Reduced arteriovenous oxygen difference in heart failure with preserved ejection fraction patients: Is the muscle oxidative phenotype certainly involved?

Heart failure (HF) is intrinsically associated with impaired functional exercise capacity, leading to substantial morbidity and mortality.1 Understanding the aetiology and pathophysiology of exercise intolerance (EI) in HF remains essential for the advent of novel therapies, particularly concerning ∼50% of patients presenting with HF with preserved ejection fraction (HFpEF), for whom no effective treatment has been found thus far.2 Peak oxygen (O2) consumption (VO2peak), as elicited by incremental dynamic exercise involving >50% of total muscle mass (treadmill, cycle ergometer), is considered a hallmark of exercise tolerance with high prognostic value.3 Theoretically, the impairment of any step of the O2 transport and utilisation chain could explain the limitation of VO2peak in HFpEF. An exhaustive scrutiny of the determinants of VO2peak is necessary to ascertain which mechanisms should be primarily targeted for therapy. Regarding HFpEF patients, markedly low VO2peak levels are consistently associated with diminished Qpeak and subsequent inability to deliver sufficient O2 to cope with increased metabolic demands.5–9 Moreover, anaemia, as diagnosed by haemoglobin (Hb) concentration affects ≥43% of HFpEF patients.10 Decreased total circulating red blood cell volume (RBCV) is observed in nine out of 10 HFpEF patients with Hb-based anaemia,11 hence presenting with a deficit in total blood O2 carrying capacity. Along with cardiac and haematological factors, a peripheral limitation in the form of reduced peak a-vO2diff is also apparent in a fraction of HFpEF studies.7–9 Importantly, reduced peak a-vO2diff has been portrayed as the dominant feature underlying EI in HFpEF8,9 with anomalies in skeletal muscle (SM) morphology and function as primary culprits.12 Furthermore, in recent exercise training (ET) guidelines for HF, it has been posited that local (knee extensor) exercise primarily targeting SM adaptations might be more effective to improve VO2peak than well-established aerobic ET involving large SM mass (e.g. walking, cycling).13 Given the relevance of the ‘peripheral’ paradigm for future ET prescription in HFpEF patients, we would like to ponder current evidence on the basis of sound integrative physiology with the attempt to facilitate a forthcoming robust understanding of EI in HFpEF. Peak a-vO2diff is commonly seen as a proxy of the capacity of SM to extract (and metabolise) O2 from the circulation. Variables facilitating O2 diffusion from capillary into SM fibres and O2 utilisation such as capillarisation and mitochondrial volume density (MitoVD) are frequently deemed to determine peak a-vO2diff.7,8 Support or refutation of this construct can be derived from classic experimental models delving into muscle physiology.14 For instance, substantial decreases in SM capillarisation, MitoVD and oxidative capacity are typically observed with prolonged bed rest in healthy individuals.15 Notwithstanding, peak a-vO2diff remains unaltered after bed rest.15,16 This apparently counterintuitive outcome could be related to the fact that SM mitochondrial oxidative capacity largely exceeds O2 delivery at VO2peak in healthy humans.17,18 Likewise, recent catheterisation studies demonstrate a two-fold functional reserve in SM O2 extraction at VO2peak in untrained individuals.19 In this line, manipulation of murine SM mitochondrial content (–30% relative to normal levels) and oxidative capacity (–60% to −85%) did not reduce VO2peak, whereas anaemia did so.20 Human and animal evidence thus coincide in that partially impaired SM oxidative phenotype may not limit peak a-vO2diff, which denotes certain ‘muscle surplus’ during large SM mass exercise, known to be primarily limited by the convective O2 delivery capacity.18,21 Reduced SM oxidative capacity might however affect O2 extraction with exercise involving smaller SM mass (<50% of total SM mass), thus less constrained by the finite O2 delivery, such as two-arm cranking or one-leg knee extension in that in vivo O2 consumption reaches SM mitochondrial oxidative capacity measured ex vivo.17 Nonetheless, as concerns the established assessment of exercise capacity by means of large SM exercise, alterations in SM morphology and/or metabolic capacity should not unequivocally be taken to affect peak a-vO2diff. Any form of HF is partly characterised by physical inactivity. Accordingly, SM abnormalities may be present in HFpEF, plausibly attributed to physical deconditioning rather than disease-specific myopathy.22 Pioneering biopsy studies assessing SM phenotype in HFpEF patients have recently been performed.23–25 Herein, Molina et al. found lower (up to −46%) SM mitochondrial content and oxidative capacity compared with healthy age-matched controls.25 The study subjects of this investigation were not matched by physical activity levels, body size or composition, HFpEF patients presenting higher body mass index (BMI) (+33%) and percentage body fat (+43%) relative to controls.25 Indeed, the group difference in SM mitochondrial content vanished when adjusted by BMI.25 This concurs with the dominant role of body size on SM phenotype comparisons,26 and suggests that the difference in mitochondrial content is body size- or adiposity-dependent.25 The same research group has also reported elevated BMI-dependent intermuscular fat23 as well as lower percentage of type 1 fibres,24 possibly curtailing the oxidative machinery’s potential, in HFpEF vs age- but not BMI-matched control individuals. Regardless, if exercise tolerance in HFpEF were limited by any SM variable, significant relationships with peak a-vO2diff/VO2peak should emerge. While a-vO2diff was not measured in the aforementioned studies,23–25 it is noteworthy that no SM variable was associated with VO2peak in the HFpEF cohort alone.23-25 In contrast, VO2peak is a linear function of variables underlying convective O2 delivery such as maximal heart rate or stroke volume in HFpEF,5–9 concurring with the predominant central limitation of VO2peak in sedentary individuals.4 It yet remains unresolved whether SM divergences from the healthy phenotype (primarily) determine EI in HFpEF patients. While leg VO2 can be measured de facto, SMΔO2 is estimated in a convoluted manner.29 Mean capillary O2 pressure is approximated by a forward integration procedure using arterial and femoral venous O2 pressure on the premise that perfusion/VO2 heterogeneity and perfusional or diffusional shunt are negligible in SM and adjacent tissues28 – a currently untenable assumption.30,31 In addition, DMO2 is determined by leg blood flow (LBF), which indeed may be substantially impaired (up to −40%) in HFpEF patients.32 Reasonably, DMO2 cannot be directly determined when LBF and leg O2 delivery is limited by the pumping capacity of the failing heart. Forthcoming studies assessing DMO2 might consider the experimental manipulation of leg O2 delivery, e.g. via altering blood O2 carrying capacity while preserving arterial O2 partial pressure, in control individuals relative to HFpEF patients.33 Collectively considered, DMO2, as currently determined, is a lumped parameter that does not dissociate strict SM O2 diffusing capacity from blood flow distribution (BFD) and convective O2 delivery. Ideally, any contribution of SM oxidative phenotype to O2 extraction should be pinpointed using a-vO2diff measures across active SM. However, all the studies describing a reduced peak a-vO2diff in HFpEF patients gauged arterial and mixed venous blood or estimated a-vO2diff from the Fick equation7–9 based on VO2peak and Qpeak measurements, where the latter presents low reliability if determined non-invasively.36 Results derived from mixed venous blood or indirect estimations offer less certainty than invasive investigations sampling venous blood draining SM. Intriguingly, to date the only catheterisation study that has sampled blood from the deep femoral vein, thus maximising the proportion of analysed blood irrigating active SM, revealed near-complete O2 extraction at VO2peak in HFpEF patients, which even exceeded that of age-matched controls.37 This suggests that systemic peak a-vO2diff may be dissociated from SM a-vO2diff in HFpEF patients due to impaired BFD, a recognised trait in older individuals.34 The established age- and HFpEF-related decline of leg vascular conductance secondary to augmented muscle sympathetic nerve activity,38–40 may reduce LBF and thereby systemic peak a-vO2diff, independently of alterations in the SM oxidative phenotype. An additional methodological issue concerning current evidence of reduced peak a-vO2diff in HFpEF patients is the inclusion of a higher percentage of females in HFpEF versus control groups,7–9 a problematic aspect given the inherent impact of female lower blood O2 carrying capacity on a-vO2diff.41 Ultimately, until more (gender-matched) evidence becomes available providing direct assessments of SM O2 extraction at VO2peak, the ‘peripheral’ paradigm of EI in HFpEF, as founded upon SM abnormalities, will remain a compelling, but unproven, hypothesis. DM conceived and drafted the manuscript. CL critically revised the manuscript. Both authors gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article. The author(s) received no financial support for the research, authorship and/or publication of this article.