Finding the Road to Recovery: Therapeutic and Clinical Trial Implications of Dysfunctional Viable Myocardium in Heart Failure with Reduced Ejection Fraction

This article refers to ‘Multidimensional contractile reserve predicts adverse outcome in patients with severe systolic heart failure: a 4-year follow-up study’, by I.A. Paraskevaidis et al. and ‘The importance of myocardial contractile reserve in predicting response to cardiac resynchronization therapy’ by M. Kloosterman et al., published in this issue on pages 846–861 and 862–869. Despite the discovery of multiple evidence-based therapies, heart failure with reduced ejection fraction (HFrEF) continues to carry mortality and morbidity comparable to many cancers. However, the HFrEF population is heterogeneous, and individual patient outcomes may vary widely. Once HFrEF is diagnosed, barring sudden death, any of three clinical trajectories are possible: stabilization, acceleration to end-stage heart failure (HF) or death, or structural and functional cardiac recovery. A central goal of guideline-directed therapy for HFrEF is to maximize chances of myocardial recovery and to halt, or at least slow, HF progression. In this context, for decades, dobutamine stress echocardiography (DSE) has been incorporated into HFrEF patient care to predict likelihood of cardiac recovery. In contemporary practice, DSE protocols may be used to assess myocardial viability and contractile reserve (CR), the latter traditionally defined by the degree of stroke volume augmentation. However, while different in execution and interpretation, both viability and CR studies serve the shared overarching purpose of quantifying the degree of dysfunctional myocardial tissue that remains capable of functional improvement (i.e. ‘myocardial reserve’). Traditionally, this information has been used to guide decisions for coronary revascularization or aortic valve replacement.1, 2 Given that such procedures are invasive, costly, and carry exaggerated risks in the setting of depressed left ventricular ejection fraction (LVEF), the presence of adequate viability or CR may signal improved likelihood of clinical benefit and recovery of systolic function. Likewise, lack of this myocardial reserve may fail to justify procedural risks and thus favour a non-invasive medical or palliative approach. In this issue, we find two articles that expand existing perspective on the prognostic and therapeutic implications of myocardial viability and CR in HFrEF.3, 4 Herein, we explore the role of dysfunctional viable (DV) myocardium in HFrEF with respect to prognostication, prediction of treatment response, and enrichment of clinical trial selection criteria. In this issue, Paraskevaidis et al. present a prospective observational study evaluating DSE-related changes in echocardiographic speckle tracking indices and subsequent cardiac mortality.3 Radial strain at rest and dobutamine-induced change in global longitudinal strain (GLS) demonstrated independent associations with cardiac death after adjustment for conventional clinical and echocardiographic factors. Likewise, incorporation of both speckle parameters within multivariable predictive models yielded significant improvements in model performance beyond that seen with traditionally reported dobutamine-induced echocardiographic changes, including surrogates for stroke volume. Moreover, among the subset of patients surviving through study follow-up, both resting radial strain and GLS change with dobutamine were predictive of subsequent cardiac recovery. Notably, a relative change from baseline in GLS of >19% was associated with a mean increase in LVEF of 14% at an approximately 4-year follow-up, whereas lesser relative change in GLS with dobutamine was associated with a small continued decline in LVEF. Despite elegant statistical design and sound study execution, several limitations of these data must be recognized. First, small overall sample size limits these findings to hypothesis generating only. Further studies must externally validate the incremental prognostic value of resting and dobutamine-induced changes in speckle tracking parameters beyond traditional, and perhaps more easily attainable, clinical and echocardiographic factors. Second, this study examined an exclusively New York Heart Association class III/IV cohort with many features suggestive of advanced HFrEF, including a resting systolic blood pressure of ∼100 mmHg and left ventricular end-diastolic dimension >70 mm. Accordingly, these findings may not be generalizable to populations with less severe HFrEF. Third, data regarding baseline and follow-up medical therapy were not available and imbalances in medication use could confound outcome results. Fourth, this was a single-centre study utilizing the expertise of an academic echocardiographic laboratory. Issues with measurement accuracy, reproducibility, and availability may represent challenges at less experienced centres. Likewise, although low-dose DSE used at this study centre represents a common approach, these findings may not extend to higher dose dobutamine protocols. Paraskevaidis and colleagues add to the impressive body of literature correlating myocardial reserve with prognosis using a multitude of imaging modalities and parameters.3, 5-7 While classically associated with ischaemic cardiomyopathy to reflect ‘hibernating’ or ‘stunned’ myocardium secondary to coronary artery stenosis, the term DV myocardium is now understood to extend to non-ischaemic substrates. Indeed, the study by Paraskevaidis et al. included a nearly even mix of ischaemic and non-ischaemic cardiomyopathy patients, reiterating that evaluation of myocardial reserve need not be guided by the presence of coronary artery disease.3 Rather, a broader definition of DV myocardium calls for hypo- or akinetic tissue free of replacement fibrosis that retains intact cellular metabolism and membranes, with or without overlying ischaemia.8 Conceptually, while existence of viable myocardium does not ensure recoverability, the ability of a chronic HFrEF patient to recover is presumably contingent on its presence. Although the true prevalence and quantification of DV myocardium within the general HFrEF population have not been systematically studied, clinical characteristics, such as robust systolic blood pressure and rapid response to diuretics, suggest many patients presenting with worsening symptoms do not have an end-stage phenotype. In this sense, HF contrasts with other common examples of organ ‘failure’ such as chronic kidney disease or cirrhosis, where residual viable tissue is minimal and fibrosis predominates at the point of clinical organ failure.9 Thus, while mortality rates for HFrEF remain high, one can postulate that many patients die with ‘tissue capital’ that was not adequately revived.9 Existing evidence suggests that HFrEF patients with DV myocardium exhibit more robust response to guideline-directed medical and device-based therapy. For example, both CR by DSE and viability mapping on gadolinium-enhanced cardiac magnetic resonance (CMR) imaging predict likelihood of left ventricular recovery with beta-blocker therapy.10, 11 However, while the relationship between myocardial reserve and therapeutic response may inform medical management in some cases, the practical application of this relationship may be particularly evident in decisions related to device implantation, where procedural risks and cost-effectiveness present added considerations. Although cardiac resynchronization therapy (CRT) is of proven value in appropriately selected HFrEF populations, response to therapy is heterogeneous and many patients may derive no benefit despite meeting guideline indications. Thus, there remains significant interest in refined approaches to prospectively identify ‘responders’. Multiple studies have assessed the role of myocardial viability in predicting CRT response, largely concluding that extent of DV myocardium, as well as the extent and location of scar tissue, are helpful in further individualizing care decisions.12, 13 Likewise, multiple studies have evaluated CR by DSE, but with generally small sample size and variable definitions of CR and CRT response. In this issue, Kloosterman et al. present pooled data from such studies and provide added reassurance regarding the utility of CR in predicting response to CRT, irrespective of ischaemic or non-ischaemic status.4 While the authors correctly acknowledge that, given the lack of large definitive trials, patients should not be denied CRT based solely on absence of CR, this meta-analysis does support CR as an additional data point worth considering in the setting of an unclear risk–benefit ratio for an individual patient. The significant heterogeneity of the hospitalized and ambulatory HFrEF populations has become increasingly recognized as a key challenge to drug and device development programmes. Moving forward, optimizing clinical trial success may rely on matching the ‘right’ therapy to the ‘right’ subset of patients most likely to benefit from the therapeutic mechanism.14, 15 Rather than a traditional early phase approach including a generic HFrEF population and generic surrogate endpoints (e.g. change in natriuretic peptide level), use of a ‘mechanistic translational phase’ can be considered to facilitate improved understanding of a therapy's effects on the heart in a small homogeneous population.16 While the exact characteristics that define such a test cohort will vary according to the investigational agent, it stands to reason that a common foundation may include the presence of ample viable myocardium. Further relevant to clinical trials is the recognition that, in the absence of epicardial coronary stenosis, DV myocardium may exist for a variety of reasons. Problems can occur with any of the myocardial components, including the myocyte, interstitium, or microcirculation, as well as cellular metabolism and mitochondrial function.17 Discerning the exact mechanism underlying DV myocardium may optimize therapeutic targeting. For example, a drug targeting endothelial function may be best studied in a population with DV myocardium driven by microcirculatory impairment. Although the execution of such precise ‘matchmaking’ between patient and investigational therapy remains to be seen, advances in cardiac imaging may offer the future possibility of such a strategy. CMR, in particular, can offer granular detail on myocardial structure and function. Relevant applications of CMR may include detection of replacement fibrosis or non-viable ‘scar’ with gadolinium enhancement, quantification of interstitial fibrosis with T1 mapping, and diagnosis of microcirculatory dysfunction with perfusion imaging.17, 18 A recent phase IIA study of mesenchymal stem cells in non-ischaemic cardiomyopathy patients may serve as a preliminary example for how CMR and DV myocardium assessment may be incorporated within an early phase clinical trial.19 To enrich the study population with HFrEF patients deemed most likely to benefit, the protocol screened patients with delayed-enhancement CMR imaging to exclude those with any evidence of non-viable replacement scarring.19 Such an approach may become increasingly practical with the advent of CMR-compatible pacemakers and defibrillators. In aggregate, the two analyses presented in this issue of the journal reiterate the critical importance of myocardial reserve in informing patient prognosis and response to HFrEF therapy. Traditional and novel applications of DSE and other imaging modalities have allowed for greater capture of information relevant to individualized patient care. Moreover, these articles add to the compelling case favouring DV myocardium as the optimal cardiac substrate for early phase clinical trial testing of novel HFrEF therapeutics. As such, increasing consideration of myocardial reserve assessments in patient-specific care decisions and early phase clinical trial selection criteria is warranted. While population-level outcomes for HFrEF continue to be poor, we remain optimistic that the presence of DV myocardium signals that HFrEF need not always be a progressive condition, but one with the hope of recovery. Conflict of interest: S.J.G. and M.V. report no conflicts. M.G. has been a consultant for Abbott Laboratories, Astellas, AstraZeneca, Bayer HealthCare AG, CorThera, Cytokinetics, DebioPharm S.A., Errekappa Terapeutici, GlaxoSmithKline, Ikaria, Johnson & Johnson, Medtronic, Merck, Novartis Pharma AG, Otsuka Pharmaceuticals, Palatin Technologies, Pericor Therapeutics, Protein Design Laboratories, Sanofi-Aventis, Sigma Tau, Solvay Pharmaceuticals, Takeda Pharmaceutical, and Trevena Therapeutics.