Tamas Erdei

A systematic review of diastolic stress tests in heart failure with preserved ejection fraction, with proposals from the EU-FP7 MEDIA study group

Cardiac function should be assessed during stress in patients with suspected heart failure with preserved ejection fraction (HFPEF), but it is unclear how to define impaired diastolic reserve.

Pathophysiological rationale and diagnostic targets for diastolic stress testing

Cardiopulmonary functional reserve measured as peak oxygen uptake is predicted better at rest by measures of cardiac diastolic function than by systolic function. Normal adaptations in the trained heart include resting bradycardia, increased LV end-diastolic volume and augmented early diastolic suction on exercise. In normal populations early diastolic relaxation declines with age and end-diastolic stiffness increases, but in healthy older subjects who have exercised throughout their lives diastolic function can be well preserved. The mechanisms by which LV diastolic filling and pressures can be impaired during exercise include reduced early diastolic recoil and suction (which can be exacerbated by increased late systolic loading), increased preload and reduced compliance. Abnormal ventricular-arterial coupling and enhanced ventricular interaction may contribute in particular circumstances. One common final pathway that causes breathlessness is an increase in LV filling pressure and left atrial pressure. Testing elderly subjects with breathlessness of unknown aetiology in order to detect worsening diastolic function during stress is proposed to diagnose heart failure with preserved EF. In invasive studies, the most prominent abnormality is an early and rapid rise in pulmonary capillary wedge pressure. A systematic non-invasive diagnostic strategy would use validated methods to assess different mechanisms of inducible diastolic dysfunction and not just single parameters that offer imprecise estimates of mean LV filling pressure. Protocols should assess early diastolic relaxation and filling as well as late diastolic filling and compliance, as these may be affected separately. Better refined diagnostic targets may translate to more focused treatment.

Characterization of myocardial motion patterns by unsupervised multiple kernel learning

&NA; We propose an independent objective method to characterize different patterns of functional responses to stress in the heart failure with preserved ejection fraction (HFPEF) syndrome by combining multiple temporally‐aligned myocardial velocity traces at rest and during exercise, together with temporal information on the occurrence of cardiac events (valves openings/closures and atrial activation). The method builds upon multiple kernel learning, a machine learning technique that allows the combination of data of different nature and the reduction of their dimensionality towards a meaningful representation (output space). The learning process is kept unsupervised, to study the variability of the input traces without being conditioned by data labels. To enhance the physiological interpretation of the output space, the variability that it encodes is analyzed in the space of input signals after reconstructing the velocity traces via multiscale kernel regression. The methodology was applied to 2D sequences from a stress echocardiography protocol from 55 subjects (22 healthy, 19 HFPEF and 14 breathless subjects). The results confirm that characterization of the myocardial functional response to stress in the HFPEF syndrome may be improved by the joint analysis of multiple relevant features. HighlightsMultiple myocardial velocity patterns from a stress protocol are jointly analyzed.Unsupervised Multiple Kernel Learning is used to reduce the complexity of the data.The variability analysis on the learnt space unravels healthy/diseased differences.The joint analysis of multiple patterns notably improves the characterization. Graphical abstract Figure. No caption available.

Characterization of Myocardial Velocities by Multiple Kernel Learning: Application to Heart Failure with Preserved Ejection Fraction

The present study aims at improving the characterization of myocardial velocities in the context of heart failure with preserved ejection fraction (HFPEF) by combining multiple descriptors. It builds upon a recent extension of manifold learning known as multiple kernel learning (MKL), which allows the combination of data of different natures towards the learning. Such learning is kept unsupervised, thus benefiting from all the inherent explanatory power of the data without being conditioned by a given class. The methodology was applied to 2D sequences from a stress echocardiography protocol from 33 subjects (21 healthy controls and 12 HFPEF subjects). Our method provides a novel way to tackle the understanding of the HFPEF syndrome, in contrast with the diagnostic issues surrounding it in the current clinical practice. Notably, our results confirm that the characterization of the myocardial functional response to stress in this syndrome is improved by the joint analysis of multiple relevant features.

Machine Learning Analysis of Left Ventricular Function to Characterize Heart Failure With Preserved Ejection Fraction

Background: Current diagnosis of heart failure with preserved ejection fraction (HFpEF) is suboptimal. We tested the hypothesis that comprehensive machine learning (ML) of left ventricular function at rest and exercise objectively captures differences between HFpEF and healthy subjects. Methods and Results: One hundred fifty-six subjects aged >60 years (72 HFpEF+33 healthy for the initial analyses; 24 hypertensive+27 breathless for independent evaluation) underwent stress echocardiography, in the MEDIA study (Metabolic Road to Diastolic Heart Failure). Left ventricular long-axis myocardial velocity patterns were analyzed using an unsupervised ML algorithm that orders subjects according to their similarity, allowing exploration of the main trends in velocity patterns. ML identified a continuum from health to disease, including a transition zone associated to an uncertain diagnosis. Clinical validation was performed (1) to characterize the main trends in the patterns for each zone, which corresponded to known characteristics and new features of HFpEF; the ML-diagnostic zones differed for age, body mass index, 6-minute walk distance, B-type natriuretic peptide, and left ventricular mass index (P<0.05) and (2) to evaluate the consistency of the proposed groupings against diagnosis by current clinical criteria; correlation with diagnosis was good (&kgr;, 72.6%; 95% confidence interval, 58.1–87.0); ML identified 6% of healthy controls as HFpEF. Blinded reinterpretation of imaging from subjects with discordant clinical and ML diagnoses revealed abnormalities not included in diagnostic criteria. The algorithm was applied independently to another 51 subjects, classifying 33% of hypertensive and 67% of breathless controls as mild-HFpEF. Conclusions: The analysis of left ventricular long-axis function on exercise by interpretable ML may improve the diagnosis and understanding of HFpEF.

Non-invasively estimated left atrial stiffness is associated with short-term recurrence of atrial fibrillation after electrical cardioversion

BACKGROUND As atrial stiffness (Kla) is an important determinant of cardiac pump function, better mechanical characterization of left atrial (LA) cavity would be clinically relevant. Pulmonary venous ablation is an option for atrial fibrillation (AF) treatment that offers a powerful context for improving our understanding of LA mechanical function. We hypothesized that a relation could be detected between invasive estimation of Kla and new non-invasive deformation parameters and traditional LA and left ventricular (LV) function descriptors, so that Kla can be estimated non-invasively. We also hypothesized that a non-invasive surrogate of Kla would be useful in predicting AF recurrence after cardioversion. METHODS In 20 patients undergoing AF ablation, LA pressure-volume curves were derived from invasive pressure and echocardiographic images; Kla was calculated during ascending limb of V-loop as ΔLA pressure/ΔLA volume. 2D-speckle-tracking echocardiographic LA and LV longitudinal strains and volumes, ejection fraction (EF) and ventricular stiffness (Klv), as obtained from mitral deceleration time, were tested as non-invasive Kla predictors. In 128 sinus rhythm patients 1 month after electrical cardioversion for persistent AF, non-invasively estimated Kla (computed-Kla) was tested as predictor of recurrence at 6 months. RESULTS Tertiles of mean LA pressure correlated with increasing Kla (trend, p=0.06) and decreasing LA peak strain, LVEF, and LV longitudinal strain (p=0.029, p=0.019, and p=0.024). There were no differences in LA and LV volumes and Klv across groups. Multiple regression analysis identified LV longitudinal strain as the only independent predictor of Kla (p=0.014). Patients in highest quartile of computed-Kla (estimated as [log]=0.735+0.051×LV strain) tended to have highest AF recurrence rate (25%) as compared with remaining 3 quartiles (9%, 9%, 3%, p=0.09). CONCLUSION Kla can be assessed invasively in patients undergoing AF ablation and it can be estimated non-invasively using LV strain. AF recurrence after cardioversion tends to be highest in highest quartile of computed-Kla.

Diagnosis of Heart Failure With Preserved Ejection Fraction: Machine Learning of Spatiotemporal Variations in Left Ventricular Deformation

Background: Stress testing helps diagnose heart failure with preserved ejection fraction (HFpEF), but there are no established criteria for quantifying left ventricular (LV) functional reserve. The aim of this study was to investigate whether comprehensive analysis of the timing and amplitude of LV long‐axis myocardial motion and deformation throughout the cardiac cycle during rest and stress can provide more informative criteria than standard measurements. Methods: Velocity, strain, and strain rate traces were measured from all 18 LV segments by echocardiographic myocardial velocity imaging at rest and during semisupine bicycle exercise in 100 subjects aged 69 ± 7 years, including patients with HFpEF and healthy, hypertensive, and breathless control subjects. A machine‐learning algorithm, composed of an unsupervised statistical method and a supervised classifier, was used to model spatiotemporal patterns of the traces and compare the predicted labels with the clinical diagnoses. Results: The learned strain rate parameters gave the highest accuracy for allocating subjects into the four groups (overall, 57%; for patients with HFpEF, 81%), and into two classes (asymptomatic vs symptomatic; area under the curve, 0.89; accuracy, 85%; sensitivity, 86%; specificity, 82%). Machine learning of strain rate, compared with standard measurements, gave the greatest improvement in accuracy for the two‐class task (+23%, P < .0001), compared with +11% (P < .0001) using velocity and +4% (P < .05) using strain. Strain rate was also best at predicting 6‐min walk distance as an independent reference criterion. Conclusions: Machine learning of spatiotemporal variations of LV strain rate during rest and exercise could be used to identify patients with HFpEF and to provide an objective basis for diagnostic classification.

16 Chronotropic and dromotropic responses to exercise in heart failure with preserved ejection fraction (hfpef) patients and controls

Introduction The hallmark of Heart Failure with Preserved Ejection Fraction (HFpEF) is exercise intolerance. The mechanisms for this are numerous, but chronotropic incompetence, defined as a failure to reach at least 70% of the age-predicted maximum heart rate (HR) on maximal exercise, has been reported to contribute. Impaired Heart Rate Reserve, a measure of HR achieved on maximal exercise compared with age-predicted maximum heart rate, is correlated with negative cardiovascular outcomes. In normal subjects, the PR interval shortens during exercise as parasympathetic tone reduces. This is known as dromotropy and is reduced in subjects with coronary artery disease. We formed the hypothesis that HFpEF patients may also have impaired PR shortening and that this may contribute to exercise intolerance. Methods and results HFpEF patients and controls (healthy (H), hypertensive (HT) and breathless controls (BC)) from the MEDIA (Metabolic Road to Diastolic Heart Failure, EU FP7) trial at our centre underwent maximal semi-supine bicycle stress tests whilst on standard treatment. Electrocardiograms were examined by a single, blinded investigator for PR-interval and heart rate (HR) at: rest; submaximal exercise (HR 100 min-1); peak exercise; and 2 and 5 min after exercise. 110 subjects were recruited, of whom 24 were excluded (unable to exercise/atrial arrhythmia). Data on 86 subjects were analysed. Baseline characteristics are given in the Table. Resting HR was similar between groups, but maximal HR was lower in patients than all controls (p<0.05), as was HR (p<0.0001) (figure 1). HR recovery was impaired in patients versus healthy controls at 2 min (H=−34.8, p<0.001) and 5 min (H=−17.9, p=0.04) of recovery. PR interval decreased across all groups from rest to peak (mean −7.2%, p=0.025). PR interval was similar at rest; it shortened less in patients (5.2%) versus healthy controls (11.4%), but not when corrected for difference in peak HR. PR/HR did not vary by group at any stage. Beta blockers did not affect PR interval, but reduced HR and HR at all stages (35% reduction from rest to peak) (figure 2). The difference in maximal HR between patients and controls persisted after excluding beta blocked subjects. Patients were 13% older than controls (p<0.01). Subjects from all groups of less than 70 years had significantly increased PR shortening (mean difference −9.1 ms, p=0.02) and change in HR (mean difference 15.7 ms, p<0.001) between rest and peak than those over 70 years. Patients had significantly higher Body Mass Indices (BMI) than hypertensive or healthy subjects (H=−3.6, p=0.02, H=−3.5, p=0.003 respectively), as did breathless controls (H=3.1, p=0.012, H=2.8, p=0.027). Conclusions Patients with HFpEF have chronotropic incompetence and impaired HR recovery, but there was no evidence of dromotropic incompetence in this study.Abstract 16 Figure 1 a) Median heart rate and b) Mean PR interval by stage of exercise for each patient group. Bars represent 95% Confidence Interval. P = Patients, BC = Breathless Controls, HT = Hypertensive Subjects, H = Healthy Controls.Abstract 16 Figure 2 Effect of beat blockers on a) heart rate and b) PR interval by stage of exercise. Bars represent 95% confidence limits. Abstract 16 Table 1