Need for Critical Reappraisal of Intra-aortic Balloon Counterpulsation

Despite considerable organizational, technological, and pharmacological advancements in the treatment of patients with acute ST-segment elevation myocardial infarction (STEMI) and even with percutaneous coronary intervention (PCI), the efficacy of reperfusion is still suboptimal, mortality remains high, and novel therapeutic interventions are needed. Intra-aortic balloon counterpulsation (IABC) was first used to treat cardiogenic shock¹ in 1968; since then, it has been used in various clinical conditions to provide mechanical cardiac assistance. Use of IABC is associated with immediate hemodynamic effects leading to increased diastolic pressure, increased coronary perfusion pressure, and reduced left ventricular afterload. In the setting of STEMI, left ventricular unloading by IABC may prevent early infarct extension and ventricular remodeling.² Experimental studies have shown that IABC reduces infarct size.³

Currently, IABC is widely used in patients with STEMI complicated by cardiogenic shock. In addition, IABC has been used as a circulatory support in PCI procedures for patients at high risk for hemodynamic instability such as those with complex coronary lesions, post-MI refractory angina, and severely compromised left ventricular function as well as an intervention in the last remaining vessel and in unprotected left main coronary artery disease. Nevertheless, the issue of IABC use during high-risk PCI procedures or in STEMI remains controversial. The report of the Counterpulsation to Reduce Infarct Size Pre-PCI Acute Myocardial Infarction (CRISP AMI) trial by Patel and colleagues⁴ in this issue of JAMA helps to clarify this controversy.

CRISP AMI was an open-label, multicenter randomized controlled trial that included 337 patients with STEMI involving the anterior wall who presented within 6 hours of chest pain onset and without cardiogenic shock. Patients were randomly assigned to receive IABC, which was placed prior to PCI and was continued for at least 12 hours, or primary PCI alone. The primary outcome of the trial was infarct size measured by cardiac magnetic resonance imaging. The mean infarct size (estimated a median of 4 days after the onset of symptoms) was 42.1% of the left ventricle in patients assigned to IABC plus PCI and 37.5% of the left ventricle in the patients assigned to PCI alone (P = .06), showing no benefit with IABC. No significant differences between the IABC plus PCI group and the PCI alone group were observed regarding clinical end points. The authors concluded that the routine use of IABC in patients with anterior wall STEMI without cardiogenic shock does not lead to a reduction in infarct size or to an improvement in clinical outcomes at 6 months.⁴

Autopsy studies have shown that cardiogenic shock is generally associated with loss of 40% or more of the left ventricular mass,^{5 - 6} and the association between necrosis of 40% or more and the development of cardiogenic shock has been accepted in the field of cardiology during the last 40 years. Therefore, the finding in the CRISP AMI trial that the mean infarct size estimated by cardiac magnetic resonance imaging was 40% and patients were hemodynamically stable at presentation seems surprising. Moreover, infarct size was recorded after primary PCI, a reperfusion strategy that leads to an average salvage index (proportion of area at risk salvaged by reperfusion) of about 50% according to studies with serial measurements of both the initial area at risk and the final infarct size after reperfusion.⁷ Based on the T2-weighted imaging obtained from the single magnetic resonance imaging study performed after reperfusion, the authors computed a mean salvage index of 35%. Although the reasons underlying these findings remain largely unclear, possible explanations include an overestimation of infarct size by cardiac magnetic resonance imaging,⁸ premature timing of imaging at a point when the recovery of myocardial perfusion is not complete,⁹ or dissociation of the link between myocardial necrosis and the development of cardiogenic shock in the contemporary era of STEMI care.

Five previous randomized trials have assessed the role of IABC in patients with AMI without cardiogenic shock.^{10 - 14} However, neither these trials nor the CRISP AMI trial were sufficiently powered for the evaluation of clinical outcomes with IABC, with only 21 deaths among 518 patients in the IABC groups and 21 deaths among 536 patients in the control groups across the studies.^{10 - 14}

Explanations for failure of IABC to promote myocardial salvage and improve clinical outcome in patients with STEMI and no cardiogenic shock remain speculative. Although the potential for an increase in infarct size due to the extra time needed to insert the IABC cannot be entirely refuted, the effect of a 10-minute time delay in patients receiving IABC seems to be modest and not of sufficient magnitude to offset the beneficial effects of IABC. The suggestion in the CRISP AMI trial that the left ventricular unloading by IABC might have occurred too late after symptom onset to salvage considerable amounts of myocardium at risk contradicts a recent study that demonstrated considerable myocardial salvage by primary PCI at a much later time.¹⁵

Other putative explanatory mechanisms may also be offered. There are indications that the increase in the diastolic perfusion pressure may not lead to an increase in the coronary blood flow. Provided that coronary autoregulation is intact, the variation in the coronary perfusion pressure across a wide range (from 25-125 mm Hg) does not lead to changes in the coronary blood flow.² A randomized angiographic study of predefined high-risk patients with AMI but without shock undergoing urgent PCI did not demonstrate any improvement in coronary flow, left ventricular function, ST-segment recovery, or clinical outcome in patients allocated to treatment with IABC.¹⁴ On the other hand, even if coronary blood flow is increased, it may lead to increased cardiac contractility and myocardial oxygen consumption due to an increased stretch of myocardial fibers (Gregg phenomenon).¹⁶ To what extent this mechanism may be operative in the setting of IABC use remains unknown.

In addition, primary PCI with routine use of modern antithrombotic drugs and frequent use of thrombectomy is a highly effective reperfusion strategy; so far it has been difficult to achieve further significant improvement in myocardial salvage capacity with adjunctive interventions. A meta-analysis of observational studies of patients with STEMI complicated by cardiogenic shock demonstrated a beneficial effect of IABC adjunctive to thrombolysis but not to primary PCI; however, the latter finding was subject to considerable confounding by the younger age of patients, patient crossover, and selection bias.¹⁷

The clear message from the CRISP AMI trial is that the routine use of IABC neither reduces infarct size nor improves clinical outcome among patients with STEMI without cardiogenic shock; accordingly, use of this device should be discouraged in these patients. For patients with STEMI and cardiogenic shock, IABC remains a frequently used intervention. An ongoing randomized trial plans to enroll 600 patients with STEMI and shock (clinicaltrials.gov Identifier: NCT00491036) and aims to provide evidence to support or refute the strong recommendation for the use of IABC contained in the current practice guidelines.¹⁸

Other research avenues for the treatment of patients with STEMI remain attractive. Organizational efforts to increase the availability of primary PCI and reduce ischemia time remain of paramount clinical importance. Furthermore, the no-reflow phenomenon and microvascular dysfunction following primary PCI for STEMI remain prime candidates for future research due to their high incidence and strong association with mortality.