Heart failure with preserved ejection fraction: neither deformation nor timing are preserved

This editorial refers to ‘Myocardial systolic and diastolic consequences of left ventricular mechanical dyssynchrony in heart failure with normal left ventricular ejection fraction’ by Morris et al., on pages 556–567. Despite its frequency, heart failure with preserved ejection fraction (HFpEF) remains a riddle.1 This condition is highly heterogeneous and probably not uncommonly misdiagnosed, especially if the diagnosis is one of exclusion.2 For these and many more reasons, the pharmacological treatment of HFpEF remains difficult, with multiple negative survival studies, even though some medications have been linked to functional benefit.3 In this context, the paper of Morris et al.4 on the association of left ventricular mechanical dyssynchrony with HFpEF is interesting, and may prove to be very important. In this work, the authors studied 325 patients, 85 with HFpEF and 240 with asymptomatic diastolic dysfunction in the absence of heart failure symptoms. The diagnostic criteria were proposed by the European Society of Cardiology,5 and patients with confounding diseases were excluded. In comparison to those with asymptomatic diastolic dysfunction, patients with HFpEF were more likely to be hypertensive, have type 2 diabetes mellitus, and have a history of coronary artery disease. They had LV hypertrophy, lower tissue velocity and lower systolic and diastolic myocardial deformation, as documented previously. HFpEF patients showed both systolic and diastolic dyssynchrony, and the presence of dyssynchrony within the HFpEF group was associated with worse systolic and diastolic strain than in patients with synchrony, with systolic dysfunction being three times more common, and diastolic dysfunction being approximately twice as frequent in the presence of dyssynchrony. Dyssynchrony was also associated with high filling pressures and worse functional class. All of these observations are understandable in a teleologic sense—heterogeneity of contraction and relaxation could be expected to lead to inefficient systole and diastole. These findings introduce some interesting questions. First, what is the driver of this mechanical dyssynchrony in these patients? As the authors point out, cardiomyopathic processes are associated with myocardial fibrosis and this may be patchy rather than diffuse, leading to dispersion of mechanical activation. This link has been shown to be prognostically important.6 However, a more mundane explanation may be relevant. Nearly two thirds of patients with HFpEF had a history of coronary disease, and this alone could be expected to produce mechanical dyssynchrony. While the dyssynchrony index was an independent predictor of systolic and diastolic dysfunction, LV filling pressures and functional class, even after correction for the history of coronary disease, it is not so much history but scar extent and location that may be the determinants of mechanical dyssynchrony. The aetiology of dyssynchrony is of critical importance, as resynchronization therapy is suitable for patients with electrical dyssynchrony, and rather unlikely to improve the timing of contractility in patients with mechanical dyssynchrony.7 The second question which might arise from these findings is to what extent this is truly a characteristic of HFpEF, given that at least a part of the enigma of HFpEF is attributable to the fact that ejection fraction is a suboptimal marker of left ventricular systolic function. Indeed, it is known from the literature on heart failure with reduced ejection fraction that dyssynchrony is associated with worse function and worse outcome.8 Conversely, while differences between HFpEF and asymptomatic diastolic dysfunction have been documented in this paper, dyssynchrony is certainly evident in a variety of disease entities that are not necessarily associated with heart failure symptoms. Pai et al.9 described regional delay in patients with secondary left ventricular hypertrophy, and an association between non-uniformity and LV filling has been documented in hypertensive heart disease.10 Similar reports of abnormal diastolic timing have been made in hypertrophic cardiomyopathy and coronary disease.11,12 One previous study has documented persistence of long axis tension at the same time as short axis lengthening.13 Thus, dyssynchrony is not rare, and we might be moved to question whether this is an epiphenomenon of another process that drives symptoms. Third, to what extent is the nature of the technique used for defining dyssynchrony important? In a group of subjects with type 2 diabetes mellitus, Burgess et al.14 compared deformation characteristics and diastolic dyssynchrony in individuals with impaired and normal left ventricular filling. While diastolic dyssynchrony was associated with delayed relaxation, the association of functional capacity with this was much less strong than the association with disturbances of longitudinal function. Diastolic dyssynchrony was not associated with impaired exercise capacity. The difference between these findings and those of the current study may arise from either the use of tissue velocity imaging, which is subject to tethering and translational effects, or the selection of subjects with diabetes mellitus, who may have a different mechanism of disturbance than the patients investigated by the authors. Finally, how is this information to be used in clinical management? Studies have already reported failure of cardiac resynchronization therapy in patients with a narrow QRS, who would overlap with this population.15 Indeed, if the lessons of patients with heart failure with reduced ejection fraction can be extrapolated to this population, patients without a bundle branch block, and with coronary disease can be expected to have little response to CRT. Additional studies are currently in progress relating to the use of mechanical markers for measuring synchrony, and no doubt much will be learned from this work. Nonetheless it has been very difficult to show convincingly that mechanical dyssynchrony should be a criterion for resynchronization therapy, particularly given the limitations in the accuracy and reproducibility of current synchrony measures. The heterogeneity of HFpEF has been one of the reasons why defining appropriate management has been so difficult. In this respect, the study published by Morris et al.4 may prove a useful step in permitting more effective subcategorization of HFpEF. Whether the characterization of mechanical dyssynchrony permits a specific management strategy remains to be determined. Conflict of interest: none declared.