Practical Cardiovascular Medicine. Elias B. Hanna

injury that aggravates myocardial depression. Coronary reocclusion, coronary perforation with tamponade, and bleeding complications are also considered.

       For a given LV contractility, the more severe impairment of LV compliance in acute MI leads to a more severe rise of LV filling pressure, which further reduces coronary perfusion pressure.

       Transient hypotension (drugs, arrhythmia, sedation) in an initially stable patient may transiently reduce coronary blood flow and thus initiate a vicious circle of progressive myocardial ischemia that sustains the hypotension. Furthermore, since the pulmonary edema of MI results from volume redistribution rather than florid volume overload, aggressive diuresis may precipitate shock in MI. Also, β-blockers, ACE-Is, and other vasodilators, including sedatives administered during PCI or during intubation, may precipitate shock in a pre-shock patient who depends on the compensatory vasoconstriction and tachycardia. This partly explains why cardiogenic shock often develops after hospital admission.While mild hypotension unloads the LV and may be tolerated in chronic LV failure, it is not well tolerated in a patient with acute ischemia and unstable CAD or in a patient with RV failure.

       In over 25% of MI-associated cardiogenic shock, SVR is inappropriately low or normal rather than elevated, despite the use of vasopres-sors (SVR ≤1000 dyn.s.cm–5).¹⁰⁵ This mismatch between myocardial depression and inappropriate vasodilatation (or lack of compensa- tory vasoconstriction) may result in cardiogenic shock. Also, 18% of patients, mainly those with a low initial SVR, go on to develop a clinical picture of sepsis with fever or leukocytosis 2–4 days later, mostly with positive bacterial cultures. Thus, inappropriate vasodilata- tion is initiated by a systemic inflammatory response syndrome (SIRS) secondary to MI early on, then a septic process later on, and contributes to shock in a substantial proportion of patients. This implies a role for vasopressors in this subset of patients. High levels of cytokines and inducible nitric oxide synthase, beyond the healthy levels of endothelial nitric oxide synthase, precipitate vasodilatation and further myocardial depression. The initial vasodilatation, per se, is associated with an increased risk of later sepsis (bacterial translocation?).¹⁰⁵

      Some patients with a pre-shock state before PCI develop a full-blown shock after PCI. PCI may initiate a reperfusion injury with further activation of inducible nitric oxide synthase, and thus vasodilatation and myocardial depression. This is a temporary phenomenon, as the benefit from PCI eventually takes over. Also, the use of sedatives and supine positioning may precipitate shock during PCI.

      C. Management of LV-related cardiogenic shock

      A shock is defined as sustained SBP <90 mmHg for at least 30 minutes, with signs of low perfusion (oliguria <30 ml/h, cold/clammy extremi- ties, impaired mentation, increased lactate >2 mmol/l).⁷³ LV-related cardiogenic shock is characterized by additional features suggestive of increased left-sided filling pressure, such as clinical or radiographic pulmonary edema. Supportive hemodynamic features consist of a car- diac index ≤2.2 l/min and PCWP ≥15 mmHg, but right heart catheterization was not absolutely required in the SHOCK trial when pulmo- nary congestion was evident in a patient with anterior MI. Echo may be used to assess left-sided filling pressures and confirm the diagnosis, in addition to ruling out mechanical complications.

      In the SHOCK trial, patients with cardiogenic shock and STEMI or Q-wave MI of less than 36 hours’ duration were randomized to emergent revascularization vs. medical therapy. The median time from the onset of MI to shock was 5.6 hours, from MI to PCI 11 hours, and from MI to CABG 19 hours. Approximately 65% of patients had three-vessel disease and ~20% had left main disease. Revascularization with PCI or CABG, as appropriate, reduced the absolute 6-month mortality by a drastic 13% (30-day mortality 46% vs. 56%, with a lower mortality of 38% if successful PCI was performed; 6-month mortality 50% vs. 63%). Patients who survive the acute phase of cardiogenic shock have a good long-term survival, two-thirds being alive at 6 years.^106,107

      Notably, fibrinolytics were administered to 63% of the SHOCK trial patients managed medically and were associated with a marked and significant 40% mortality reduction in comparison to no fibrinolytic therapy.¹⁰⁸ In addition, fibrinolytics were administered to ~50% of patients managed with revascularization, as half of SHOCK trial patients presented to non-PCI-capable hospitals; the mortality benefit of fibrinolytics in this subgroup was less clear. In the SHOCK registry, fibrinolytic therapy was associated with a mortality reduction even in those who eventually underwent revascularization.¹⁰⁹ The impaired systemic perfusion may impede the lytic delivery to its target; fibrinolysis remains, nonetheless, effective, particularly if IABP is used.

In the SHOCK trial, 37% of patients received emergent CABG rather than PCI, and CABG was performed briskly, at a median of 2.7 hours after randomization. CABG was associated with the same survival as PCI, despite the higher prevalence of extensive CAD.¹¹⁰ However, this quick CABG is not feasible at many institutions and the CABG rate in the community is lower.

      IABP was recommended in all SHOCK trial patients, including medically treated patients and those initially presenting to a non-PCI hospital, and was used in 86% of patients. In the SHOCK registry, IABP was associated with a reduced mortality.¹⁰⁹ However, randomized trials such as CRISP-AMI failed to show a benefit of IABP in STEMI patients with LV failure,¹¹¹ and the IABP-SHOCK II trial failed to show a benefit of IABP even in MI patients with cardiogenic shock.¹¹² The failure of the IABP-SHOCK II trial may be due to the heterogeneity of cardiogenic shock and the inclusion of patients whose shock was not purely related to LV dysfunction (median EF was 35%, 33% of patients had NSTEMI), or whose organ failure was advanced (45% had post-cardiac arrest shock); also, IABP was mostly placed after rather than before PCI. In patients with true LV shock, it is reasonable to place the IABP before PCI as an early measure to stabilize the patient, unload the LV and reduce O₂ demands, and perform a safer PCI with potentially less reperfusion injury. Also, outside shock, IABP is useful in patients with a large STEMI who have persistent ischemia/slow coronary flow after primary PCI.

      In sum, the following strategy is recommended in cardiogenic shock:

      1 Emergent revascularization of the infarct-related artery only, irrespective of the time of presentation (even >24 hours after MI onset). Ongoing ischemia and necrosis are typical of the vicious cycle that characterizes shock, even late shock. Non-culprit artery PCI is preferably avoided as it adds contrast load and procedural time and therefore worsens LV volume overload and possibly myocardial injury (CULPRIT-SHOCK trial, STEMI [62%] or NSTEMI).69 Even patients with RCA culprit and severe LAD non-culprit do not appear to benefit from non-culprit PCI. This does not apply to cases without a clearly identifiable culprit, especially NSTEMI, where multivessel PCI may be required. Also, non-culprit PCI may be required in the patient with severe and refractory cardiogenic shock requiring multiple vasopressors despite culprit PCI, although a ventricular assist device is probably more useful in this case.

      2 In patients presenting to non-PCI-capable hospitals within 12 hours of MI onset, administer fibrinolytics whenever transfer delays are expected. Also, IABP may be placed before transfer. Augmentation of blood pressure with IABP or vasopressors may increase coronary flow and facilitate fibrinolysis.

      3 In patients with multivessel CAD on angiography:Severe three-vessel CAD (>90%) or critical left main disease dictates immediate CABG if possible. Alternatively, PCI of only the infarct-related artery is performed. Balloon angioplasty of the culprit artery followed by early CABG is another alternative.

      4 IABP may be placed during or before the revascularization procedure and may be tailored to patients who continue to decline and require higher inopressor doses, yet are not in extremis and are not post-cardiac arrest.

      5 Mechanical ventilation is often necessary to reduce the respiratory work, improve oxygenation, and reduce LV preload and afterload.

      6 Inotropic support: dobutamine is used in patients whose SBP is >80 mmHg, while norepinephrine is required in patients with SBP <80 mmHg or inappropriately low or normal SVR. While vasoconstriction may be harmful, the maintenance of an appropriate systemic perfusion pressure, including coronary perfusion pressure, is a priority and justifies the use of