Ventricular tachyarrhythmias (VAs) most commonly occur early in ischaemia, and patients presenting with an acute MI and ventricular arrhythmias are a group that has a significantly increased risk of mortality.1,2 Thrombolysis primary percutaneous coronary intervention (PCI) and use of beta-blockers, while resulting in the modification of the natural history of an infarct, have also reduced the incidence of sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) occurring within 48 hours of the onset of an acute coronary syndrome (ACS), over the past decades.3 The prevention and treatment of haemodynamically significant VA in the post-infarct period, and of sudden cardiac death (SCD) remote from the event, remain areas of ongoing study.
Prompt revascularisation and drug therapy, including anti-platelets, statins, angiotensin converting enzyme (ACE)-inhibitors and betablockers, have markedly reduced the incidence of VA.4–8 Nevertheless, approximately 10% of post-MI survivors remain at high risk of dying in the first months or years following hospital discharge (mortality >25% at 2 years).9 Sudden death secondary to sustained VT or VF accounts for about 50% of all deaths in these high-risk patients.10
There remain three vulnerable classes of patients: patients presenting after a long period of chest pain, patients who have undergone only partial revascularisation and those with a pre-existing arrhythmogenic substrate.11 VA is seen more often in those with cardiogenic shock, related to the size of the infarct. A genetic predisposition to VA in the context of ischaemia may also exist. The finding of early repolarisation (ER) changes being more prevalent in idiopathic VF survivors and their relatives, and on the ECGs of patients with coronary artery disease and ST-segment elevation MI (STEMI) who experience VA perhaps alludes to this.12,13
There is a temporal distribution to VA post-acute MI: an early, or acute phase, of up to 48–72 hours, which is a time of very dynamic ischaemia and reperfusion. From 72 hours to a few weeks up to a month post event, and a more chronic phase beyond that, where remodelling continues to occur. Premature ventricular contractions (PVCs) are common in the early phase. The significance of these and the occurrence of non-sustained or sustained VA in terms of shortand long-term prognosis have been debated over the years. It appears there needs to exist a combination of biochemical, electrophysiological, autonomic and, as yet unknown, genetic factors culminating in a so-called ‘perfect storm’ resulting in arrhythmia in the post-MI period. Patients who develop sustained VF or VT >48 hours after their index MI have a significantly higher rate of all-cause mortality; however, the relationship between early (within 48 hours of the index MI) VF/VT and mortality remains controversial. Some data have suggested that sustained ventricular arrhythmias during the early post-MI period may be associated with increased 30-day mortality, but without a protracted risk over the long term.14–16
Non-sustained VT in the early-phase post-MI does not contribute to risk assessment in this group of patients.17 The prognostic implication of VA post PCI was studied in the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONSAMI) Trial. In this, 5.2% patients developed VA post primary PCI with 85% of the VA happening in the first 48 hours. They reported that sustained VT/VF after primary PCI was not significantly associated with 3-year mortality or major adverse clinical events.18
Mechanisms of Ischaemic Arrhythmogenesis: From the Cell Upwards
Acute MI is characterised histopathologically by coagulative necrosis of the myocardium. In a non-reperfused MI, this is seen within 30 to 40 minutes of sustained ischaemia, with the changes only visible at the resolution electron microscopy. From 2 weeks, scar develops from the periphery to the centre, and its formation is complete after the second month.19 Any attempt at reperfusion potentially alters this process.
Thus there is a temporal distribution to the occurrence and postulated mechanisms of ventricular arrhythmia in the post-MI period. In animal studies, an early, potentially reversible, phase within the first 30 minutes following epicardial coronary artery occlusion was identified. This is followed by an irreversible phase from 90 minutes to 72 hours, during which there is rapid evolution of the characteristics of the infarcted tissue.19 Reperfusion contributes to the profound electrophysiological changes.
Acute ischaemia causes hypoxia, which results in an intracellular depletion of adenosine triphosphate and the consequent accumulation of adenosine diphosphate and the products of anaerobic glycolysis, leading to intracellular acidosis. This drop in pH activates the Na+/H+ and Na+/Ca++ion exchange channels, with expulsion of hydrogen ions in exchange for sodium, which passes into the cell and is then exchanged for calcium, resulting in cell swelling and calcium overload. This is accompanied by the build-up of extracellular potassium, cathecholamines and lysophosphatidylcholine. This results in depolarisation of the cell membrane and reduction of the fast inward sodium current and increase in the late sodium current initially prolonging the action potential duration (APD). Ultimately, abbreviation of the APD, seen during ischaemia, results from decreased inward calcium currents (inhibited by the acidosis) and enhanced outward ATP-sensitive potassium current due to reduction in intracellular ATP, following hypoxia. A lower reduced resting transmembrane potential, intracellular calcium mishandling and reduced functional gap junctions also result. The cessation of anaerobic glycolysis marks the next phase resulting in low glycogen and high lactic acid intracellular content, reduction of ATP levels to below 10%, sodium and calcium overload and further accumulation of extracellular potassium. Spontaneous calcium oscillations trigger early and late afterdepolarisation-induced ventricular ectopics.11,19
Surviving purkinje fibres with shortened APD or reduced amplitude, depolarised membrane potentials and reduced Vmax are thought to be the source of automatic foci for VA. The partial and temporal dispersal repolarisation contribute to a re-entrant mechanism based on regions of unidirectional conduction block, fractionation of cellular electrograms and shortened APDs.11 Tissue heterogeneity is particularly marked in the peri-infarct or ‘border zone’ and it is here that arrhythmogenesis is thought to arise.11,19 Of note, both human and animal studies have shown abnormal sympathetic activity in these border zones. These nerve terminals are more susceptible to ischaemic damage than myocytes.20
It has been well recognised that re-entry through a stable circuit involving the infarct scar tissue is the most-likely mechanism of sustained monomorphic ventricular tachycardia.21 In acute myocardial ischaemia, with no previous scar, zones of slow conduction and block may create conditions for re-entry. These patients are likely to have large infarct areas and a very large acute ischaemic zone may create the conditions for a transiently stable re-entry circuit capable of sustaining a monomorphic re-entrant tachycardia.15 Alternatively, mechanical stretching of a failing ventricle alongside high sympathetic drive associated with MI can result in VA because of focal triggers.22
Identifying those at Risk of Post-infarction Ventricular Tachyarrhythmia
Various invasive and non-invasive tools have been used to identify patients most at risk of SCD following MI. Left ventricular ejection along with inducibility of VA during programmed electrical stimulation have been most cited as factors identifying patients most at risk. Due to the dynamic phase of repair and remodelling after the acute episode, the acute assessment of these parameters may not reliably predict longterm risk of arrhythmic death.
It is estimated that around 15–20% of all AMI patients will have LVEF ≤35% at the time of revascularisation for AMI.23 However, recovery of reduced LVEF after AMI with or without immediate PCI is often unpredictable. In a study by Solomon et al., 3 months after AMI, 22% of all patients with abnormal LV function at the time of AMI recovered to normal LV function.24 Assessment of LVEF to risk stratify and subsequent implantation of ICD early (within 40 days) post-MI has not been shown to be of prognostic benefit.25,26
Left Ventricular Ejection Fraction
The incidence of VA is directly proportional to the size of an infarct and inversely related to the LVEF. Late presenters, and in those whom there is failure to achieve adequate patency of the culprit artery, are the most at risk. In addition, over a third of patients with STEMI, and the majority of those with STEMI complicated by cardiogenic shock, have significant bystander coronary disease, and appear to be at increased risk of VA,11 presenting a valid argument for early full revascularisation. Transthoracic echocardiography is used routinely to assess the extent of infarct and to risk stratify patients on the basis of LVEF. When quantifying LVEF alone, there is greater operator variability in echo compared with MRI. Irrespective of the mode of LVEF determination, there are limitations in its use in identifying those at risk of VA and SCD.
The multicenter automatic defibrillator implantation trial (MADIT) identified a mortality benefit from ICD in 5.6% of patients with a postinfarct LVEF of 30% or lower over 27 months from the index event.27 The sudden cardiac death in heart failure trial (SCD-HeFT) identified a mortality benefit of 7.3% over 60 months in those with an LVEF of 35% or less.28 Apart from being quite modest numbers, most patients who suffer a cardiac arrest post-MI have an LVEF higher than 35%.29 When looking at VA and SCD risk in the longer term it is worth noting that less than 20% of ICD recipients in the above-mentioned studies received appropriate ICD therapies in their respective follow-up periods.27
Additionally in the setting of chronic coronary artery disease, an analysis of the multicenter unsustained tachycardia trial (MUSTT) study would seem to advise caution in attributing risk based on LVEF alone. Those who experienced non-sustained ventricular tachycardia (NSVT), a condition of recruitment into the study, and an LVEF of anywhere between 30% and 40%, demonstrated higher risk of arrhythmic death or cardiac arrest compared with LVEF of or less than 30% when other factors were taken into account.30 The variables having the greatest prognostic impact in multivariable analysis were functional class, history of heart failure, NSVT not related to bypass surgery, EF, age, LV conduction abnormalities, inducible sustained ventricular tachycardia, enrolment as an inpatient and AF. Non-invasive assessments of scar burden, ventricular conduction and repolarisation, as well as autonomic tone, have been explored in risk prediction. Cardiac magnetic resonance is far superior at characterising infarcted tissue and assessing scar burden compared with other imaging modalities. Increased tissue heterogeneity31 and a larger peri-infarct or ‘border zone’ has been found to correlate with increased mortality risk and MRI is better able to assess these.32 Large-scale data assessing how primary prevention ICD therapy could be guided by MRI beyond LVEF assessment is lacking. A study of 48 patients with known coronary artery disease referred for PES did find that infarct surface area and mass measured by cardiac MRI more accurately identified patients with a substrate for monomorphic VT, compared with LVEF.33 The majority of patients were studied beyond a month post-MI. DETERMINE-ICD is a current large, randomised trial looking into prophylactic ICD therapy in post-MI in patients with an LVEF greater than 35% and extensive scarring assessed by MRI.
Programmed Electrical Stimulation
Current data only support the use of PES in post-MI patients with LVEF of 40% or less. The timing of programmed electrical stimulation (PES) is an area of debate. In MADIT I, patients with inducible VA and LVEF of 35% or less late post-MI were found to derive most mortality benefit from an ICD.34 The same finding in beta-blocker Strategy plus Implantable Cardioverter Defibrillator (BEST+ICD) less than a month post-MI suggested that inducibility early after the event may do the same but was really unable to predict benefit.35 The Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infarction (CARISMA) trial, however, found that inducible VT at 6 weeks post-MI was a strong predictor of future arrhythmic events.36 Beyond the fact that differences in stimulation protocol is one factor that can account for the variation in study findings, PES is still marred by low sensitivity demonstrated in part by the fact that more than a quarter of patients with an LVEF of 35% or less and a negative study went on to have serious events.37
Other Risk Assessment Modalities
None of the non-invasive assessments of ventricular conduction and repolarisation or autonomic tone have yet been individually found to be of use in accurately predicting risk of VA or to guide ICD therapy.29 Measures of ventricular conduction include QRS duration, signal-averaged ECG and Wedensky modulation (the identification of local perturbations within the QRS following delivery of a subthreshold impulse during the ventricular refractory period). Measures of ventricular repolarisation include QT variability and dispersion, T loop morphology variations, T wave variance, the QT/RR slope and T wave alternans. Measures of autonomic tone include assessment of linear and non-linear heart rate variability (HRV), baroreflex sensitivity, heart rate turbulence and deceleration capacity. Risk Estimation following Infarction Non-invasive Estimation (REFINE)-ICD has been designed to evaluate prophylactic ICD therapy in post-MI patients with LVEF of 36% to 49% and abnormal Holter T wave alternans and heart rate turbulence.29
An increased incidence of ER on the ECG in the form of a slurring or notching in the terminal portion of the QRS complex in the inferior and/or lateral leads has been found in patients with coronary artery disease. Patel et al. compared 50 individuals with VA within the first 72 hours post-STEMI and 50 individuals without VA. Arrhythmias included sustained VT, non-sustained VT and VF. When looking at ECGs recorded 1 year prior to the MI, they found a higher prevalence of ER among those with VA, even after adjusting for creatine kinase (CK)-MB levels and LVEF.38 Another group reported a higher prevalence of ER in over 400 SCD victims ascribed to acute coronary syndrome following postmortem.39 This was in addition to the findings that more victims were male smokers with a lower BMI, higher heart rate, with prolonged QRS durations and a lower prevalence of history of cardiovascular disease. Early repolarisation syndrome (ERS) may be a marker of vulnerable substrate. The same finding of increased incidence of ER in idiopathic VF survivors40,41 and their relatives42 would suggest that there exists a phenotype with a predisposition to VA in the context of ACS.
The incidence of sustained VT and VF occurring within 48 hours of the onset of an ACS seems to have decreased over the past decade, likely due to the widespread availability of revascularisation therapy, limiting the size of infarction and to an increased use of beta-blockers.43
Robust evidence for anti-arrhythmic drug (AAD) use in the early dynamic phase of ischaemia and reperfusion within the first 48–72 hours post-MI is lacking in comparison to use in the chronic phase. Despite the availability of early revascularisation and beta-blocker use, 6% of patients in this early phase are still affected by sustained VA.44 While the immediate treatment for VA with haemodynamic collapse remains direct current cardioversion, recurrent sustained VA in the absence of need for further revascularisation and normal electrolytes usually calls for some form of drug therapy. AADs are not without their side effects in addition to their effects on transmembrane voltage and heart rate, all of which can further exacerbate instability.
Beta-blockers have been effectively used in patients with acute coronary events, reducing major cardiac events including SCD.45 In a meta-analysis by Huang et al., use of beta-blockers was associated with reduction of all-cause death in patients with acute MI undergoing PCI. The benefit was restricted to those with reduced EF, low use of other secondary prevention drugs or with none STEMI. The association between the use of beta-blockers and improved survival rate was significant only in <1-year follow-up duration. They concluded there was a lack of evidence to support routine use of beta-blockers in all patients with AMI who underwent PCI.46 In the carvedilol post-infarct survival control in left ventricular dysfunction study (CAPRICORN) trial, Carvedilol was shown to have significant anti-arrhythmic effect after AMI. It suppressed both atrial and ventricular arrhythmias in these patients.47
There have been conflicting reports concerning the class Ib drug lidocaine of either a significant trend towards a lower risk of death in the early period post-MI,48,49 and less VA and a survival benefit post-cardiac arrest when used prophylactically to a neutral effect on overall mortality or a trend towards excess mortality.49,50 Prophylactic lidocaine use has largely been discouraged although it remains a potential intravenous treatment of recurrent sustained VA post-MI. The class Ic drugs like flecainide and propafenone cause significant slowing of conduction, which may exacerbate VA in the post-MI setting and should not be used.50
Amiodarone remains the most commonly used AAD post-MI and is particularly useful in the presence of severe structural disease. However, it does take time to reach therapeutic levels. Its use following out-of-hospital cardiac arrest in patients with shock refractory VF was associated with a survival benefit in comparison with lidocaine.51 In patients who survived more than 3 hours after MI, use of amiodarone was associated with increased short- (30 days) and long-term (6 months) mortality compared with lidocaine for use in the ACS setting.49 No added survival benefit has been shown with amiodarone use over and above concomitant beta-blocker therapy post-MI52 and its significant side-effect profile has been shown to increase mortality in the longer term. There are no data to support the use of dronaderone in the post-MI period. Other class III drugs, such as dofetilide, prolong cardiac repolarisation and do suppress VA,51–53 but there are no data showing added beneficial effects of their use. All class III drugs prolong the QT interval thus risking polymorphic VT, though the incidence of polymorphic VT is low with amiodarone.54
Ranolazine is a relatively new entry into the arena. The metabolic efficiency with ranolazine for less ischemia in non−ST-elevation acute coronary syndromes (MERLIN)-TIMI 36 trial did not show a significant difference in the combined primary endpoint of cardiovascular death, MI or recurrent ischaemia but it did significantly reduce the incidence of non-sustained VT when compared with placebo.55 More investigation is needed, including how it stacks up against current drug therapy
Similarly, eplerenone 25 mg/day, in addition to conventional therapy, significantly reduced all-cause mortality 30 days after randomisation in patients with an LVEF ≤40% and signs of heart failure. There was a 37% relative risk reduction of SCD in patients receiving epleronone.56
When considering drug therapy, the use of deep sedation must not be forgotten. In addition to the necessary use if a conscious patient is to undergo direct current cardioversion, a reduction of the sympathetic drive associated with post-MI VA afforded by it makes it a viable therapeutic option.
If the above measures fail to suppress VA in the early post-MI period, temporary overdrive pacing may be used. An automatic focus may be captured and suppressed, or exit block achieved by making surrounding myocardium refractory. Alterations of conduction and refractoriness due to pacing may terminate a tachycardia caused by a re-entrant mechanism. This measure can be used in refractory VA to reduce the need for recurrent cardioversion while waiting for drug therapy to take effect, or prior to further revascularisation or catheter ablation.
Catheter ablation for VA in the acute phase is not commonly performed. The acute success rate is in the region of 70% and carries with it a peri-procedural mortality of 3% in unstable patients, and a long-term mortality of 18% due to decompensated heart failure.57,58
Ablation is primarily subendocardial and in the border zone. The targets are the re-entrant circuits in the heterogenous myocardium and the after-depolarisations and automatic foci arising from purkinje fibres. Activation mapping can be performed in the presence of frequent PVCs. Pace mapping can be performed against prior recorded PVCs if they are less frequent. Endpoints include suppression of the triggering PVC and loss of the purkinje potential. On the occasions that PVCs are not present, routine induction manoeuvres including PES or drug provocation can be non-specific in the acute period is often unsuccessful and can be non-specific. In these situations, substrate ablation guided by voltage mapping can be undertaken.59 This cohort of patient is often haemodynamically unstable and the complexity of the procedure means it is best undertaken in highvolume centres by experienced electrophysiologists, with the use of 3D electroanatomical mapping systems and, one would argue, advanced supportive care including the ability to provide mechanical circulatory support if needed.
Mechanical Circulatory Support
VA is common in MI complicated by cardiogenic shock and is associated with high short-term mortality. Sustained VT occurs in 17–21% and VF is seen slightly more often (24–29%) in selected patients with cardiogenic shock and acute MI, undergoing thrombolysis or primary PCI.60 Revascularisation improves survival. The use of inotropes can exacerbate VA and the amount needed can potentially be reduced if used in conjunction with mechanical support. Beyond supporting revascularisation procedures, mechanical support may help maintain cardiac output adequate for tissue perfusion in the post-MI period.
The most widely used form is the intra-aortic balloon pump (IABP). This counter-pulsation device primarily reduces afterload, augments diastolic coronary perfusion and contributes towards the cardiac output. It is unable to provide support in VF, whereas other forms of mechanical support can. Use of the Impella assist device in this cohort of patients was associated with reduction of tissue hypoxia, haemodynamic stabilisation and improvement of neurological outcome.61,62 Veno-arterial extracorporeal membrane oxygenation (VA–ECMO)-assisted PCI was shown to improve survival in patients with cardiogenic shock and refractory VT/VF compared with IABP.61–63 These forms of support can act as bridges to recovery or eventual heart transplantation.
The Implantable Cardioverter-Defibrillator
Current guidance advocates ICD implantation from 40 days post-MI in patients with an LVEF of 35% or less in New York Heart Association (NYHA) class 1, 2 or 3. The Defibrillator in Acute Myocardial Infarction Trial (DINAMIT) and the Immediate Risk Stratification Improves Survival (IRIS) showed that there was no survival benefit in implantation under 40 days from the event.25,26 In the DINAMIT trial, ICD implantation within 6–40 days of an acute MI (average time from MI to randomisation of 18 days) was compared with conventional medical therapy. This was a primary prevention study and excluded people who had VA post-48 hours after MI. The DINAMIT study included a measurement of HRV, with an EF of <35% as study inclusion criteria. The study demonstrated a reduction in the arrhythmic death, which was largely balanced by an increase in the non-arrhythmic cardiac death in the ICD arm when compared with the control group, but there was no reduction in the total mortality.
Similarly, the IRIS trial enrolled patients, 5–31 days after MI. The inclusion criteria included a reduced LVEF (≤40%) and a heart rate of 90 or more. This was also a primary prevention study and failed to show any benefit of prophylactic ICD in this group of patients, though the rate of arhhythmic deaths was lower in patients with ICD.
Though both the trials showed no significant benefit of ICD in this group of patients, they did not include people who had VA after 48 hours of the myocardial event and such patients need to be studied further. Risk stratification tools are required to assess these people who may be at a higher risk of VA. In the absence of robust risk stratification tools, wearable defibrillators may have a role in the prevention of SCD, particularly in patients with depressed LV function, but such a strategy will need to be evaluated in further randomised studies.
The risk of VA and consequent ICD therapy appears to reduce with increasing time from the infarct and the majority of patients fitted with a primary prevention ICD (and without further ischaemia) do not have sustained VA, implying that there might be other factors at play beyond LV function and stable scar.64
An ICD prevents sudden death from VA but does not prevent VA itself. It certainly does not prevent death from progressive pump failure and may in fact just allow for change in the mode of death. Both appropriate and inappropriate shocks have been associated with increased mortality, whereas ATP-treated arrhythmias were not. The psychological impact of ICD therapy must not be underestimated. More aggressive ATP, extended detection and redetection algorithms, onset and stability criteria and morphology discrimination can help reduce the frequency of inappropriate shocks.65
Further ischaemia or infarction may precipitate an electrical storm in patients who already have an ICD. In such situations, treatment should be as for someone without an ICD but an added consideration may be to reprogramme the device to deal with increased frequency or altered characteristics of arrhythmia and perhaps even switching device therapies off in the short term in a well-monitored environment to prevent excessive or inappropriate ATP and/or shocks.
A greater understanding of the cellular and electrophysiological mechanisms of arrhythmia in the context of acute ischaemia is needed. It is clear that risk stratification based on LVEF and PES alone is inadequate and more robust investigations used in combination are likely to be required. Current therapy is limited to drugs with significant side effects and catheter ablation techniques.
The concept of a genetic predisposition and therapies targeted at autonomic modulation are fascinating and warrant further research. Timely and thorough revascularisation appears to be a strategy to prevent death due to VA but it cannot be achieved in all patients presenting post-infarct.
Perhaps the future for some does lie in molecular and stem cell therapy, with the potential to regenerate lost or damaged myocardium.66 Stem cell therapy has been used to treat heart failure. In early trials in an animal model (rat), and subsequently in humans, stem cell therapy was shown to increase propensity to ventricular arrhythmias (ventricular ectopics and non-sustained VA), perhaps due to lack of effective integration to the connexin network.67,68
Though further studies using stem cell therapies in the animal model and subsequently in the humans have been reported to either reduce propensity for cardiac arrhythmias or show no change in incidence of sustained ventricular arrhythmias, further studies are required.69,70 Ventricular arrhythmias in patients with acute coronary syndrome could influence the immediate and long-term mortality in these patients. Appropriate risk assessment is needed to identify patients at risk of further risk of VA and sudden cardiac death, and research is needed in this field.