• heart failure with preserved ejection fraction

Heart failure with preserved ejection fraction (HFpEF) afflicts millions worldwide and clinical trials have failed to identify effective treatments.1 This failure is related partly to pathophysiological heterogeneity and partly to incomplete pathophysiological understanding of the different phenotypes. Elevation in left ventricular filling pressures during exercise is the one pathophysiological component that is shared by all HFpEF phenotypes,1 2 but high filling pressures do not appear to explain everything that goes wrong in people with this syndrome.

In order to perform work, mitochondria in skeletal must generate ATP through oxidative metabolism.3 As the work of exercise increases, there is matching increase in the amount of oxygen consumed (VO₂) to provide fuel. Increases in VO₂ are achieved by enhancing O₂ delivery through several integrated steps involving multiple organ systems. The lungs are responsible for oxygenation of erythrocytes, and the heart accelerates these erythrocytes through the circulation to the tissues in a process termed convective O₂ transport. This ‘bulk transfer’ step is analogous to shipment of a package from the central post office to the local post office closer to the recipient’s home. The convective impairment in O₂ transport (the ability to increase cardiac output, Qc) is typically impaired in patients with HFpEF.1–5

But that is not the end of the story (or the postal metaphor). After being ejected from the heart through the conduit and resistance arteries, the O₂-rich erythrocytes then need to be distributed to capillary beds in skeletal muscle, where O₂ will diffuse from haemoglobin molecule to skeletal myocyte to be utilised by mitochondria. This step in the O₂ pathway is analogous to transport of our package from the destination post office to the recipient’s doorstep (capillary), where it is then opened and read by the recipient (mitochondria). Just as there are central cardiac limitations in HFpEF, recent studies have revealed abnormalities in this more peripheral component of O₂ transport.3 One would expect that these impairments might require different treatments, but a simple, reliable, non-invasive way to identify patients with these deficits has been lacking.

In this issue of Heart, Hearon and colleagues report exciting new data that may help address this unmet need.6 The authors performed exercise testing in patients with HFpEF (n=18) and age/sex-matched healthy controls (n=18). Qc was assessed non-invasively by the acetylene rebreathe technique along with VO₂. The arterial–venous O₂ difference (a-vO₂ difference) was calculated noninvasively based on the Fick principle. Oxygen uptake kinetics were characterised by the mean response time (MRT) from continuously measured VO₂, where prolonged MRT connotes greater impairment in VO₂ kinetics, and accumulation of a greater O₂ deficit during exercise (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Schematic representation of oxygen consumption (VO₂) kinetics. VO₂ (red line) is plotted as a function of time during constant workload exercise. The mean response time (MRT) represents the time required for VO₂ to rise to ~63% of steady-state level at a given exercise workload. In health, VO₂ rises very rapidly during exercise, because both the convective and diffusive components of the O₂ transport pathway are able to respond rapidly and efficiently to increasing metabolic demand, with a pattern that more closely resembles an idealised square wave (upper left bounds). In patients with abnormal VO₂ kinetics, there is a progressive lag in the increase in VO₂, resulting in prolongation of the MRT and an increase in the gap between the square wave and the observed increase in VO₂ (red shaded area). This gap is termed the oxygen deficit, and may be caused by convective or diffusive abnormalities in O₂ transport.

The authors found that during submaximal exercise MRT was 60% longer in HFpEF as compared with age and sex-matched controls.6 There was no difference in Qc. The oxygen deficit was accordingly 66% higher in patients with HFpEF, which may lead to local energy depletion and accumulation of metabolic byproducts from anaerobic glycolysis that promote the perception of dyspnea and fatigue. In other words, the package was delivered (bulk O₂ transport or Qc was preserved), but for some reason the package either did not reach the mailbox (capillaries), or it reached the mailbox, but was not opened or read by the recipient (mitochondria).

The authors then performed an exploratory subanalysis comparing HFpEF patients with normal versus prolonged MRT (n=9 each).6 They found that during submaximal exercise, a-vO₂ difference was lower in the prolonged MRT HFpEF group. Arterial O₂ saturation was not reported, but if we assume that patients did not desaturate, the lower a-vO₂ difference would signify a higher O₂ saturation in the venous circulation. The blood returning to the right heart would therefore be brighter red, meaning that several of the packages (haemoglobin-bound O₂) were being ‘returned to sender’ (the heart) without being opened. In other words, the patients with HFpEF and prolonged MRT required more packages to be delivered for a given amount of mail to be read (higher Qc to achieve a given VO₂). As a consequence, the hearts of these patients must do more hydraulic work to achieve a given level of tissue-level O₂ delivery. This could be a big problem for a left ventricle that struggles to fill with and pump blood adequately, as is the case in HFpEF.2 4 5 Hearon and colleagues conclude that, on average, patients with HFpEF display slowed O₂ kinetics, and that use of MRT may be helpful improve individual patient phenotyping.6

The authors are to be commended for this outstanding study and important contribution to the field.6 While the sample size is small, the methods applied were incredibly rigorous, performed in one of the leading exercise physiology laboratories worldwide. There are, however, some points to discuss before we consider how to apply the results to the broader cohort of patients with HFpEF in the community. First, patients with atrial fibrillation and most patients with coronary artery disease were excluded.6 While this improves the homogeneity of the study cohort, it decreases the applicability of these data to the broader HFpEF population. Indeed, patients with atrial fibrillation (40% of all HFpEF) and coronary disease (65% of all HFpEF) typically display greater impairments in Qc reserve, and they more commonly present with right heart dysfunction, pulmonary hypertension and endothelial dysfunction.1 The current results may therefore not apply to this majority of patients.

Control participants were rigorously screened to exclude cardiovascular diseases. This resulted in a number important baseline differences that confound interpretation.6 The most striking of these is the dramatic difference in body mass, which may independently influence O₂ kinetics. Other comorbidities of interest such as diabetes and metabolic syndrome were not reported, but might be expected to be greater in the HFpEF group, given the mean body mass index of 35. Thus, we are left wondering whether part of the difference in VO₂ kinetics was ascribable to adiposity rather than HFpEF. Future studies should include a control population, that is, matched for these relevant comorbidities, particularly obesity and diabetes, which also influence peripheral O₂ delivery and metabolism.

In this study, Qc was similar in HFpEF and controls during exercise, which is in an important negative finding that suggests that the observed prolongation in MRT was related in these patients to peripheral deficits rather than the heart.6 The authors state in their discussion that Qc is uniformly preserved during submaximal exercise in HFpEF, but the literature shows that this is not the case; numerous studies have shown that submaximal Qc is reduced in HFpEF, even at very low workloads in the range of 20–40 W.1 2 4 5 Furthermore, the increase in Qc relative to VO₂ is not preserved in HFpEF, but rather reduced, when assessed using the gold standard invasive Fick method.2 4

In this regard, the 18 HFpEF patients enrolled in the current study sample with normal Qc during exercise were somewhat atypical, and differ from those we see in HFpEF patients with more severe cardiac limitations.1 2 4 5 However, this should not detract from the substantial clinical importance of the authors’ findings, because the combination of both impaired convective and diffusive O₂ delivery in patients with HFpEF and more severe central limitations would only be expected to cause greater O₂ deficit in the tissues, which would accordingly lead to more severe symptoms and functional impairments.

The most important and exciting clinical implication of these data centres on the potential ability to use VO₂ kinetics to non-invasively identify patients with greater peripheral deficits, who might then be treated differently, assuming we can come up with specific treatments. While the analysis packages to determine MRT are not currently part of the standard ‘plug-and-play’ clinical software, this could easily be added to what is obtained as part of standard exercise testing, and would require only minor modifications of currently used exercise protocols. However, before this becomes part of everyday practice, it will be essential to conduct future prospective studies in larger cohorts of patients to validate the authors’ findings, with broader representation of typical HFpEF patients, preferably utilising multicentre design and including those with less sophisticated exercise laboratories that reflect real world practice.

In summary, there are many steps in the O₂ transport pathway that become compromised in people with HFpEF. Hearon and colleagues have reminded us how the final steps in this pathway are often just as limiting as the more central steps closer to the heart, and in the process, they have elegantly demonstrated how we may be able to use VO₂ kinetics to help identify the patients where these peripheral steps become most compromised.6 If the mail is not being delivered, we need to know why, and the same is true if O₂ transport is constrained. It is hoped that with more robust phenotyping methods, using VO₂ kinetics and other novel methods, we will 1 day be in a position to better understand the nature of deficits in both the heart and periphery at the individual patient level, enabling better tailoring of therapies to improve clinical status for this large and ever expanding cohort.