the performance of GEE, we conducted a series of simulations, on data generated adopting the kinship matrix (correlation structure within each family) from the Strong Heart Family Study (SHFS). We performed a fivefold cross-validation to further evaluate the GEE predictive power on data from the SHFS. A Bayesian modeling approach, with direct integration of the kinship matrix, was also included to contrast with GEE. Our simulation studies revealed that GEE performs well on a binary outcome from families having a relatively simple kinship structure. However, data with a binary outcome generated from families with complex kinship structures, especially with a large genetic variance, can challenge the performance of GEE.

Biomarkers and cardiovascular outcomes in chimeric antigen receptor T-cell therapy recipients.

AIMS: Chimeric antigen receptor T-cell therapy (CAR-T) harnesses a patient's immune system to target cancer. There are sparse existing data characterizing death outcomes after CAR-T-related cardiotoxicity. This study examines the association between CAR-T-related severe cardiovascular events (SCE) and mortality. METHODS AND RESULTS: From a multi-centre registry of 202 patients receiving anti-CD19 CAR-T, covariates including standard baseline cardiovascular and cancer parameters and biomarkers were collected. Severe cardiovascular events were defined as a composite of heart failure, cardiogenic shock, or myocardial infarction. Thirty-three patients experienced SCE, and 108 patients died during a median follow-up of 297 (interquartile range 104-647) days. Those that did and did not die after CAR-T were similar in age, sex, and prior anthracycline use. Those who died had higher peak interleukin (IL)-6 and ferritin levels after CAR-T infusion, and those who experienced SCE had higher peak IL-6, C-reactive protein (CRP), ferritin, and troponin levels. The day-100 and 1-year Kaplan-Meier overall mortality estimates were 18% and 43%, respectively, while the non-relapse mortality (NRM) cumulative incidence rates were 3.5% and 6.7%, respectively. In a Cox model, SCE occurrence following CAR-T was independently associated with increased overall mortality risk [hazard ratio (HR) 2.8, 95% confidence interval (CI) 1.6-4.7] after adjusting for age, cancer type and burden, anthracycline use, cytokine release syndrome grade ≥ 2, pre-existing heart failure, hypertension, and African American ancestry; SCEs were independently associated with increased NRM (HR 3.5, 95% CI 1.4-8.8) after adjusting for cancer burden. CONCLUSION: Chimeric antigen receptor T-cell therapy recipients who experience SCE have higher overall mortality and NRM and higher peak levels of IL-6, CRP, ferritin, and troponin.

Postoperative pericardial effusion, pericardiotomy, and atrial fibrillation: An explanatory analysis of the PALACS trial.

BACKGROUND: In the Posterior left pericardiotomy for the prevention of atrial fibrillation after cardiac surgery (PALACS) trial, posterior pericardiotomy was associated with a significant reduction in postoperative atrial fibrillation (POAF) after cardiac surgery. We aimed to investigate the mechanisms underlying this effect. METHODS: We included PALACS patients with available echocardiographic data (n = 387/420, 92%). We tested the hypotheses that the reduction in POAF with the intervention was associated with 1) a reduction in postoperative pericardial effusion and/or 2) an effect on left atrial size and function. Spline and multivariable logistic regression analyses were used. RESULTS: Most patients (n = 307, 79%) had postoperative pericardial effusions (anterior 68%, postero-lateral 51.9%). The incidence of postero-lateral effusion was significantly lower in patients undergoing pericardiotomy (37% vs 67%; P < .001). The median size of anterior effusion was comparable between patients with and without POAF (5.0 [IQR 3.0-7.0] vs 5.0 [IQR 3.0-7.5] mm; P = .42), but there was a nonsignificant trend towards larger postero-lateral effusion in the POAF group (5.0 [IQR 3.0-9.0] vs 4.0 [IQR 3.0-6.4] mm; P = .06). There was a non-linear association between postero-lateral effusion and POAF at a cut-off at 10 mm (OR 2.70; 95% CI 1.13, 6.47; P = .03) that was confirmed in multivariable analysis (OR 3.5, 95% CI 1.17, 10.58; P = 0.02). Left atrial dimension and function did not change significantly after posterior pericardiotomy. CONCLUSIONS: Reduction in postero-lateral pericardial effusion is a plausible mechanism for the effect of posterior pericardiotomy in reducing POAF. Measures to reduce postoperative pericardial effusion are a promising approach to prevent POAF.

Baseline Intraoperative Left Ventricular Diastolic Function Is Associated with Postoperative Atrial Fibrillation after Cardiac Surgery.

BACKGROUND: Detailed understanding of the association between intraoperative left atrial and left ventricular diastolic function and postoperative atrial fibrillation is lacking. In this post hoc analysis of the Posterior Left Pericardiotomy for the Prevention of Atrial Fibrillation after Cardiac Surgery (PALACS) trial, we aimed to evaluate the association of intraoperative left atrial and left ventricular diastolic function as assessed by transesophageal echocardiography (TEE) with postoperative atrial fibrillation. METHODS: PALACS patients with available intraoperative TEE data (n = 402 of 420; 95.7%) were included in this cohort study. We tested the hypotheses that preoperative left atrial size and function, left ventricular diastolic function, and their intraoperative changes were associated with postoperative atrial fibrillation. Normal left ventricular diastolic function was graded as 0 and with lateral e' velocity 10 cm/s or greater. Diastolic dysfunction was defined as lateral e' less than 10 cm/s using E/e' cutoffs of grade 1, E/e' 8 or less; grade, 2 E/e' 9 to 12; and grade 3, E/e' 13 or greater, along with two criteria based on mitral inflow and pulmonary wave flow velocities. RESULTS: A total of 230 of 402 patients (57.2%) had intraoperative diastolic dysfunction. Posterior pericardiotomy intervention was not significantly different between the two groups. A total of 99 of 402 patients (24.6%) developed postoperative atrial fibrillation. Patients who developed postoperative atrial fibrillation more frequently had abnormal left ventricular diastolic function compared to patients who did not develop postoperative atrial fibrillation (75.0% [n = 161 of 303] vs. 57.5% [n = 69 of 99]; P = 0.004). Of the left atrial size and function parameters, only delta left atrial area, defined as presternotomy minus post-chest closure measurement, was significantly different in the no postoperative atrial fibrillation versus postoperative atrial fibrillation groups on univariate analysis (-2.1 cm2 [interquartile range, -5.1 to 1.0] vs. 0.1 [interquartile range, -4.0 to 4.8]; P = 0.028). At multivariable analysis, baseline abnormal left ventricular diastolic function (odds ratio, 2.02; 95% CI, 1.15 to 3.63; P = 0.016) and pericardiotomy intervention (odds ratio, 0.46; 95% CI, 0.27 to 0.78, P = 0.004) were the only covariates independently associated with postoperative atrial fibrillation. CONCLUSIONS: Baseline preoperative left ventricular diastolic dysfunction on TEE, not left atrial size or function, is independently associated with postoperative atrial fibrillation. Further studies are needed to test if interventions aimed at optimizing intraoperative left ventricular diastolic function during cardiac surgery may reduce the risk of postoperative atrial fibrillation.

Atrial fibrillation onset before heart failure or vice versa: what is worst? A nationwide register study.

AIMS: Atrial fibrillation (AF) and heart failure (HF) often coexist. However, whether AF onset before HF or vice versa is associated with the worst outcome remains unclear. A consensus of large studies can guide future research and preventive strategies to better target high-risk patients. METHODS AND RESULTS: We included all Danish cases with the coexistence of AF and HF (2005-17) using nationwide registries. Patients were divided into three separate groups (i) AF before HF, (ii) HF before AF, or (iii) AF and HF diagnosed concurrently (±30 days). Adjusting landmark Cox analyses (index date was the time of the latter diagnosis of AF or HF) were used for evaluating the association of the three groups with a composite outcome of ischaemic stroke or death. Among a total of 49 042 patients included, 40% had AF before HF, 27% had HF before AF, and 33% had AF and HF diagnosed concurrently. The composite endpoint accrued more often in patients with HF before AF compared to the two other groups (<0.001), and this remained significant in the adjusted analyses with hazard ratios (95% confidence intervals) of 1.26 (1.22-1.30) compared to AF before HF. Finally, antihypertensive treatment, oral anticoagulants, amiodarone, statins, and AF ablation were associated with a lower hazard ratio of the composite endpoint (all < 0.001). CONCLUSIONS: In this large Danish national cohort, diagnosis of HF before AF was associated with an increased absolute risk of death compared to AF before HF and AF and HF diagnosed concurrently. Antihypertensive treatment, oral anticoagulants, amiodarone, statins, and AF ablation may improve prognosis.

Dietary lead modulates the mouse intestinal microbiome: Subacute exposure to lead acetate and lead contaminated soil.

The effect of dietary lead on the intestinal microbiome has not been fully elucidated. To determine if there was an association between microflora modulation, predicted functional genes, and Pb exposure, mice were provided diets amended with increasing concentrations of a single lead compound, lead acetate, or a well characterized complex reference soil containing lead, i.e. 6.25-25 mg/kg Pb acetate (PbOAc) or 7.5-30 mg/kg Pb in reference soil SRM 2710a having 0.552 % Pb among other heavy metals such as Cd. Feces and ceca were collected following 9 days of treatment and the microbiome analyzed by 16 S rRNA gene sequencing. Treatment effects on the microbiome were observed in both feces and ceca of mice. Changes in the cecal microbiomes of mice fed Pb as Pb acetate or as a constituent in SRM 2710a were statistically different except for a few exceptions regardless of dietary source. This was accompanied by increased average abundance of functional genes associated with metal resistance, including those related to siderophore synthesis and arsenic and/or mercury detoxification. Akkermansia, a common gut bacterium, was the highest ranked species in control microbiomes whereas Lactobacillus ranked highest in treated mice. Firmicutes/Bacteroidetes ratios in the ceca of SRM 2710a treated mice increased more than with PbOAc, suggestive of changes in gut microbiome metabolism that promotes obesity. Predicted functional gene average abundance related to carbohydrate, lipid, and/or fatty acid biosynthesis and degradation were greater in the cecal microbiome of SRM 2710a treated mice. Bacilli/Clostridia increased in the ceca of PbOAc treated mice and may be indicative of increased risk of host sepsis. Family Deferribacteraceae also was modulated by PbOAc or SRM 2710a possibly impacting inflammatory response. Understanding the relationship between microbiome composition, predicted functional genes, and Pb concentration, especially in soil, may provide new insights into the utility of various remediation methodologies that minimize dysbiosis and modulate health effects, thus assisting in the selection of an optimal treatment for contaminated sites.

Aortic Root Dilatation in Hypertensive Patients with Left Ventricular Hypertrophy-Application of A New Multivariate Predictive Model. The Life Study.

BACKGROUND: Available nomograms to predict aortic root (AoR) diameter for body surface area have limitations. The purpose of this study was to evaluate the use of a new multivariate predictive model to identify AoR dilatation in hypertensive patients with left ventricular hypertrophy. METHODS: 943 of 961 patients in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) echocardiographic sub-study had the necessary baseline characteristics and echocardiographic 2D measurements of AoR size to be included. RESULTS: Predicted AoR (Sinus of Valsalva) diameter was 1.519 + (age [years] × 0.010) + (height [cm] × 0.010) - (gender [1 = M, 2 = F] × 0.247), and a measured AoR diameter exceeding the 97.5-percentile of this estimate was considered dilated. Measured AoR diameter was larger in men than in women (3.75 vs. 3.48 cm, p < 0.001) and AoR diameter predicted by the model was larger than predicted by nomogram (3.52 vs. 3.28 cm, p < 0.001). Using the multivariate model to identify patients with AoR dilatation, the prevalence was 13.7% in men and 12.3% in women (p = 0.537). There was consensus of AoR phenotype (normal/dilated) between model and nomogram in 92.8% of the patients. In multivariate logistic regression, AoR dilatation by model definition was predicted by presence of aortic regurgitation (OR 2.67, p < 0.001) and SD increase in age (OR 0.75, p = 0.023), pulse pressure (OR 0.64, p < 0.001), left ventricular mass index (OR 1.36, p = 0.08) and stroke volume (OR 1.45, p = 0.002), but not by body weight. CONCLUSIONS: Using the proposed model the prevalence of AoR dilatation was equal in men and women and the model seems to address the effects of gender, age and body size on AoR size. CLINICAL TRIAL REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT00338260.

Subclinical atherosclerosis in adolescents and young adults and the risk of cardiovascular disease: The Strong Heart Family Study (SHFS).

BACKGROUND AND AIMS: Rates of cardiovascular disease (CVD) among American Indians (AI) have been increasing. Although we have observed an association between atherosclerosis and CVD in older adults, the potential association among young AI is unclear. Therefore, we aim to describe the prevalence of atherosclerosis among young AI and determine its association with CVD and all-cause mortality. METHODS AND RESULTS: We evaluated AI participants from the Strong Heart Family Study (SHFS), who were <40 years old and CVD free at the baseline examination, 2001-2003 (n = 1376). We used carotid ultrasound to detect baseline atherosclerotic plaque. We identified CVD events and all-cause mortality through 2019, with a median follow-up of 17.8 years. We used shared frailty Cox Proportional Hazards models to assess the association between atherosclerosis and time to CVD event or all-cause mortality, while controlling for covariates. Among 1376 participants, 71 (5.2%) had atherosclerosis at baseline. During follow-up, 120 (8.7%) had CVD events and 104 (7.6%) died from any cause. CVD incidence was higher in participants who had baseline atherosclerosis (13.51/1000 person-years) than in those who did not (4.95/1000 person-years, p = 0.0003). CVD risk and all-cause mortality were higher in participants with atherosclerosis, while controlling for covariates (CVD HR = 1.85, 95%CI = 1.02-3.37, p = 0.0420; all-cause mortality HR = 2.04, 95%CI = 1.07-3.89, p = 0.0291). CONCLUSIONS: Among young AI, atherosclerosis was independently associated with incident CVD and all-cause mortality later in life. Thus, atherosclerosis begins early in life and interventions in adolescents and young adults to slow the progression of disease could prevent or delay CVD events later in life.

Relative Impact of Surgical Mitral Repair and MitraClip on Annular Remodeling-A Potential Mechanism for Therapeutic Response to Mitral Repair for Degenerative Mitral Regurgitation.

OBJECTIVES: Three-dimensional transesophageal echocardiography (TEE) is widely used to guide decision-making for mitral repair. The relative impact of surgical mitral valve repair (MVr) and MitraClip on annular remodeling is unknown. The aim was to determine the impact of both mitral repair strategies on annular geometry, including the primary outcome of annular circumference and area. DESIGN: This was a retrospective observational study of patients who underwent mitral intervention between 2016 and 2020. SETTING: Weill Cornell Medicine, a single, large, academic medical center. PARTICIPANTS: The population comprised 50 patients with degenerative mitral regurgitation (MR) undergoing MVr. INTERVENTIONS: Elective MVr and TEE. MEASUREMENTS AND MAIN RESULTS: Patients undergoing MitraClip or surgical MVr were matched (1:1) for sex and coronary artery disease. Mitral annular geometry indices were quantified on intraprocedural three-dimensional TEE. Mild or less MR on follow-up transthoracic echocardiography defined optimal response. Patients undergoing MitraClip were older (80 ± eight v 66 ± six years; p < 0.001) but were otherwise similar to surgical patients. Patients undergoing MitraClip had larger baseline left atrial and ventricular sizes, increased tenting height, and volume (p < 0.01), with a trend toward increased annular area (p = 0.23). MitraClip and surgery both induced immediate mitral annular remodeling, including decreased area, circumference, and tenting height (p < 0.001), with greater remodeling with surgical repair. At follow-up (4.1 ± 9.0 months) optimal response (≤ mild MR) was ∼twofold more common with surgery than MitraClip (81% v 46%; p = 0.02). The relative reduction in annular circumference (odds ratio [OR] 1.05 [1.00-1.09] per cm; p = 0.04) and area (OR 1.03 [1.00-1.05] per cm2; p = 0.049) were both associated with optimal response. CONCLUSIONS: Surgical MVr and MitraClip both reduce annular size, but repair-induced remodeling is greater with surgery and associated with an increased likelihood of optimal response.

Left Atrial Strain Quantification by Intraoperative Transesophageal Echocardiography: Validation With Transthoracic Echocardiography.

OBJECTIVE: Whereas left atrial (LA) strain has been well-validated using transthoracic echocardiography (TTE), its detection using transesophageal echocardiography (TEE) has not been studied. Conventional transesophageal views are known to be limited due to the posterior location of the LA. Here, the feasibility and accuracy of the deep transgastric long-axis LA focused view for peak atrial longitudinal strain (PALS) quantification was tested. DESIGN: This was a retrospective study of patients who underwent elective cardiac surgery between 2018 and 2020. TEE deep transgastric long-axis view was compared to TTE 4-chamber atrial focused view as the reference standard. LA area, volume, and PALS were quantified independently. SETTING: At Weill Cornell Medicine, a single, large academic medical center. PARTICIPANTS: The population comprised 42 patients undergoing cardiac surgery who had a TTE and TEE within 14.9 ± 20.8 days. INTERVENTIONS: TTE, TEE, and cardiac surgery. MEASUREMENTS AND MAIN RESULTS: TEE-derived PALS strongly correlated with TTE- derived PALS (r = 0.92, p < 0.001), though absolute PALS were lower (20.7 ± 6.0% v 25.7 ± 6.8%; p < 0.001). Mean TEE-derived atrial length was similar to TTE-derived length (5.18 ± 0.61 cm v 5.24 ± 0.61 cm; p = 0.38), but mean LA area was significantly smaller (16.7 ± 3.5 cm2v 18.9 ± 3.7 cm2; p < 0.001), with significant correlations between the 2 modalities for both (r = 0.74, 0.74, respectively; all p < 0.001). CONCLUSION: This exploratory study supported the feasibility of TEE for assessing LA longitudinal strain. There was an excellent correlation between atrial strain derived via TEE versus TTE, although values tended to be smaller on TEE, and bias between values was highly variable, suggesting that the values were not interchangeable.

Human Immunodeficiency Virus-Associated Myocardial Diastolic Dysfunction and Soluble ST2 Concentration in Tanzanian Adults: A Cross-Sectional Study.

BACKGROUND: The aims of this study were (1) to compare the prevalence of myocardial diastolic dysfunction (DD) in antiretroviral therapy (ART)-naive people living with human immunodeficiency virus (PLWH) to human immunodeficiency virus (HIV)-uninfected adults in East Africa and (2) to determine the association between serum concentration of the cardiac biomarkers ST2 and DD. METHODS: In this cross-sectional study, we enrolled PLWH and uninfected adults at a referral HIV clinic in Mwanza, Tanzania. Standardized history, echocardiography, and serum were obtained. Regression models were used to quantify associations. RESULTS: We enrolled 388 ART-naive PLWH and 461 HIV-uninfected adults with an average age of 36.0 ± 10.2 years. Of PLWH in the third, fourth, and fifth decades of life, 5.0%, 12.5%, and 32.7%, respectively, had DD. PLWH had a higher prevalence of DD (adjusted odds ratio, 2.71 [95% confidence interval, 1.62-4.55]; P < .0001). PLWH also had a higher probability of dysfunction with one or fewer traditional risk factors present. Serum ST2 concentration was associated with dysfunction in PLWH but not uninfected participants (P = .04 and P = .90, respectively). CONCLUSIONS: In a large population of young adults in sub-Saharan Africa, DD prevalence increased starting in the third decade of life. HIV was independently associated with dysfunction. Serum ST2 concentration was associated with DD in PLWH but not HIV-uninfected participants. This pathway may provide insight into the mechanisms of HIV-associated dysfunction.

Comparing Vascular Brain Injury and Stroke by Cranial Magnetic Resonance Imaging, Physician-Adjudication, and Self-Report: Data from the Strong Heart Study.

BACKGROUND: Epidemiologic studies often use self-report as proxy for clinical history. However, whether self-report correctly identifies prevalence in minority populations with health disparities and poor health-care access is unknown. Furthermore, overlap of clinical vascular events with covert vascular brain injury (VBI), detected by imaging, is largely unexamined. METHODS: The Strong Heart Study recruited American Indians from 3 regions, with surveillance and adjudication of stroke events from 1989 to 2013. In 2010-2013, all 817 survivors, aged 65-95 years, underwent brain imaging, neurological history interview, and cognitive testing. VBI was defined as imaged infarct or hemorrhage. RESULTS: Adjudicated stroke was prevalent in 4% of participants and separately collected, self-reported stroke in 8%. Imaging-defined VBI was detected in 51% and not associated with any stroke event in 47%. Compared with adjudication, self-report had 76% sensitivity and 95% specificity. Participants with adjudicated or self-reported stroke had the poorest performance on cognitive testing; those with imaging-only (covert) VBI had intermediate performance. CONCLUSION: In this community-based cohort, self-report for prior stroke had good performance metrics. A majority of participants with VBI did not have overt, clinically recognized events but did have neurological or cognitive symptoms. Data collection methodology for studies in a resource-limited setting must balance practical limitations in costs, accuracy, feasibility, and research goals.

Seasonal Distribution of Cyanobacteria in Three Urban Eutrophic Lakes Results from an Epidemic-like Response to Environmental Conditions.

Cyanobacterial communities of three co-located eutrophic sandpit lakes were surveyed during 2016 and 2017 over season and depth using high-throughput DNA sequencing of the 16S rRNA gene. All three lakes were stratified except during April 2017 when the lakes were recovering from a strong mixing event. 16S rRNA gene V4 sequences were parsed into operational taxonomic units (OTUs) at 99% sequence identity. After rarefaction of 139 samples to 25,000 sequences per sample, a combined total of 921,529 partial 16S rRNA gene sequences were identified as cyanobacteria. They were binned into 19,588 unique cyanobacterial OTUs. Of these OTUs, 11,303 were Cyanobium. Filamentous Planktothrix contributed 1537 and colonial Microcystis contributed 265. The remaining 6482 OTUs were considered unclassified. For Planktothrix and Microcystis one OTU accounted for greater than 95% of the total sequences for each genus. However, in both cases the non-dominant OTUs clustered with the dominant OTUs by date, lake, and depth. All Planktothrix OTUs and a single Cyanobium OTU were detected below the oxycline. All other Cyanobium and Microcystis OTUs were detected above the oxycline. The distribution of Cyanobium OTUs between lakes and seasons can be explained by an epidemic-like response where individual OTUs clonally rise from a diverse hypolimnion population when conditions are appropriate. The importance of using 99% identity over the more commonly used 97% is discussed with respect to cyanobacterial community structure. The approach described here can provide another valuable tool for assessing cyanobacterial populations and provide greater insight into the controls of cyanobacterial blooms.

Differential Effects of Aortic Valve Replacement on Aortic Circumferential Strain in Aortic Stenosis and Aortic Insufficiency.

OBJECTIVES: Aortic valve replacement (AVR) potentially can modify pulse-wave propagation to the distal aorta. Echo-derived global circumferential strain (GCS) was used to test whether AVR for aortic stenosis (AS) or aortic insufficiency (AI) resulted in differential aortic biomechanics in the descending thoracic aorta. DESIGN: This was a prospective observational study of patients who underwent cardiac surgery between 2016 and 2019. SETTING: Weill Cornell Medicine, a single large academic medical center. PARTICIPANTS: The population comprised 48 patients undergoing AVR (62 ± 15 y/o, 79% male; 22 with AI and 26 with AS) and 11 patients undergoing coronary bypass surgery as controls. INTERVENTIONS: Elective cardiac surgery, transesophageal echocardiography (TEE), pulmonary artery catheter. MEASUREMENTS AND MAIN RESULTS: Pre- and postprocedural TEEs were collected. Descending aorta short-axis images were analyzed for GCS, time-to-peak strain, aortic end-diastolic, end-systolic area, and fractional area changes. Pulse pressure (PP) and stroke volume were quantified. Preprocedural GCS significantly differed between patients with AI and AS, with AI patients having greater GCS (median/interquartile range, 9.6 95.3,13.6) than patients with AS (4.3 [3.4-5.1]). After AVR, in AI patients, strain significantly decreased (5.5 [3.8,8.2], p = 0.001), along with PP (mean ± standard deviation) (66.4 ± 0.8 to 54.1 ± 13.7, p < 0.001), and PP corrected strain did not (GCS/PP = 14.8 [6.9-19.9] v 12.7[8.2-18.6], p = 0.34). In AS patients, GCS significantly increased after AVR to (5.45 [4.2-6.8], p = 0.003), as did PP-corrected strain (6.9 [5.8-9.2] v 9.7 [6.5-13.4], p = 0.016). Surgical AVR produced decrements in time-to-peak strain in AI and AS groups (both p < 0.001). CONCLUSIONS: After AVR for AI and AS, the direction of change in distal aortic strain from baseline depends on valve pathology. This finding may have important clinical implications in terms of indication for surgery and postoperative surveillance, especially in patients with aortopathies.

Machine Learning Prediction of Stroke Mechanism in Embolic Strokes of Undetermined Source.

BACKGROUND AND PURPOSE: One-fifth of ischemic strokes are embolic strokes of undetermined source (ESUS). Their theoretical causes can be classified as cardioembolic versus noncardioembolic. This distinction has important implications, but the categories' proportions are unknown. METHODS: Using data from the Cornell Acute Stroke Academic Registry, we trained a machine-learning algorithm to distinguish cardioembolic versus non-cardioembolic strokes, then applied the algorithm to ESUS cases to determine the predicted proportion with an occult cardioembolic source. A panel of neurologists adjudicated stroke etiologies using standard criteria. We trained a machine learning classifier using data on demographics, comorbidities, vitals, laboratory results, and echocardiograms. An ensemble predictive method including L1 regularization, gradient-boosted decision tree ensemble (XGBoost), random forests, and multivariate adaptive splines was used. Random search and cross-validation were used to tune hyperparameters. Model performance was assessed using cross-validation among cases of known etiology. We applied the final algorithm to an independent set of ESUS cases to determine the predicted mechanism (cardioembolic or not). To assess our classifier's validity, we correlated the predicted probability of a cardioembolic source with the eventual post-ESUS diagnosis of atrial fibrillation. RESULTS: Among 1083 strokes with known etiologies, our classifier distinguished cardioembolic versus noncardioembolic cases with excellent accuracy (area under the curve, 0.85). Applied to 580 ESUS cases, the classifier predicted that 44% (95% credibility interval, 39%-49%) resulted from cardiac embolism. Individual ESUS patients' predicted likelihood of cardiac embolism was associated with eventual atrial fibrillation detection (OR per 10% increase, 1.27 [95% CI, 1.03-1.57]; c-statistic, 0.68 [95% CI, 0.58-0.78]). ESUS patients with high predicted probability of cardiac embolism were older and had more coronary and peripheral vascular disease, lower ejection fractions, larger left atria, lower blood pressures, and higher creatinine levels. CONCLUSIONS: A machine learning estimator that distinguished known cardioembolic versus noncardioembolic strokes indirectly estimated that 44% of ESUS cases were cardioembolic.

Prognostic role of diastolic dysfunction in patients undergoing transcatheter aortic valve replacement.

OBJECTIVES: Prior studies have shown that left ventricular diastolic dysfunction (DD) is associated with increased mortality after surgical aortic valve replacement but studies on transcatheter aortic valve replacement (TAVR) are limited and have not taken into account mitral annular calcification (MAC), which limits the use of mitral valve annular tissue Doppler imaging. We performed a single-center retrospective analysis to better evaluate the role of baseline DD on outcomes after TAVR. METHODS: After excluding patients with atrial fibrillation, mitral valve prostheses and significant mitral stenosis, 359 consecutive TAVR patients were included in the study. Moderate-to-severe MAC was present in 58% of the patients. We classified patients into severe versus nonsevere DD based on the evaluation of elevated left ventricular filling pressure. The outcome measure was all-cause mortality or heart failure hospitalization. RESULTS: Over a mean follow-up time of 13 months, severe DD was associated with an increased risk for the outcome measure (HR 2.02 [1.23-3.30], p = .005). However, this association was lost in a propensity-matched cohort. In multivariate analysis, STS score was the only independent predictor of all cause mortality of heart failure hospitalization (HR 1.1 [1.05-1.15], p < .001). CONCLUSIONS: We evaluated the role of baseline DD on outcomes after TAVR by taking into account the presence of MAC. Severe DD was associated with increased all-cause mortality or heart failure hospitalization but not independently of other structural parameters and known predictors of the outcome measure.

On-treatment HDL cholesterol predicts incident atrial fibrillation in hypertensive patients with left ventricular hypertrophy.

Purpose: Hypertensive patients are at increased risk of atrial fibrillation (AF). Although low baseline high density lipoprotein (HDL) cholesterol has been associated with a higher risk of AF, this has not been verified in recent population-based studies. Whether changing levels of HDL over time are more strongly related to the risk of new AF in hypertensive patients has not been examined.Material and methods: Incident AF was examined in relation to baseline and on-treatment HDL levels in 8267 hypertensive patients with no history of AF, in sinus rhythm on their baseline electrocardiogram, randomly assigned to losartan- or atenolol-based treatment. HDL levels at baseline and each year of testing were categorised into quartiles according to baseline HDL levels.Results: During 4.7 ± 1.10 years of follow-up, 645 patients (7.8%) developed new AF. In univariate Cox analyses, compared with the highest quartile of HDL levels (>1.78 mmol/l), patients with on-treatment HDL in the lowest quartile (≤ 1.21 mmol/l) had a 53% greater risk of new AF. Patients with on-treatment HDL in the second and third quartiles had intermediate increased risks of AF. Baseline HDL in the lowest quartile was not a significant predictor of new AF (hazard ratio (HR): 1.14, 95% confidence interval (CI): 0.90-1.43). In multivariable Cox analyses adjusting for multiple baseline and time-varying covariates, the lowest quartile of on-treatment HDL remained associated with a nearly 54% increased risk of new AF (HR: 1.54, 95% CI: 1.16-2.05) whereas a baseline HDL≤ ⩽1.21 mmol/l was not predictive of new AF (HR: 1.01, 95% CI: 0.78-1.31).Conclusion: Lower on-treatment HDL is strongly associated with risk of new AF. These findings suggest that serial assessment of HDL can estimate AF risk better than baseline HDL in hypertensive patients with left ventricular hypertrophy. Future studies may investigate whether therapies that increase HDL can lower risk of developing AF.Clinical Trials Registration: http://clinicaltrials.gov/ct/show/NCT00338260?order=1.

Development of novel machine learning model for right ventricular quantification on echocardiography-A multimodality validation study.

PURPOSE: Echocardiography (echo) is widely used for right ventricular (RV) assessment. Current techniques for RV evaluation require additional imaging and manual analysis; machine learning (ML) approaches have the potential to provide efficient, fully automated quantification of RV function. METHODS: An automated ML model was developed to track the tricuspid annulus on echo using a convolutional neural network approach. The model was trained using 7791 image frames, and automated linear and circumferential indices quantifying annular displacement were generated. Automated indices were compared to an independent reference of cardiac magnetic resonance (CMR) defined RV dysfunction (RVEF < 50%). RESULTS: A total of 101 patients prospectively underwent echo and CMR: Fully automated annular tracking was uniformly successful; analyses entailed minimal processing time (<1 second for all) and no user editing. Findings demonstrate all automated annular shortening indices to be lower among patients with CMR-quantified RV dysfunction (all P < .001). Magnitude of ML annular displacement decreased stepwise in relation to population-based tertiles of TAPSE, with similar results when ML analyses were localized to the septal or lateral annulus (all P ≤ .001). Automated segmentation techniques provided good diagnostic performance (AUC 0.69-0.73) in relation to CMR reference and compared to conventional RV indices (TAPSE and S') with high negative predictive value (NPV 84%-87% vs 83%-88%). Reproducibility was higher for ML algorithm as compared to manual segmentation with zero inter- and intra-observer variability and ICC 1.0 (manual ICC: 0.87-0.91). CONCLUSIONS: This study provides an initial validation of a deep learning system for RV assessment using automated tracking of the tricuspid annulus.

Descending aortic strain quantification by intra-operative transesophageal echocardiography: Multimodality validation via cardiovascular magnetic resonance.

BACKGROUND: Whereas cardiac magnetic resonance (CMR) imaging provides high temporal resolution imaging of aortic distensibility (strain), transesophageal echocardiography (TEE) is widely used for intra-operative aortic imaging and provides a clinical alternative for aortic assessment. We tested intra-operative global circumferential aortic strain (GCS) measured on TEE in relation to the reference of CMR-derived strain among patients undergoing surgical graft repair of ascending aortic aneurysms. METHODS: CMR (3T) was prospectively performed in patients scheduled for aortic repair. TEE was performed intra-operatively; images were co-localized with MRI. GCS on CMR and TEE was quantified independently, blinded to results of the other modality. RESULTS: 25 patients (54 ± 10 year-old, 88% male) were studied, inclusive of 13 genetically mediated and 12 degenerative aneurysms: CMR and TEE were performed within 12 ± 9 days. Pulse pressure (PP)-adjusted descending aortic TEE-derived GCS strongly correlated with cine-CMR-derived GCS (r = .75, P = .002) though absolute GCS and PP-adjusted values were slightly lower (5.40 ± 1.11 vs 6.49 ± 1.43% and 11.55 ± 3.04 vs 13.99 ± 4.53%, respectively). Similarly, TEE yielded slightly lower end-diastolic area (EDA [5.1 ± 1.7 cm2 vs 5.8 ± 1.3 cm2 , P = .004]) and end-systolic area (ESA [6.1 ± 1.9 cm2 vs 6.5 ± 1.7 cm2 , P = .10]), with significant correlations between the two modalities (r = .73, .76, P < .05 for all). CONCLUSIONS: This exploratory study supports feasibility of TEE for assessing aortic GCS in a surgical at-risk population, as well as magnitude of agreement between intra-operative TEE and preoperative CMR. We found that there is a significant correlation between GCS and EDA and ESA aortic areas, but that TEE-derived parameters underestimated CMR values by a small but significant amount.

Differential myocardial strain in the early postoperative period in patients receiving arterial vs venous bypass grafts: A hypothesis-generating study.

OBJECTIVE: Revascularization via coronary artery bypass grafting (CABG) remains a common therapy for coronary artery disease. CABG-based revascularization is most commonly performed via either single arterial graft (SAG) or multiple arterial grafting (MAG) strategies. Echo-derived global and regional longitudinal strain was used to test where SAG or MAG results in immediate differences in left ventricular (LV) function after CABG. MATERIALS AND METHODS: Pre- and postprocedural intraoperative transesophageal echos were prospectively collected. Two-dimensional LV images were analyzed for global and regional longitudinal strain (GLS), LV ejection fraction, end-diastolic volume, end-systolic volume, and stroke volume (SV). RESULTS: Twenty patients underwent open, on-pump CABG (63.9 ± 10 years old, 85% male; 10 with SAG and 10 with MAG. Preprocedural GLS significantly differed between patients with SAG and MAG, with patients with MAG having greater GLS (mean [standard deviation, SD], 20.41 [5.54]) than patients with SAG (16.28 [3.48]). After CABG, in patients with MAG, LV strain decreased both globally (-1.13 [3.15]) and regionally in the anterior-lateral (-1.22 [3.84]) and inferior-lateral regions (-1.32 [5.69]), along with LVEF. In patients with SAG, LV strain increased after CABG globally (1.34 [2.73]) and regionally in the anterior-lateral (1.20 [6.49]) and inferior-lateral regions (0.39 [7.26]), as did LVEF and SV. Postprocedure, more patients with MAG were given vasopressor (100% vs 60%) and inotrope infusions (70% vs 40%) than patients with SAG. CONCLUSIONS: After CABG, LV function quantified through GLS changes both globally and regionally increased after SAG and decreased after MAG. This finding may have important clinical implications in terms of optimizing intraoperative management for patients with CABG and have the potential to guide the improvement of clinical outcomes.