Editor’s Choice- Pathophysiology and therapy of myocardial ischaemia/reperfusion syndrome

The epidemiological perspective

The implementation of reperfusion and adjuvant pharmacological (acute and maintenance) therapy has resulted in an impressive improvement in the prognosis of STEMI patients in western countries.^1,2 Paradoxically, while inhospital death rates for STEMI have substantially dropped over the past few decades, ³ there is an inversely proportional increase of chronic heart failure (CHF) rates as a result. Patients with severely depressed hearts after AMI would formerly die during hospitalisation, but currently many of them survive at the cost of a high number of subsequent long-term complications. The incidence of AMI is growing disproportionately in some geographical regions due to the increased cardiovascular risk burden, turning the incidence of STEMI into a major health problem in developing countries.^4,5 Either because of the increase in mortality rates associated with the growing incidence or because of the increase in morbidity and the socioeconomic burden associated with the development of CHF in STEMI survivors,^6,7 there is a worldwide need to reduce the life-long impact of AMI.

Besides the necessity to reduce and treat cardiovascular risk factors in the long term, the implementation of timely reperfusion is currently the cornerstone therapy to improve mortality and morbidity in AMI patients. ² It is not debatable that the shortening of symptom-to-reperfusion time results in better clinical outcomes (‘time is muscle’),^2,8,9 and there is in many ways still room to improve both the implementation and the timing of primary percutaneous coronary intervention (PPCI; the best reperfusion technique). ¹⁰ However, in some regions, this strategy seems to have reached its own ceiling and, despite having an efficient early PPCI programme, one-year mortality remains excessively high, reaching 15% in some countries. ¹¹ In regions where the time from STEMI diagnosis to PPCI-mediated reperfusion has been shortened to about 90 minutes, further shortening of this time has not been demonstrated to improve mortality. ¹² While it is true that sicker patients now undergo PPCI and this might contribute to the plateau in mortality despite better response times, ¹³ these figures highlight the need for novel therapies to be administered as adjuncts to PPCI in order both to improve patient survival and prevent the onset of heart failure.

The hypothesis that a reduction in myocardial IS translates into an improvement in clinical outcomes has recently been demonstrated in a patient-level meta-analysis, in which a steep gradient was found in the combined endpoint of all-cause mortality and hospitalisation for heart failure at one year across quartiles of IS, with 8.8% of events for the top quartile and 1.2% for the bottom quartile.^14,15 Because timely and complete reperfusion is the most well-established way of limiting IS and the subsequent ventricular remodelling, the focus of this review will be on those cardioprotective interventions beyond reperfusion.

The clinical perspective

The beginning of the reperfusion era in AMI patients starts in 1972, with the seminal studies by Ross and co-workers, who demonstrated in a dog model that reperfusion after 3 hours’ coronary occlusion limited the progression of necrosis.^16–18 Soon afterwards Braunwald and Kloner popularised the concept that reperfusion itself induces additional injury in the vulnerable ischaemic tissue. ¹⁹ Since then, it is well accepted that the extension of irreversibly injured myocardium (i.e. IS) is the result of ischaemia and reperfusion-related damage, thus the term ‘I/R injury’. After several decades of intense research on the topic, it is still debated as to what is the relative contribution of ischaemia and reperfusion-related injuries to the final IS. One reason for this uncertainty is the close interplay between both types of injury: as the degree of ischaemic injury at the time of blood flow restoration is a main factor contributing to reperfusion-related damage, it is impossible to separate the two types of damage. Another factor contributing to this vagueness is the absence of tools to differentiate ischaemia and reperfusion-related injuries. Cardiac magnetic resonance (CMR) imaging is able to visualise some features of reperfusion-related injury, such as oedema formation or microvascular obstruction,^20–24 but others such as cardiomyocyte mitochondrial damage are not amenable to CMR imaging at present.

The GISSI and ISIS-2 trials concurrently demonstrated that reperfusion through intravenous thrombolysis was not just feasible, ²⁵ but was also able to improve clinical outcomes in STEMI patients.^26,27 Mechanical reperfusion using PPCI was later demonstrated to be superior to thrombolysis and was thus the preferred mode of reperfusion, if available in a timely fashion. ²⁸ However, while early reperfusion therapy solves part of the problem (ischaemia-related injury), the other part (reperfusion-related injury) still remains to be clinically addressed and has become one of the top 10 unmet clinical needs in cardiology. ²⁹ One clinical observation led to a change in the way researchers envisioned damage associated with AMI: patients with pre-infarction angina had a better prognosis than patients whose first symptom was at the index AMI. ³⁰ This clinical observation was experimentally mimicked in dogs by Murry et al. ³¹ by inducing brief periods of coronary occlusion and reperfusion before prolonged occlusion (i.e. index AMI insult), something known as ischaemic preconditioning (IPC). Preconditioning resulted in massive reductions in IS. This simple experiment, emulating the clinical scenario, demonstrated that, beyond early reperfusion, an intervention could reduce IS. IPC has been shown to protect when applied in the actual heart, as in the seminal experiment by Murry et al., ³¹ and also when applied in a distant organ (something known as remote ischaemic conditioning; RIC). In addition, it has been shown that conditioning stimulus can be applied at the end of the ischaemic duration (post-conditioning) and even at the middle of it in a remote organ (remote per-conditioning).

The experimental perspective

The impact of myocardial reperfusion at the histological level was first postulated in 1960 by Jennings et al. in a landmark study describing pathology features of the reperfused ischaemic canine myocardium, such cell swelling, contracture of myofibrils, disruption of the sarcolemma and the appearance of intra-mitochondrial calcium phosphate particles. ³² Later, the concept of reperfusion injury was further expanded to other manifestations, ³³ such as myocardial stunning, reperfusion arrhythmias and the no-reflow phenomenon related to microvascular damage. However, the reperfusion-triggered death of cardiomyocytes that have been reversibly injured during ischaemia has been considered the paradigm target for all developing cardioprotective therapies during past decades. This phenomenon (known as lethal reperfusion injury) is driven by a cytosolic and mitochondrial calcium overload, oxidative stress and rapid restoration of intracellular pH resulting in the opening of the mitochondrial permeability transition pore (mPTP). Researchers have focused on investigating molecular targets in the cardiomyocyte, under the hypothesis that their activation at reperfusion onset ameliorates the impact of myocardial I/R, reducing the resultant IS (and subsequently having an improvement in clinical outcomes).

Cardiomyocyte death: the end stage of I/R injury

The heart is a contractile organ, with cardiomyocytes being the contractile units. Cardiomyocytes are thus at the centre of the heart’s function. Loss of contractile units leads to failure of cardiac function. As mentioned before, cardiomyocytes represent up to 80% of heart volume and thus the loss of these cells has a significant impact on cardiac function. Despite I/R injury being a multi-compartment syndrome, cardiomyocytes are at the end of the chain of events. For this reason, most of the attempts to find therapies able to reduce I/R injury have been focused on cardiomyocytes. Deciphering the intra-cardiomyocyte events occurring during I/R injury have led to the identification of potential targets to protect these cells. Experimental studies over the past three decades have identified a complex signalling map within the cardiomyocyte to explain how most cardioprotective therapies exert their protective effect against I/R injury. ³⁷ These findings have been first described to elucidate the mechanism underlying some forms of ischaemic conditioning, but they have been subsequently extended to most pharmacological interventions. Briefly, there is a consensus to recognise three hierarchical levels of signal transduction: ³⁸ (a) triggers (usually sarcolemmal membrane receptors, such as G-protein coupled receptors and tyrosine kinase receptors); (b) intracellular mediators (the signalling cascades that help initiate and propagate the signal); and (c) end-effectors (mechanisms that actually cause the attenuation of cellular injury and death during the lethal ischaemic insult). This sequential three-step mechanism can be easily illustrated with a well-known cardioprotective therapy: insulin (the trigger) activates PI3Kα, which in turn recruits Akt and its downstream cascade (mediators) to end up inhibiting the mPTP (the end-effector), which is in the last link of the chain to avoid cardiomyocyte death. ³⁹

There have been attempts to manipulate these three levels of signal transduction pharmacologically. At the trigger level, insulin or adenosine have been used to activate the cardioprotective response. ³⁹ At mediator level, three pro-survival cascades have received special attention:^38,40 the reperfusion injury salvage kinase (RISK) pathway (comprising PI3K-Akt and MEK1/2-ERK1/2),^41–43 the survivor activator factor enhancement (SAFE) pathways (comprising TNFα and JAK-STAT3) ⁴⁴ and the PKG/eNOS signalling cascade. ⁴⁵ Importantly, these pro-survival pathways are activated at the onset of reperfusion,^46,47 hence getting a translational value if they can be activated either in the ambulance or the cath lab. Finally, at end-effector level, all the three pro-survival cascades seem to converge in the mitochondria. ⁴⁸

Non-cardiomyocyte compartments involved in I/R injury

While cardiomyocytes are the end stage of I/R injury, there are several processes occurring upstream that contribute to the loss of contractile units. No matter how ‘strong’ is the cardiomyocyte (e.g. from the intracellular signalling pathway perspective) if there is no efficient tissue perfusion due to microvascular damage the cell will not survive the episode. Similarly, a cardiomyocyte able to survive the ischaemic insult that is surrounded by highly activated neutrophils of other inflammatory cell mediators is at high risk of dying hours/days after reperfusion. Recent evidence has shown that therapies targeting non-cardiomyocyte compartments can be efficient in reducing I/R injury and ultimately IS. ³⁶

Due to the high metabolic demand, the heart has an extensive microvascular blood supply system. Some advocate that the endothelium might have a relevant role in cardioprotection due both to its optimal situation to interact with blood signals and its paracrine ability. ⁴⁹ As first point of contact between the myocardium and humoral factors, the endothelium constitutes a ‘blood–heart barrier’. ⁵⁰ There is some evidence demonstrating that the effluent collected from preconditioned endothelial cells is able to provide some protection against I/R in naive primary cardiomyocytes. ⁵¹ Similarly, Teng et al. demonstrated with a transgenic mouse model, which restricted the expression of the erythropoietin (EPO) receptor to haematopoietic and endothelial cells, that the administration of EPO in these mice can protect the heart against I/R, therefore suggesting a major role for the endothelial cell response to EPO to achieve an acute infarct-limiting effect. ⁵² As a paracrine organ, the endothelium has been demonstrated to trigger protection in cardiomyocytes through receptor/ligand interaction and gaseotransmitter. The endothelin-1 (ET1) receptor and bradykynin B(2) receptor are both present in cardiomyocytes,^53,54 and when pharmacologically activated both trigger a preconditioning-like effect.^53–55 With regard to gaseous signals, nitric oxide has long been associated with ischaemic conditioning through the role of eNOS (the endothelial isoform of nitric oxide synthase), as demonstrated in eNOS knockout mice; ⁵⁶ however, therapies with inhaled nitric oxide have failed in their translation to the clinical arena. ⁵⁷

Besides being a provider of protective triggers and activated mediators to cardiomyocytes, there is also the possibility for the endothelium to be a target itself for cardioprotection. ⁴⁹ Most signalling pathways and end-effector mechanisms described in cardioprotection are most likely not specific to cardiomyocytes. Some publications have shown a higher vulnerability to I/R of the endothelium when compared to cardiomyocytes.^58,59 Targeting endothelial receptors through adenosine agonists A1 and A3 or angiotensin II ⁶⁰ preserve not only endothelium-dependent vasodilation but also cardiomyocyte viability. Indeed, preserving microvascular function will provide further blood supply to the injured cardiomyocytes. In the clinical setting, pre-infarct angina (a preconditioning clinical manifestation) has been associated with the attenuation of no-reflow in STEMI patients undergoing PPCI. ⁶¹

Other cells can also be relevant in cardioprotection, such as platelets, in which recent experimental data have demonstrated that P2Y12 inhibitors are protective at the onset of reperfusion through RISK activation,^62,63 or neutrophils, that can be targeted by metoprolol to inhibit neutrophil–platelet interaction. ³⁶

Mechanical interventions

The phenomenon whereby the myocardium can endogenously be protected from lethal I/R was firstly reported by Murry et al. in 1986. ³¹ In that landmark study, the myocardial IS reduction obtained from the application of several brief cycles of non-injurious ischaemia and reperfusion before the subsequent sustained index ischaemic insult was given the term ‘ischaemic preconditioning’ (IPC). ³¹ This finding, first described in dogs but subsequently replicated in numerous pre-clinical models, ⁶⁴ became the cornerstone of the field for two main reasons: (a) the concept of local IPC evolved into ‘ischaemic conditioning’, a broader term that encompasses a number of related endogenous cardioprotective strategies, applied either to the heart (IPC or postconditioning) ⁶⁵ or to a distant organ (remote ischaemic pre, per or postconditioning);^65,66 (2) its underlying signalling architecture has been extrapolated to several cardioprotective therapies and has helped to identify molecular targets amenable to pharmacological modulation. Table 1 illustrates the landmark studies in I/R and conditioning-related cardioprotective therapies in chronological order.

OPEN IN VIEWER

Table 1.

Landmark studies in I/R and conditioning-related cardioprotective therapies.

Year	Author	Achievement
1972	Maroko et al. 16	First evidence that reperfusion limits extent of necrosis
1977	Reimer et al. 126	Description of the wave front progression of necrosis (from endocardium to epicardium)
1986	Murry et al. 31	First evidence that ischaemic preconditioning reduces IS (first window of protection)
1993	Marber et al. 127	Description of the second window of protection
1993	Przyklenk et al. 128	First evidence that remote ischaemic preconditioning within the heart reduces IS
1997	Birnbaum et al. 129	First evidence that the application of non-coronary remote conditioning confers protection to the heart
2002	Schulman et al. 41	Description of the reperfusion injury salvage kinase (RISK) pathway
2003	Zhao et al. 130	First evidence that ischaemic postconditioning reduces IS
2005	Lecour et al. 131	First evidence of a RISK-independent signalling pathway, later on coined as the survivor activator factor enhancement (SAFE) pathway
2010	Bøtker et al. 114	First evidence that remote ischaemic perconditioning increases myocardial salvage in man
2017	García-Prieto et al. 36	First demonstration that the β1 selective blocker metoprolol reduces MVO and IS by targeting neutrophils
2019	CONDI2/ERIC-PPCI study 73	Ongoing randomised clinical trial involving 5400 STEMI patients undergoing PPCI assessing improvement in long-term clinical outcomes following the application RIC

I/R: ischaemia/reperfusion injury; IS: infarct size; mPTP: mitochondrial permeability transition pore; MVO: microvascular obstruction; PPCI: primary percutaneous coronary intervention; RIC: remote ischaemic conditioning; STEMI: ST-segment elevation myocardial infarction.

The translational potential of local IPC is inevitably limited by the necessity to apply the intervention before the index ischaemia, which is unpredictable in many clinical scenarios such as STEMI. Ischaemic postconditioning has already been tested in the clinical setting, with mixed results in proof-of-concept studies^22,67,68 but overall disappointing findings when assessing hard clinical outcomes. ⁶⁹ On the contrary, RIC has emerged as a non-invasive alternative that can be applied either before (preconditioning), during (perconditioning) or after (postconditioning) index ischaemia. ⁷⁰ Overall, large clinical trials assessing the impact of RIC on hard endpoints in the context of cardiac surgery have been disappointing^71,72 – most likely because of the small amount of myocardium at risk, the short ischaemic time, the routine use of cardioplegia and hypothermia and the type of death in this clinical scenario (more related to surgical complications than to the myocardial injury). In the end, cardiac surgery-induced infarct myocardial damage is no longer a relevant problem with current approaches. However, it is acknowledged that the pathophysiology underlying STEMI is completely different and there are great expectations for the outcomes of two large ongoing clinical trials which have been combined into a single one, namely the CONDI2/ERIC-PPCI study ⁷³ – that trial has already finished the recruitment of 5400 STEMI patients undergoing PPCI in Europe (Denmark, UK, Spain and Serbia). Patients were randomly assigned to RIC or control, and the primary outcome is the composite of cardiac death and heart failure hospitalisation over 12 months’ follow-up. The results are expected to be reported late in 2019.

The application of high mechanical index impulses using a regular clinically available ultrasound transducer during commercially available intravenous micobubble infusion (so-called sonothrombolysis) has been shown to be effective in reducing cardiac injury in STEMI patients. ⁷⁴ The recent and promissing MRUSMI trial concluded that sonothrombolysis has an important role restoring epicardial flow and reducing IS. ⁷⁴ This effect is probably due to the microbubble growth and collapse produced by the ultrasound stimulation. This mechanical intervention promotes cavitation forces capable of thrombus dissolution. ⁷⁵

Pharmacological strategies

Several pharmacological strategies have been tested in experimental and pilot clinical trials with promising results. ⁷⁶ Many of them have failed as they progressed in the clinical arena. We are focusing this section on the therapies that are still promising and continue in the race to find a robust cardioprotective agent. We also briefly comment on therapies that were highly promising but withdrew from this aim.

Promising agents not fulfilling expectations

From the long list of failures, some have been especially painful because they were preceded by solid experimental and pilot clinical experiences.^9,76 Some of them are presented here for the sake of historical perspective.

Cardioprotection: an unsuccessful translational story to date

Many cardioprotective therapies aimed at reducing I/R have been successfully evaluated in the experimental setting. It is important to remark that not all these experimental experiences were performed with rigor according to accepted preclinical guidelines. ¹⁰⁴ Despite attenuating I/R at the bench, not all of them have subsequently demonstrated an IS-limiting effect at the bedside, and none to date have demonstrated clear benefits in terms of mortality or heart failure rehospitalisation. ¹⁰⁵ The reasons for these disappointing translations from experimental and proof-of-concept trials to clinical practice have been widely discussed elsewhere and are a matter of intense debate.^93,105–107 They are briefly summarised in Table 2. One common denominator of many of the failures is that the experimental setting where the therapies were proved beneficial are very different from the clinical scenario in which they were validated. As an illustrative example, therapies targeting mPTP should reach the cardiomyocyte before reperfusion or at immediate reperfusion at the latest. This is achievable at the experimental level but not at the clinical level. It has recently been shown that the administration of a bloodless oxygen-carrying solution (enriched with other nutrients) can stop ongoing necrosis without immediate blood flow restoration. ¹⁰⁸ This strategy allows the total control of conditions (temperature, pH, absence of inflammatory cells or mediators, etc.) at early stages of reperfusion. This platform could serve to test some of these agents that are only beneficial if they reach the myocardium before full-blown reperfusion conditions occur.

OPEN IN VIEWER

Table 2.

Potential reasons explaining the translational failure of cardioprotective therapies.

1	Differences between AMI experimental models and patients with STEMI (i.e. atherothrombosis vs. artery ligation)
2	Differences in animal physiology (i.e. a different IPC effect size gradient was observed across species in the CAESAR initiative) 132
3	Reductionist approaches for experimental designs to an increasingly complex clinical situation (i.e. single interventions are tested without the background of comorbidities or other therapies) 104
4	Performance of proof-of-concept clinical trials without solid experimental evidence or large-animal data 9
5	Lack of reproducibility and poor reporting133
6	Lack of ‘strict’ validation of surrogate endpoint in clinical trials 102
7	Randomised clinical trial design: eligibility criteria for patient selection (i.e. anterior infarct would benefit the most), misuse of some composite endpoints

AMI: acute myocardial infarction; CAESAR: Consortium for preclinicAl assESsment of cARdioprotective therapies; IPC: ischaemic preconditioning; STEMI: ST-segment elevation myocardial infarction.

The epidemiological perspective

The clinical perspective

The experimental perspective

New targets

There are some novel therapeutic targets that are currently under intense investigation, such as the immune system (monocytes, macrophages, extracellular DNA and RNA, inflammasomes), platelets – inflammatory cell interactions, exosomes and microvesicles, G-protein coupled receptors, Toll-like receptors and proteases such as matrix metalloproteinases and calpains among others, with especial emphasis on mitochondria. ¹²³ A large number of functional mitochondria are needed for the myocardium to keep producing the necessary adenosine triphosphate for a regular contractile function, in addition to participating in other basic biological processes, such as ion homeostasis and calcium exchange. ¹²⁴ Among all the events triggered by I/R injury, mitochondrial disturbances play a major role, probably being the final confluence of many simultaneous pathways, as happens with ischaemic conditioning cascades finishing in the mitochondria as an end-effector. ¹²⁴ Some drugs directly targeting the mitochondria have been tested in myocardial I/R injury, such as the mPTP inhibitor cyclosporine A, although they have failed in their final clinical translation to date. ⁴⁸ Despite these initial disappointing results, mitochondria still remain an attractive target in AMI, and new approaches using drugs aimed to modify mitochondrial dynamics (fission and fusion) ¹²⁵ or even transplantating complete organelles have recently been proposed. ¹²⁶

The complexity of the underlying pathophysiological processes in the myocardium subjected to I/R injury, as well the different cell types involved and the unknown impact of time over both components of the injury suggest that mimicking a single pathway is probably too simplistic an approach. Therefore, the synergistic application of more than one therapy either acting on different processes involved in the I/R injury, or just adding further effect into the same biological process is getting popular in some laboratories – i.e. in a pig AMI model, the combination of RIC with either GIK or exenatide at the time of reperfusion has been shown to reduce myocardial IS to a greater extent than either intervention alone. ¹²⁷ The rationale for using combination therapies has been expanded elsewhere. ¹²⁸ These positive results have led these researchers to conduct a clinical trial combining the two therapeutic interventions (the COMBINATION Therapy in Myocardial Infarction, or COMBAT-MI Trial).