Left atrial strain predicts exercise capacity in heart failure independently of left ventricular ejection fraction.

AIMS: We hypothesized that left atrial (LA) remodelling and function are associated with poor exercise capacity as prognostic marker in chronic heart failure (CHF) across a broad range of left ventricular ejection fraction (LVEF). METHODS AND RESULTS: One hundred seventy-one patients with CHF were analysed [age 65 ± 11 years, 136 males (80%); 86 heart failure with reduced ejection fraction (HFrEF), 27 heart failure with mid-range ejection fraction (HFmrEF), 58 heart failure with preserved ejection fraction (HFpEF)]. All patients underwent echocardiography and maximal cardiopulmonary exercise testing and were classified according to a prognostic cut-off of peak VO2 (pVO2 ; 14 mL/kg/min). Seventy-seven (45%) patients reached pVO2  < 14 and 94 (55%) pVO2  ≥ 14 mL/kg/min. Between the two groups, there was a considerable difference in both left atrial volume (LAVi, 53 ± 24 vs. 44 ± 18 mL/m2 , P = 0.005) and function (LA reservoir strain 12 ± 5 vs. 20 ± 10%, P < 0.0001). Receiver-operating characteristic curves identified LA reservoir strain (area under the curve: 0.73 [0.65-0.80], P < 0.0001) as strong predictor for impaired pVO2 among all echocardiographic variables; LA reservoir strain < 23% had 37% specificity but a very high sensitivity (96%) in identifying a severely reduced pVO2 . In logistic regression analysis, LA reservoir strain < 23% was associated with a highly increased risk of pVO2  < 14 mL/kg/min (odds ratio 16.0 [4.7-54.6]; P < 0.0001). The multivariate analysis showed that a reduced LA reservoir strain was associated with pVO2  < 14 mL/kg/min after adjustment for age, body mass index (BMI), and clinical variables, that is, New York Heart Association class, atrial fibrillation, haemoglobin, and creatinine (b 0.22 [95% confidence interval, CI, 0.12-0.31]; P < 0.0001), and after adjustment for echocardiographic variables, that is, LVEF or left ventricular global longitudinal strain (LVGLS) and tricuspid annular plane systolic excursion (TAPSE) (b 0.16 [95% CI 0.08-0.24]; P < 0.0001). Patients with HFrEF, HFmrEF, and HFpEF were separately analysed. Among LA reservoir strain, LAVi, LVEF, LVGLS, and TAPSE, LA reservoir strain was the only one significantly associated with pVO2 in all subgroups (after adjustment for sex and BMI, P = 0.003, 0.04, and 0.01, respectively). CONCLUSIONS: In patients with CHF, an impaired LA reservoir function is independently associated with a severely reduced pVO2 . LA dysfunction represents a marker of poor prognosis across LVEF borders in the CHF population.

Left atrial function and maximal exercise capacity in heart failure with preserved and mid-range ejection fraction.

AIMS: Exercise intolerance is the leading manifestation of heart failure with preserved or mid-range ejection fraction (HFpEF or HFmrEF), and left atrial (LA) function might contribute to modulating left ventricular filling and pulmonary venous pressures. We aim to assess the association between LA function and maximal exercise capacity in patients with HFpEF or HFmrEF. METHODS AND RESULTS: Sixty-five patients, prospectively enrolled in the German HFpEF Registry, were analysed. Inclusion criteria were New York Heart Association functional class ≥ II, left ventricular ejection fraction > 40%, structural heart disease or diastolic dysfunction, and elevated levels of N terminal pro brain natriuretic peptide (NT-proBNP). LA function was evaluated through speckle-tracking echocardiography by central reading in the Charité Academic Echocardiography core lab. All patients underwent maximal cardiopulmonary exercise test and were classified according to a peak VO2 cut-off of prognostic value (14 mL/kg/min). NT-pro-BNP was measured. Twenty-nine patients (45%) reached a peak VO2  < 14 mL/kg/min (mean value 12.4 ± 1.5) and 36 patients (55%) peak VO2  ≥ 14 mL/kg/min (mean value 19.4 ± 3.9). There was no significant difference in left ventricular ejection fraction (60 ± 9 vs. 59 ± 8%), left ventricular mass (109 ± 23 vs. 112 ± 32 g/m2 ), LA volume index (45 ± 17 vs. 47 ± 22 mL/m2 ), or E/e´ (13.1 ± 4.7 vs. 13.0 ± 6.0) between these groups. In contrast, all LA strain measures were impaired in patients with lower peak VO2 (reservoir strain 14 ± 5 vs. 21 ± 9%, P = 0.002; conduit strain 9 ± 2 vs. 13 ± 4%, P = 0.001; contractile strain 7 ± 4 vs. 11 ± 6%, P = 0.02; reported lower limits of normality for LA reservoir, conduit and contractile strains: 26.1%, 12.0%, and 7.7%). In linear regression analysis, lower values of LA reservoir strain were associated with impaired peak VO2 after adjustment for age, sex, body mass index, heart rhythm (sinus/AFib), and log-NTproBNP [ß 0.29, 95% confidence interval (CI) 0.02-0.30, P = 0.02], with an odds ratio 1.22 (95% CI 1.05-1.42, P = 0.01) for peak VO2  < 14 mL/kg/min for LA reservoir strain decrease after adjustment for these five covariates. Adding left ventricular ejection fraction, it did not influence the results. On the other hand, the addition of LA strain to the adjustment parameters alone described above provided a significant increase of the predictive value for lower peak VO2 values (R2 0.50 vs. 0.45, P = 0.02). With receiver operating characteristic curve analysis, we identified LA reservoir strain < 22% to have 93% sensitivity and 49% specificity in predicting peak VO2  < 14 mL/kg/min. Using this cut-off, LA reservoir strain < 22% was associated with peak VO2  < 14 mL/kg/min in logistic regression analysis after comprehensive adjustment for age, sex, body mass index, heart rhythm, and log-NTproBNP [odds ratio 95% CI 10.4 (1.4-74), P = 0.02]. CONCLUSIONS: In this HFpEF and HFmrEF cohort, a reduction in LA reservoir strain was a sensible marker of decreased peak exercise capacity. Therefore, LA reservoir strain might be of clinical value in predicting exercise capacity in patients with HFpEF or HFmrEF.

Left atrial strain as sensitive marker of left ventricular diastolic dysfunction in heart failure.

AIMS: The purpose of this retrospective analysis was to examine the association of left atrial (LA) strain (i.e. LA reservoir function) with left ventricular diastolic dysfunction (DD) in patients with heart failure with reduced and preserved left ventricular ejection fraction (LVEF). METHODS AND RESULTS: We analysed the baseline echocardiographic recordings of 300 patients in sinus rhythm from the SOCRATES-PRESERVED and SOCRATES-REDUCED studies. LA volume index was normal in 89 (29.7%), of whom 60.6% had an abnormal LA reservoir strain (i.e. ≤23%). In addition, the extent of LA strain impairment was significantly associated with the severity of DD according to the 2016 American Society of Echocardiography recommendations (DD grade I: LA strain 22.2 ± 6.6, rate of abnormal LA strain 62.9%; DD grade II: LA strain 16.6 ± 7.4, rate of abnormal LA strain 88.6%; DD grade III: LA strain 11.1 ± 5.4%, rate of abnormal LA strain 95.7%; all P < 0.01). In line with these findings, LA strain had a good diagnostic performance to determine severe DD [area under the curve 0.83 (95% CI 0.77-0.88), cut-off 14.1%, sensitivity 80%, specificity 77.8%], which was significantly better than for LA volume index, LA total emptying fraction, and the mitral E/e' ratio. CONCLUSIONS: The findings of this analysis suggest that LA strain could be a useful parameter in the evaluation of DD in patients with heart failure and sinus rhythm, irrespective of LVEF.

Diastolic stress test echocardiography in patients with suspected heart failure with preserved ejection fraction: a pilot study.

AIMS: The purpose of this pilot study was to assess the potential usefulness of diastolic stress test (DST) echocardiography in patients with suspected heart failure with preserved ejection fraction (HFpEF). METHODS AND RESULTS: Patients with suspected HFpEF (left ventricular ejection fraction ≥ 50%, exertional dyspnoea, septal E/e' at rest 9-14, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) at rest < 220 pg/mL; n = 13) and a control group constituted from asymptomatic patients with arterial hypertension (n = 19) and healthy subjects (n = 18) were included. All patients were analysed by two-dimensional and Doppler echocardiography at rest and during exercise (DST) and underwent cardiopulmonary exercise testing and NT-proBNP analysis during exercise. HFpEF during exercise was defined as exertional dyspnoea and peak VO2  ≤ 20.0 mL/min/kg. In patients with suspected HFpEF at rest, 84.6% of these patients developed HFpEF during exercise, whereas in the group of asymptomatic patients with hypertension and healthy subjects, the rate of developed HFpEF during exercise was 0%. Regarding the diagnostic performance of DST to detect HFpEF during exercise, an E/e' ratio >15 during exercise was the most accurate parameter to detect HFpEF (accuracy 86%), albeit a low sensitivity (45.5%). Nonetheless, combining E/e' with tricuspid regurgitation (TR) velocity > 2.8 m/s during exercise provided a significant increase in the sensitivity to detect patients with HFpEF during exercise (sensitivity 72.7%, specificity 79.5%, and accuracy 78%). Consistent with these findings, an increase of E/e' was significantly linked to worse peak VO2 , and the combination of an increase of both E/e' and TR velocity was associated with elevated NT-proBNP values during exercise. CONCLUSIONS: The findings of this pilot study suggest that DST using E/e' ratio and TR velocity could be of potential usefulness to diagnose HFpEF during exercise in patients with suspected HFpEF at rest.

Cardiac anomaly detection based on time and frequency domain features using tree-based classifiers.

OBJECTIVE: Recent advantages in mHealth-enabled ECG recorders boosted the demand for algorithms, which are able to automatically detect cardiac anomalies with high accuracy. APPROACH: We present a combined method of classical signal analysis and machine learning which has been developed during the Computing in Cardiology Challenge (CinC) 2017. Almost 400 hand-crafted features have been developed to reflect the complex physiology of cardiac arrhythmias and their appearance in single-channel ECG recordings. For the scope of this article, we performed several experiments on the publicly available challenge dataset to improve the classification accuracy. We compared the performance of two tree-based algorithms-gradient boosted trees and random forests-using different parameters for learning. We assessed the influence of five different sets of training annotations on the classifiers performance. Further, we present a new web-based ECG viewer to review and correct the training labels of a signal data set. Moreover, we analysed the feature importance and evaluated the model performance when using only a subset of the features. The primary data source used in the analysis was the dataset of the CinC 2017, consisting of 8528 signals from four classes. Our best results were achieved using a gradient boosted tree model which worked significantly better than random forests. MAIN RESULTS: Official results of the challenge follow-up phase provided by the Challenge organizers on the full hidden test set are 90.8% (Normal), 84.1% (AF), 74.5% (Other), resulting in a mean F1-score of 83.2%, which was only 1.6% behind the challenge winner and 0.2% ahead of the next-best algorithm. Official results were rounded to two decimal places which lead to the equal-second best F1 F -score of 83% with five others. SIGNIFICANCE: The algorithm achieved the second-best score among 80 algorithms of the Challenge follow-up phase equal with five others.