Cardiogenic shock (CS) is a state of systemic tissue hypoperfusion due to cardiac failure in the presence of adequate intravascular volume.1 Left ventricular pump failure leading to CS remains one of the most serious and challenging conditions over time in cardiology and is still the most common cause of hospital mortality associated with myocardial infarction (MI).2 This review focuses on CS in the setting of acute ST-elevation MI (STEMI).
An incidence of CS of around 5–8% with mortality in excess of 50% in patients with STEMI1 has remained constant for many years; it has only started to decline in the last decade.3–5 The most important turning point was the 1999 publication of the landmark SHOCK (SHould we emergently revascularize Occluded Coronaries in cardiogenic shocK?) trial.6 This trial suggested that early revascularisation can decrease hospital mortality and, intuitively, the incidence of CS in comparison with initial medical stabilisation including thrombolysis (see Table 1). The study showed a 13% absolute risk reduction in mortality at six months, which was maintained at one year.7 Although the time-frame for inclusion in the SHOCK trial was as long as 48 hours after MI and 18 hours after shock onset, there was an apparent increase in long-term mortality as time to revascularisation increased from zero to eight hours.8
Since then, emergency mechanical revascularisation, mainly by percutaneous coronary intervention (PCI) but also by coronary artery bypass graft (CABG) if coronary anatomy is suitable, has became established as the preferred treatment for patients with STEMI. For CS, the American College of Cardiology/American Heart Association (ACC/AHA) 2004 guidelines9 place PCI or CABG as a Class I A recommendation in patients less than 75 years of age and Class IIa B in patients 75 years of age or older. According to the European Society of Cardiology 2008 guidelines,10 PCI for CS is a Class I B recommendation. Therefore, the main focus on the management of CS has turned to prevention by a strategy of early diagnosis and prompt reperfusion of patients with STEMI.11
Several non-modifiable risk factors for developing CS have been recognised, such as anterior STEMI in older patients with hypertension or diabetes.12 These are usually patients with multivessel coronary artery disease and previous angina, MI or an electrocardiographic pattern of left bundle-branch block. A late hospital presentation of older patients with anterior STEMI, increased heart rate and low blood pressure may preclude the possibility of subsequent CS. Although in the SHOCK trial6 the benefit was not clear in patients over 75 years of age, due to the ageing population it is expected that early revascularisation should not be denied to the elderly as they have the highest mortality rates. Recent data from the Melbourne Interventional Group Registry13 suggest that one-year survival of elderly patients with STEMI complicated by CS undergoing PCI is similar to that of younger patients.
The only modifiable risk factor is time to reperfusion, as demonstrated in a trial comparing pre-hospital thrombolysis and primary PCI, where reperfusion therapy performed within the first two hours of symptom onset significantly reduced the incidence of CS.14 Several MI programmes designed to improve compliance with STEMI guidelines for earlier diagnosis and mechanical reperfusion are being implemented in many countries. The programmes will certainly confirm the impact of improved compliance on the decline of CS.15
Although CS on admission is an independent predictor of in-hospital mortality, the majority of patients develop CS within hospital.16 Early suspicion based on features of haemodynamic instability (cautious use of beta-blockers, vasodilators, morphine and diuretics) should justify all established supportive and adjunctive management therapies. Such therapies include haemodynamic monitoring to volume support to ensure adequate sufficient pre-load, haemodynamic support with inotropic and vasopressor agents, ventilatory support and mechanical support by intra-aortic counterpulsation.
Haemodynamic monitoring with invasive arterial pressure and a Swan-Ganz catheter is the preferred method for assessment and guiding therapy.10,17 The benefit of the Swan-Ganz catheter is controversial, since several retrospective analyses in CS, including a large population from the GUSTO IIb and III trials, failed to demonstrate improvement in prognosis.18
Haemodynamic profiles usually include persistent hypotension (systolic blood pressure <90mmHg), low output (cardiac index <2.0l/min/m2) and elevated left ventricular filling pressure (pulmonary wedge pressure >18mmHg). Systemic vascular resistance is inadequately elevated for the level of cardiac output. This is due to the excessive release of vasodilatation mediators (nitric oxide [NO]) and inflammatory cytokines (tumour necrosis factor-alpha and interleukin-6) by endothelial cells that reduce catecholamine responsivity, increase systemic vasodilation and decrease myocardial contractility and further depress tissue perfusion pressure.19
Supportive treatment with vasopressor and inotropic agents should be promptly initiated to improve tissue perfusion. Norepinephrine is the agent of choice in patients with severe hypotension (systolic blood pressure <80mmHg) and dobutamine combined with a vasopressor agent in other patients.10,20 Dose titration should not be aimed at correcting the haemodynamic profile but should aim for an acceptable perfusion pressure (mean blood pressure >60mmHg), since sympathomimetic amines increase oxygen demand by the jeopardised myocardium and higher doses are associated with increased mortality.21
Invasive mechanical ventilation with appropriate sedation is usually necessary to ensure optimal arterial oxygenation. Besides total respiratory support, other advantages include reduction in the work of breathing, saving the low output to maintain other vital functions, decrease in pre-load and pulmonary oedema and avoidance of unnecessary suffering associated with the disease and invasive procedures.
Creatinine clearance is a powerful independent predictor of mortality in patients with CS.21 Besides its vital roles in water and salt management, the impairment of acid excretion contributes to metabolic acidosis, which further reduces myocardial contractility and promotes inappropriate vasodilatation. In patients with severe CS, the impairment in renal function usually requires the initiation of continuous venous haemofiltration. This may inhibit shock pathogenesis through the clearance of vasodilatation mediators and inflammatory cytokines.22 Intensive insulin therapy is recommended in patients who are hyperglycaemic and have complicated MI.15 Hyperglycaemia (≥200mg/dl) is associated with a poorer prognosis in patients with acute MI and may be related to the directly negative effects of glucose.
Regarding antithrombotic therapy, all patients with acute MI should receive a combination of aspirin, clopidogrel or prasugrel and heparin.15 Nevertheless, in patients with shock or impending shock at admission, the loading dose of clopidogrel/prasugrel may be deferred until there is sufficient knowledge of the coronary anatomy, since immediate CABG may be the preferred revascularisation method.
In patients undergoing PCI in the setting of CS, the use of glycoprotein IIb/IIIa inhibitors can be considered in the catheterisation laboratory, but their use upstream is uncertain.15,23 Several observations are in agreement with the long-term benefit of abciximab in MI-free survival, demonstrated in patients undergoing primary stenting for STEMI.24 Heparin should be maintained until intra-aortic balloon pump (IABP) removal. Bivalirudin is a valuable alternative in patients with a high risk of bleeding, and a low-molecular-weight heparin can be used for thromboembolism prophylaxis until adequate mobilisation of the patient is achieved.
A special place in the treatment of CS is usually given to mechanical circulatory support. The concept is intuitive and based on left ventricular pump failure. The IABP was introduced in 1968 to improve diastolic coronary perfusion and systemic blood flow and to reduce afterload and myocardial work.25 The use of IABP in CS after MI has been reported to be variable. The ACC/AHA guidelines9 list IABP therapy as a class IB recommendation and the European Society of Cardiology guidelines10 as Class I C in CS.
However, use of IABP is a typical example of the adoption of a treatment based on a concept that becomes clinical practice.26 The above-mentioned recommendations were based on non-randomised studies only, as no randomised clinical trial of IABP support for STEMI complicated by CS has been performed and no unequivocal evidence exists.
A systematic review and meta-analysis of IABP therapy in STEMI patients was recently published.27 In 10,529 patients from nine cohorts of STEMI patients with CS, IABP was associated with an 18% decrease in 30-day mortality in patients treated with thrombolysis (although they were younger and had significantly higher revascularisation/PCI rates). It was also associated with a 6% increase in 30-day mortality in patients treated with PCI.27 This striking observation must be interpreted cautiously due to the possible influence of bias and confounding factors inherent in cohort studies. A large clinical trial is planned to answer whether IABP is beneficial for the treatment of CS in addition to PCI.28 For the time being, it appears that current clinical practice should continue.
Despite developments with early emergency revascularisation strategies and continuous use of known supportive measures, mortality of CS after STEMI remains unacceptably high. Other approaches, including new pharmacological therapies based on the pathophysiology of CS, new inotropic agents and other forms of mechanical haemodynamic support, have recently been investigated.
The first approach relates to the complex pathophysiology of CS (see Table 2). The classic paradigm is that left ventricular dysfunction and low cardiac output, being initially compensated for by catecholamine release and peripheral vasoconstriction, creates a vicious circle. In addition to systemic vasoconstriction, there is activation of the neurohormonal cascade. More recently, the possibly important roles of a systemic inflammatory response, complement activation, release of inflammatory cytokines and expression of inducible NO synthase (NOS) have been considered.29
This possible overlap between CS and the systemic inflammatory response creates high levels of NO, which may cause inappropriate systemic vasodilatation as well as systemic and coronary hypoperfusion and therefore may contribute directly to myocardial depression. Patients with CS may be particularly sensitive to even slight reductions of coronary blood flow to the surviving myocardium.
Initial results of the effect of NOS inhibition in refractory CS were both promising and conflicting. In the LINCS randomised study30 of 30 patients with the non-selective NOS inhibitor L-NAME, there was a significant reduction in mortality at 30 days from 67% (control patients) to 27% (those receiving L-NAME) (p=0.008). In the SHOCK-2 randomised phase II dose-ranging study31 with 78 patients and different doses of L-NMMA (a non-selective NOS inhibitor), there was a modest increase in mean arterial pressure at 15 minutes but no difference at two hours and no differences in mortality. However, the studies confirm that an excess of NO does play a role in the hypotension associated with CS.
Results of the randomised Tilarginine Acetate Injection in a Randomized International Study in Unstable MI Patients With Cardiogenic Shock (TRIUMPH) trial32 with tilarginine (a non-selective NOS inhibitor) were disappointing. The study included 398 patients with refractory CS and, despite successful coronary revacularisation, the study was prematurely terminated after a pre-specified analysis of futility. There was a significant increase in systolic pressure at two hours, but no differences in the primary end-point of 30-day mortality (48 versus 42% – a 6% decrease) and in the secondary end-points of shock resolution, duration of shock and six-month mortality (58 versus 59%).
This might not be the final answer, however, as NO can be produced by three NOS inhibitor isoforms and only one was studied. Two (neuronal and endothelial) are constitutive enzymes. NO production is calcium-dependent and provides myocardial protection. The third inducible NOS is calcium-independent and may contribute to tissue damage. The balance between the different types of NOS inhibition may be crucial in the outcome of treatment and it is possible that the NO pathway in the setting of CS is not completely known.33
Although prompt restoration of blood flow in the infarct-related coronary artery is the main therapy for improving survival in STEMI, reperfusion may cause irreversible damage to the ischaemic myocardium, which may account for the high early mortality in CS.34 Mediators of this reperfusion injury include oxidative stress and inflammation, which trigger apoptosis.
Several metabolic strategies thought to cause more advantageous energy production from glucose failed to demonstrate any benefit.35 Nevertheless, in patients with hyperglycaemia, randomised trial evidence supports the use of insulin infusion, although the optimal level of glycaemia is uncertain.9
Complement inhibition with paxelizumab in high-risk acute STEMI also failed to improve clinical outcomes.36 A new strategy targeted to selective inhibition of d-protein kinase C, a mediator of apoptosis, significantly reduced infarct size in a phase IIA dose-finding, double-blind, randomised trial in 154 patients with anterior STEMI undergoing primary PCI.37
Levosimendan belongs to a new group of inotropic agents that enhance the sensitivity of myofilaments to calcium and induce vasodilation through the opening of potassium channels.38 The Randomized Study on Safety and Effectiveness of Levosimendan in Patients with Left Ventricular Failure after an Acute Myocardial Infarct (RUSSLAN) trial was a double-blind, randomised, placebo-controlled trial in 504 patients with acute MI and left ventricular failure. Almost 60% of patients included had signs of poor tissue perfusion. Levosimendan was associated with a significant reduction of death at 14 days with a similar rate of ischaemia and/or hypotension.39 Preliminary data in 22 consecutive patients who developed CS after primary PCI revealed that levosimendan had a better effect than dobutamine in terms of cardiac power and pulmonary capillary wedge pressure by the end of the 24-hour infusion.40 This inodilator also improves coronary blood flow and exerts antistunning effects, which may accelerate myocardial recovery after revascularisation in CS.38
Another approach relates to mechanical haemodynamic support, as IABP provides no active mechanical augmentation of cardiac output, although favourably altering the balance between myocardial oxygen supply and demand. A more recent alternative is the total and active circulatory support to the failing heart provided by percutaneous left ventricular assist devices (LVADs) and extracorporeal life support.41
There are surgically implanted (cannula placed in the left ventricular apex and connected to the ascending aorta) and percutaneously implanted LVADs. The two commonly used percutaneous devices are:
- TandemHeart – the cannula is inserted into the left atrium by transseptal access and the blood is returned to the femoral artery with retrograde perfusion of the abdominal and thoracic aorta with a centrifugal pump; and
- Impella Recover LP 2.5, which is a microaxial propeller pump implanted across the aortic valve.
Three randomised clinical trials in patients with CS complicating MI, two comparing the TandemHeart42,43 and one the Impella44 system with IABP, have been completed. The meta-analysis of these three trials was recently published.45 The trials were small, with only a total of 100 patients included, only one trial was multicentric, follow-up was short and the LVADs were different in terms of insertion, mechanical properties and mechanisms of action. With all of these limitations, the meta-analysis showed that patients with LVAD had a higher cardiac index, higher mean arterial pressure and lower pulmonary capillary wedge pressure two hours after baseline compared with IABP patients. However, there were no significant differences in 30-day mortality or in the incidence of leg ischaemia. Bleeding was significantly higher with the TandemHeart compared with IABP and there was a higher rate of multiorgan dysfunction, with a clinical profile suggesting systemic inflammatory response syndrome.
Therefore, with such limited experience in CS patients, at the moment LVAD offers no more than great promise. Larger randomised clinical trials (comparing IABPs with LVADs and between various LVADs) need to be performed to produce more convincing evidence of which is the best mechanical support system and where their place is for reducing mortality in CS after acute MI. Such trials are currently being conducted.46,47
Surgical LVADs have not been investigated in patients with CS and extracorporeal life support. Although extracorporeal circulation and a membrane oxygenator offer a more complete system to relieve the workload of the heart and lungs, it might not be suitable in the setting of CS complicating MI.
In conclusion, the most significant treatment strategy so far for patients with CS after STEMI has been early and aggressive revascularisation with primary PCI. The more complete long-term survival data are derived from the SHOCK trial.8 In the early revascularisation group there is a survival rate of 41% at three years and 32.8% at six years compared with 19.6% in the initial medical stabilisation group. For hospital survivors, rates were 62.4 and 44.4%, respectively. A strategy of early revascularisation resulted in a 13.2% absolute and a 67% relative improvement in six-year survival compared with initial medical stabilisation.
CS remains a difficult, perplexing and frequently fatal condition. Better control of risk factors, greater patient compliance with established medical therapies for coronary artery disease and subsequent decrease in MI are the current goals. So are increased patient awareness, early diagnosis and widespread availability of reperfusion therapies for acute MI, coupled with in-hospital medical compliance with the guidelines. The future holds renewed basic, translational and clinical research based on the pathophysiology of the syndrome.