A Machine-Learning Framework to Identify Distinct Phenotypes of Aortic Stenosis Severity

OBJECTIVES

The authors explored the development and validation of machine-learning models for augmenting the echocardiographic grading of aortic stenosis (AS) severity.

BACKGROUND

In AS, symptoms and adverse events develop secondarily to valvular obstruction and left ventricular decompensation. The current echocardiographic grading of AS severity focuses on the valve and is limited by diagnostic uncertainty.

METHODS

Using echocardiography (ECHO) measurements (ECHO cohort, n = 1,052), we performed patient similarity analysis to derive high-severity and low-severity phenogroups of AS. We subsequently developed a supervised machine-learning classifier and validated its performance with independent markers of disease severity obtained using computed tomography (CT) (CT cohort, n = 752) and cardiovascular magnetic resonance (CMR) imaging (CMR cohort, n = 160). The classifier's prognostic value was further validated using clinical outcomes (aortic valve replacement [AVR] and death) observed in the ECHO and CMR cohorts.

RESULTS

In 1,964 patients from the 3 multi-institutional cohorts, 1,346 (68%) subjects had either nonsevere or discordant AS severity. Machine learning identified 1,117 (57%) patients as having high-severity and 847 (43%) as having low-severity AS. High-severity patients in CT and CMR cohorts had higher valve calcium scores and left ventricular mass and fibrosis, respectively than the low-severity group. In the ECHO cohort, progression to AVR and progression to death in patients who did not receive AVR was faster in the high-severity group. Compared with the conventional classification of disease severity, machine-learning-based severity classification improved discrimination (integrated discrimination improvement: 0.07; 95% confidence interval: 0.02 to 0.12) and reclassification (net reclassification improvement: 0.17; 95% confidence interval: 0.11 to 0.23) for the outcome of AVR at 5 years. For both ECHO and CMR cohorts, we observed prognostic value of the machine-learning classifications for subgroups with asymptomatic, nonsevere or discordant AS.

CONCLUSIONS

Machine learning can integrate ECHO measurements to augment the classification of disease severity in most patients with AS, with major potential to optimize the timing of AVR.

Flexibility of microstructural adaptations in airway smooth muscle

The airway smooth muscle undergoes an elastic transition during a sustained contraction, characterized by a gradual decrease in hysteresivity caused by a relatively greater rate of increase in elastance than resistance. We recently demonstrated that these mechanical changes are more likely to persist after a large strain when they are acquired in dynamic versus static conditions; as if the microstructural adaptations liable for the elastic transition are more flexible when they evolve in dynamic conditions. The extent of this flexibility is undefined. Herein, contracted ovine tracheal smooth muscle strips were kept in dynamic conditions simulating tidal breathing (sinusoidal length oscillations at 5% amplitude) and then subjected to simulated deep inspirations (DI). Each DI was straining the muscle by either 10%, 20%, or 30% and was imposed at either 2, 5, 10, or 30 min after the preceding DI. The goal was to assess whether and the extent by which the time-dependent decrease in hysteresivity is preserved following the DI. The results show that the time-dependent decrease in hysteresivity seen pre-DI was preserved after a strain of 10%, but not after a strain of 20% or 30%. This suggests that the microstructural adaptations liable for the elastic transition withstood a strain at least twofold greater than the oscillating strain that pertained during their evolution (10% vs. 5%). We propose that a muscle adapting in dynamic conditions forges microstructures exhibiting a substantial degree of flexibility.NEW & NOTEWORTHY This study confirms that airway smooth muscle undergoes an elastic transition during a sustained contraction even when it operates in dynamic conditions simulating breathing at tidal volume. It also demonstrates that the microstructural adaptations liable for this elastic transition withstand a strain that is at least twice as large as the oscillating strain that pertains during their evolution. This degree of flexibility might be an asset with major significant impact for a tissue such as the airway smooth muscle that displays an everchanging shape due to breathing.

Left ventricular asymmetric remodeling and subclinical left ventricular dysfunction in patients with calcific aortic valve stenosis - Results from a subanalysis of the PROGRESSA study

BACKGROUND

LV asymmetric remodeling (LVAR) is a feature commonly found in AS patients and it is presumed to be mainly related to the severity of valve stenosis. The aim of this study was to determine the associated factors and impact on left ventricular (LV) systolic function of LVAR in patients with mild and moderate aortic valve stenosis (AS).

METHODS

Clinical, Doppler-echocardiographic and computed-tomographic data of 155 AS patients with preserved LV ejection fraction (≥50%) prospectively recruited in the PROGRESSA study (NCT01679431) were analyzed. LVAR was defined as a septal wall thickness ≥ 13 mm and a ratio of septal/posterior wall thickness > 1.5. LV global longitudinal strain (LV-GLS) was available in 129 patients. Plasma levels of N-terminal natriuretic B-type peptides (Nt-proBNP) were also measured.

RESULTS

Mean age was 63 ± 15 years (70% men). LVAR was present in 21% (n = 33) of patients. A series of nested multivariate analysis revealed that age was the only factor associated with LVAR (all p ≤ 0.03). Additionally, these patients had higher baseline Nt-proBNP ratio (median [25-75 percentiles]: 1.04 [0.66-2.41] vs. 0.65 [0.33-1.19], p = 0.02), and significantly reduced LV-GLS (17.9[16.6-19.5] vs. 19.3[17.4-20.7] |%|, p = 0.04). A 1:1 matched analysis showed a significant association of LVAR with reduced LV-GLS (17.9[16.6-19.5] vs. 19.8[18.1-20.7] |%|, p = 0.02) and elevated Nt-proBNP (134[86-348] vs. 83[50-179]pg/ml, p = 0.03). Multivariable analysis also revealed that LVAR remains significantly associated with reduced LV-GLS (p = 0.03) and elevated Nt-proBNP (p = 0.001). LVAR was significantly associated with increased risk of major adverse cardiac events and death (Hazard ratio [95% confidence interval]: 2.32[1.28-4.22], p = 0.006).

CONCLUSIONS

LVAR was found in ~20% of patients with mild or moderate AS and was not related to the degree of AS severity or concomitant comorbidities, but rather to older age. LVAR was significantly associated with reduced LV longitudinal systolic function, increased Nt-proBNP levels, and higher risk of major adverse events and death. These findings provide support for closer clinical and echocardiographic surveillance of patients harboring this adverse LV remodeling feature.

Sex and Race Differences in the Pathophysiology, Diagnosis, Treatment, and Outcomes of Valvular Heart Diseases

Valvular heart diseases have long been considered to be similar in men and women and across races/ethnicities. Recently, studies have demonstrated major differences between sexes. Unfortunately, studies on valvular heart diseases, as on other cardiovascular diseases, are mostly performed in Caucasian men or in cohorts with a vast majority of Caucasian men. Therefore, our knowledge on valvular diseases in women and non-Caucasians remains limited. Nevertheless, aortic stenosis has been shown to be almost as prevalent in women as in men, and less prevalent in African Americans. Men appear to have a more calcified aortic valve lesion, and women tend to have a more fibrosed one. Primary mitral regurgitation is more frequent in women who have more rheumatic and Barlow etiologies, whereas men have more fibroelastic deficiency and posterior leaflet prolapse/flail. Left ventricular remodelling due to valvular heart diseases is sex related in terms of geometry and probably also in composition of the tissue. Outcomes seem to be worse in women after surgical interventions and better than or equivalent to men after transcatheter ones. Regarding other valvular heart diseases, very few studies are available: Aortic regurgitation is more frequent in men, isolated tricuspid regurgitation more frequent in women. Rheumatic valve diseases are more frequent in women and are mostly represented by mitral and aortic stenoses. Many other sex/gender- and race/ethnic-specific studies are still needed in epidemiology, pathophysiology, presentation, management, and outcomes. This review aims to report the available data on sex differences and race specificities in valvular heart diseases, with a primary focus on aortic stenosis and mitral regurgitation.

The Canadian Women's Heart Health Alliance ATLAS on the Epidemiology, Diagnosis, and Management of Cardiovascular Disease in Women-Chapter 2: Scope of the Problem

Background

This Atlas chapter summarizes the epidemiology of cardiovascular disease (CVD) in women in Canada, discusses sex and gender disparities, and examines the intersectionality between sex and other factors that play a prominent role in CVD outcomes in women, including gender, indigenous identity, ethnic variation, disability, and socioeconomic status.

Methods

CVD is the leading cause of premature death in Canadian women. Coronary artery disease, including myocardial infarction, and followed by stroke, accounts for the majority of CVD-related deaths in Canadian women. The majority of emergency department visits and hospitalizations by women are due to coronary artery disease, heart failure, and stroke. The effect of traditional cardiovascular risk factors and their association with increasing cardiovascular morbidity is unique in this group.

Results

Indigenous women in Canada experience increased CVD, linked to colonization and subsequent social, economic, and political challenges. Women from particular racial and ethnic backgrounds (ie, South Asian, Afro-Caribbean, Hispanic, and Chinese North American women) have greater CVD risk factors, and CVD risk in East Asian women increases with duration of stay in Canada.

Conclusions

Canadians living in northern, rural, remote, and on-reserve residences experience greater CVD morbidity, mortality, and risk factors. An increase in CVD risk among Canadian women has been linked with a background of lower socioeconomic status, and women with disabilities have an increased risk of adverse cardiac events.

Transvalvular Flow, Sex, and Survival After Valve Replacement Surgery in Patients With Severe Aortic Stenosis

BACKGROUND

The respective impacts of transvalvular flow, gradient, sex, and their interactions on mortality in patients with severe aortic stenosis undergoing surgical aortic valve replacement (AVR) are unknown.

OBJECTIVES

This study sought to compare the impact of pre-operative flow-gradient patterns on mortality after AVR and to examine whether there are sex differences.

METHODS

This study analyzed clinical, echocardiographic, and outcome data prospectively collected in 1,490 patients (544 women [37%]), with severe aortic stenosis and preserved left ventricular ejection fraction who underwent AVR.

RESULTS

In this cohort, 601 patients (40%) had normal flow (NF) with high gradient (HG), 405 (27%) NF with low gradient (LG), 246 (17%) paradoxical low flow (LF)/HG, and 238 (16%) LF/LG. During a median follow-up of 2.42 years (interquartile range: 1.04 to 4.29 years), 167 patients died. Patients with LF/HG exhibited the highest mortality after AVR (hazard ratio [HR]: 2.01; 95% confidence interval [CI]: 1.33 to 3.03; p < 0.01), which remained significant after multivariate adjustment (HR: 1.96; 95% CI: 1.29 to 2.98; p < 0.01). Both LF/LG and NF/LG patients had comparable outcome to NF/HG (p ≥ 0.47). Optimal thresholds of stroke volume index were obtained for men (40 ml/m²) and women (32 ml/m²). Using these sex-specific cutpoints, paradoxical LF was independently associated with increased mortality in both women (adjusted HR: 2.05; 95% CI: 1.21 to 3.47; p < 0.01) and men (adjusted HR: 1.54; 95% CI: 1.02 to 2.32; p = 0.042), whereas guidelines' threshold (35 ml/m²) does not.

CONCLUSIONS

Paradoxical LF/HG was associated with higher mortality following AVR, suggesting that a reduced flow is a marker of disease severity even in patients with HG aortic stenosis. Early surgical AVR (i.e., before gradient attains 40 mm Hg) might be preferable in these patients. Furthermore, the use of sex-specific thresholds (<40 ml/m² for men and <32 ml/m² for women) to define low-flow outperforms the guidelines' threshold of 35 ml/m² in risk stratification after AVR.

Prognostic Value of N-Terminal Pro-B-Type Natriuretic Peptide in Elderly Patients With Valvular Heart Disease

BACKGROUND

N-terminal pro-B-type natriuretic peptide (NT-proBNP) may reflect early prognosis in patients with valvular heart disease (VHD).

OBJECTIVES

The aim of this study was to examine the association between NT-proBNP and mortality in elderly patients with VHD.

METHODS

A total of 5,983 elderly patients (age ≥60 years) with moderate or severe VHD underwent echocardiography and NT-proBNP measurement. VHD examined included aortic stenosis, aortic regurgitation, mitral stenosis, mitral regurgitation, tricuspid regurgitation, and multivalvular heart disease. NT-proBNP ratio was defined as measured NT-proBNP relative to the maximal normal values specific to age and sex. Disease-specific thresholds were defined on the basis of penalized splines and maximally selected rank statistics.

RESULTS

The cohort had a mean age of 71.1 ± 7.6 years. At 1-year follow-up, 561 deaths (9.4%) had occurred. In penalized splines, relative hazards showed a monotonic increase with greater NT-proBNP ratio for death with different VHDs (p < 0.001 for all) except mitral stenosis. Higher NT-proBNP ratio, categorized by disease-specific thresholds, was independently associated with mortality (overall adjusted hazard ratio: 1.99; 95% confidence interval: 1.76 to 2.24; p < 0.001). Different subtypes of VHD all incurred excess mortality with elevated NT-proBNP ratio, with the strongest association detected for aortic stenosis (adjusted hazard ratio: 10.5; 95% confidence interval: 3.9 to 28.27; p < 0.001). The addition of NT-proBNP ratio to the prediction algorithm including traditional risk factors improved outcome prediction (overall net reclassification index = 0.28; 95% CI: 0.24 to 0.34; p < 0.001; likelihood ratio test p < 0.001). Results remained consistent in patients under medical care, with normal left ventricular ejection fractions, and with primary VHD.

CONCLUSIONS

NT-proBNP provides incremental prognostic information for mortality in various VHDs. It could aid in risk stratification as a pragmatic and versatile biomarker in elderly patients.

Comparison of Early Surgical or Transcatheter Aortic Valve Replacement Versus Conservative Management in Low-Flow, Low-Gradient Aortic Stenosis Using Inverse Probability of Treatment Weighting: Results From the TOPAS Prospective Observational Cohort Study

Background

No randomized comparison of early (ie, ≤3 months) aortic valve replacement (AVR) versus conservative management or of transcatheter AVR (TAVR) versus surgical AVR has been conducted in patients with low‐flow, low‐gradient (LFLG) aortic stenosis (AS).

Methods and Results

A total of 481 consecutive patients (75±10 years; 71% men) with LFLG AS (aortic valve area ≤0.6 cm²/m² and mean gradient <40 mm Hg), 72% with classic LFLG and 28% with paradoxical LFLG, were prospectively recruited in the multicenter TOPAS (True or Pseudo Severe Aortic Stenosis) study. True‐severe AS or pseudo‐severe AS was adjudicated by flow‐independent criteria. During follow‐up (median [IQR] 36 [11–60] months), 220 patients died. Using inverse probability of treatment weighting to address the bias of nonrandom treatment assignment, early AVR (n=272) was associated with a major overall survival benefit (hazard ratio [HR], 0.34 [95% CI, 0.24–0.50]; P<0.001). This benefit was observed in patients with true‐severe AS but also with pseudo‐severe AS (HR, 0.38 [95% CI, 0.18–0.81]; P=0.01), and in classic (HR, 0.33 [95% CI, 0.22–0.49]; P<0.001) and paradoxical LFLG AS (HR, 0.42 [95% CI, 0.20–0.92]; P=0.03). Compared with conservative management in the conventional multivariate model, trans femoral TAVR was associated with the best survival (HR, 0.23 [95% CI, 0.12–0.43]; P<0.001), followed by surgical AVR (HR, 0.36 [95% CI, 0.23–0.56]; P<0.001) and alternative‐access TAVR (HR, 0.51 [95% CI, 0.31–0.82]; P=0.007). In the inverse probability of treatment weighting model, trans femoral TAVR appeared to be superior to surgical AVR (HR [95% CI] 0.28 [0.11–0.72]; P=0.008) with regard to survival.

Conclusions

In this large prospective observational study of LFLG AS, early AVR appeared to confer a major survival benefit in both classic and paradoxical LFLG AS. This benefit seems to extend to the subgroup with pseudo‐severe AS. Our findings suggest that TAVR using femoral access might be the best strategy in these patients.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: NCT01835028.

Multimodality Imaging for Discordant Low-Gradient Aortic Stenosis: Assessing the Valve and the Myocardium

Aortic stenosis (AS) is a disease of the valve and the myocardium. A correct assessment of the valve disease severity is key to define the need for aortic valve replacement (AVR), but a better understanding of the myocardial consequences of the increased afterload is paramount to optimize the timing of the intervention. Transthoracic echocardiography remains the cornerstone of AS assessment, as it is universally available, and it allows a comprehensive structural and hemodynamic evaluation of both the aortic valve and the rest of the heart. However, it may not be sufficient as a significant proportion of patients with severe AS presents with discordant grading (i.e., an AVA ≤ 1 cm² and a mean gradient <40 mmHg) which raises uncertainty about the true severity of AS and the need for AVR. Several imaging modalities (transesophageal or stress echocardiography, computed tomography, cardiovascular magnetic resonance, positron emission tomography) exist that allow a detailed assessment of the stenotic aortic valve and the myocardial remodeling response. This review aims to provide an updated overview of these multimodality imaging techniques and seeks to highlight a practical approach to help clinical decision making in the challenging group of patients with discordant low-gradient AS.

Impact of Left-Ventricular Dysfunction in Patients With High- and Low- Gradient Severe Aortic Stenosis Following Transcatheter Aortic Valve Replacement

BACKGROUND

Outcomes of transcatheter aortic valve replacement (TAVR) in patients with high-gradient (HG) severe aortic stenosis (AS) and reduced left-ventricular (LV) ejection fraction (EF) are unknown.

METHODS

Patients undergoing TAVR for native severe AS between 2009 and 2018 were retrospectively included and classified into 3 groups: HG (≥ 40 mm Hg) and preserved EF (≥ 50%), HG low EF (< 50%), and low gradient (LG < 40 mm Hg) low EF. The primary endpoint was a composite of cardiovascular mortality and readmission for heart failure at 1 year after TAVR.

RESULTS

Of the 526 patients included, 323 (61%) had HG preserved EF, 69 (13%) had HG low EF, and 134 (26%) had LG low EF. HG low EF group had higher prevalence of atrial fibrillation and heart failure and higher Society of Thoracic Surgeons score compared with the HG preserved EF group. Patients in the LG low EF group were older and had higher prevalence of coronary artery disease compared with those in the HG groups. All-cause mortality at 30 days (4.0%) was similar across the 3 groups. After adjustment, the risk of primary endpoint was similar in the HG low-EF vs preserved EF groups. Conversely, the risk of primary endpoint was higher in the LG low EF group vs the HG preserved EF group (hazard ratio [HR], 2.24; 95% confidence interval [CI],1.36-3.70; P = 0.002) and vs HG low EF group (HR, 3.50; 95% CI, 1.55-7.90; P = 0.003), whereas the risk of all-cause mortality was similar across the 3 groups.

CONCLUSIONS

The outcome of patients with HG low EF severe AS following TAVR is as good as that of patients with HG preserved EF.

Association of Natriuretic Peptide Levels After Transcatheter Aortic Valve Replacement With Subsequent Clinical Outcomes

Importance

Among those with aortic stenosis, natriuretic peptide levels can provide risk stratification, predict symptom onset, and aid decisions regarding the timing of valve replacement. Less is known about the prognostic significance and potential clinical utility of natriuretic peptide levels measured after valve replacement.

Objective

To determine the associations of elevated B-type natriuretic peptide (BNP) levels after transcatheter aortic valve replacement (TAVR) and change in BNP levels between follow-up time points with risk of subsequent clinical outcomes.

Design, Setting, and Participants

In this cohort study, patients with severe symptomatic aortic stenosis at intermediate, high, or prohibitive surgical risk for aortic valve replacement who underwent TAVR from the PARTNER IIA cohort, PARTNER IIB cohort, SAPIEN 3 intermediate-risk registry, and SAPIEN 3 high-risk registry were included. B-type natriuretic peptide levels were obtained at baseline and discharge as well as 30 days and 1 year after TAVR. For each measurement, a BNP ratio was calculated using measured BNP level divided by the upper limit of normal for the assay used. Outcomes were evaluated in landmark analyses out to 2 years. Data were collected from April 2011 to January 2019.

Main Outcomes and Measures

All-cause death, cardiovascular death, rehospitalization, and the combined end point of cardiovascular death or rehospitalization.

Results

Among 3391 included patients, 1969 (58.1%) were male, and the mean (SD) age was 82 (7.5) years. Most patients had a BNP ratio greater than 1 at each follow-up time point, including 2820 of 3256 (86.6%) at baseline, 2652 of 2995 (88.5%) at discharge, 1779 of 2209 (80.5%) at 30 days, and 1799 of 2391 (75.2%) at 1 year. After adjustment, every 1-point increase in BNP ratio at 30 days (approximately equivalent to an increase of 100 pg/mL in BNP) was associated with an increased hazard of all-cause death (adjusted hazard ratio [aHR], 1.11; 95% CI, 1.07-1.15), cardiovascular death (aHR, 1.16; 95% CI, 1.11-1.21), and rehospitalization (aHR, 1.08; 95% CI, 1.03-1.14) between 30 days and 2 years. Among those with a BNP ratio of 2 or more at discharge, after adjustment, every 1-point decrease in BNP ratio between discharge and 30 days was associated with a decreased hazard of all-cause death (aHR, 0.92; 95% CI, 0.88-0.96) between 30 days and 2 years.

Conclusions and Relevance

Elevated BNP levels after TAVR was independently associated with increased subsequent mortality and rehospitalizations. Further studies to determine how best to mitigate this risk are warranted.

Balloon aortic valvuloplasty as a palliative treatment in patients with severe aortic stenosis and limited life expectancy: a single center experience

Whether balloon aortic valvuloplasty (BAV) may provide an effective palliation in symptomatic high-risk patients is uncertain. Therefore, we aimed to evaluate outcomes in symptomatic high-risk patients with severe aortic stenosis (AS), who underwent BAV. All-cause mortality and length of hospitalization for heart failure (HF) up to death or to 1-year follow up were collected after BAV. One hundred thirty-two (132) patients (62% women), mean age 85±7 years, underwent BAV with a substantial reduction of the peak-to-peak aortic gradient from 53±21 to 29±15 mmHg (p<0.001). The median of days of HF hospitalization prior to BAV was 9 (0-19), and decreased after BAV to 0 (0-9), p<0.001. During 1-year follow-up patients with untreated CAD (85, 64%) had a higher mortality compared to patients with insignificant/treated CAD (47, 36%): 1-year survival: 45±7% vs. 66± 7%; p=0.02. After adjustment for STS risk score and severity of residual AS, patients with untreated CAD remained at higher risk of mortality (adjusted HR 1.74 [1.01-2.91]; p=0.04). Thus, in this series of symptomatic high-risk patients, BAV was associated with a significant reduction in aortic valve gradient and hospitalization time for HF post-BAV. In patients with significant CAD, percutaneous intervention might be considered in order to improve survival.

Attenuated Mitral Leaflet Enlargement Contributes to Functional Mitral Regurgitation After Myocardial Infarction

BACKGROUND

Mitral leaflet enlargement has been identified as an adaptive mechanism to prevent mitral regurgitation in dilated left ventricles (LVs) caused by chronic aortic regurgitation (AR). This enlargement is deficient in patients with functional mitral regurgitation, which remains frequent in the population with ischemic cardiomyopathy. Maladaptive fibrotic changes have been identified in post-myocardial infarction (MI) mitral valves. It is unknown if these changes can interfere with valve growth and whether they are present in other valves.

OBJECTIVES

This study sought to test the hypothesis that MI impairs leaflet growth, seen in AR, and induces fibrotic changes in mitral and tricuspid valves.

METHODS

Sheep models of AR, AR + MI, and controls were followed for 90 days. Cardiac magnetic resonance, echocardiography, and computed tomography were performed at baseline and 90 days to assess LV volume, LV function, mitral regurgitation and mitral leaflet size. Histopathology and molecular analyses were performed in excised valves.

RESULTS

Both experimental groups developed similar LV dilatation and dysfunction. At 90 days, mitral valve leaflet size was smaller in the AR + MI group (12.8 ± 1.3 cm² vs. 15.1 ± 1.6 cm², p = 0.03). Mitral regurgitant fraction was 4% ± 7% in the AR group versus 19% ± 10% in the AR + MI group (p = 0.02). AR + MI leaflets were thicker compared with AR and control valves. Increased expression of extracellular matrix remodeling genes was found in both the mitral and tricuspid leaflets in the AR + MI group.

CONCLUSIONS

In these animal models of AR, the presence of MI was associated with impaired adaptive valve growth and more functional mitral regurgitation, despite similar LV size and function. More pronounced extracellular remodeling was observed in mitral and tricuspid leaflets, suggesting systemic valvular remodeling after MI.