Current real-life use of vasopressors and inotropes in cardiogenic shock - adrenaline use is associated with excess organ injury and mortality

Cardiogenic shock (CS) is a devastating complication of not only acute myocardial infarction (AMI) but also other cardiac emergencies [1]. While the prognosis of CS caused by AMI has improved, possibly due to increasing use of primary percutaneous coronary intervention (PCI) as early revascularization, mortality still remains unacceptably high [2, 3].

Cardiogenic shock is a result of severe impairment of cardiac output, and also of neurohormonal and cytokine activation, leading to systemic inflammatory response syndrome [1]. Fluid resuscitation, vasopressor and inotrope medications, and mechanical circulatory support have been used to stabilize severely impaired hemodynamics. However, the most commonly used form of mechanical support, the intra-aortic balloon pump (IABP) was not demonstrated to be of benefit in the recent IABP SHOCK II trial [4, 5]; nor have other devices yet been demonstrated to improve the prognosis of patients in CS [6]. Thus, use of vasopressors and inotropes remains a cornerstone of hemodynamic support in CS [7‐10]. There are few data, however, to guide clinicians in the choice of specific agents. The overall benefit of vasoactive medications is unclear and safety concerns remain, as they may lead to increased myocardial oxygen and energy consumption. Moreover, they can have cardiotoxic effects and provoke arrhythmias [11‐13], for instance, via beta-adrenergic stimulation resulting in excessive increase in free cytosolic Ca²⁺ and Ca²⁺ channel activation in cardiac myocytes.

We analyzed the current real-life use of vasoactive medications in a prospective observational study of patients with CS of various etiological causes. The associations between outcome and the use of vasopressors and inotropes were assessed to detect possible differences in safety profiles of vasopressor and inotropic therapy. We also assessed changes in hemodynamic parameters and biomarkers associated with the use of these medications.

In the CardShock study (Clinicaltrials.gov identifier: NCT01374867) 219 patients with CS were prospectively enrolled at nine hospitals in eight European countries (Czech Republic, Denmark, Finland, Greece, Italy, Poland, Portugal, and Spain) in collaboration with the Global REsearch on Acute Conditions Team (GREAT) network. Recruitment started in October 2010 and ended on 31 December 2012 [14].

In addition to an acute cardiac cause, the inclusion criteria consisted of systolic blood pressure (SBP) <90 mmHg (in the absence of hypovolemia or after adequate fluid challenge) for at least 30 minutes, or need of vasopressor therapy to maintain adequate perfusion pressure, and signs of hypoperfusion (any of the following: altered mental status/confusion, cold periphery, oliguria, blood lactate >2 mmol/l). Patients had to be over 18 years of age and be included within 6 h of the first identification of shock. Informed consent was obtained from patients for study participation (apart from Copenhagen; see “Acknowledgements”). The main exclusion criteria were CS after cardiac surgery and ongoing hemodynamically significant arrhythmia as the cause of hypotension.

We recorded basic demographic data, previous medical history, and clinical, biochemical, and hemodynamic parameters on detection of shock and until 96 h after the study baseline. In addition cardiac index was registered in patients with a pulmonary artery catheter (n = 82 (38 %)). The etiology of CS was classified as acute coronary syndrome (ACS) or non-ACS; ACS was defined as ST elevation myocardial infarction (STEMI) or myocardial infarction without ST elevation (non-STEMI). Management of CS was registered in detail including intravenous medications, mechanical circulatory support, and ventilatory support. The use of vasoactive medications was recorded during the first 96 h after the detection of shock.

Noradrenaline, adrenaline, dopamine, vasopressin, and terlipressin were categorized as vasopressors, whereas dobutamine, levosimendan, and milrinone and enoximone (phosphodiesterase 3 inhibitors (PDE3i)) were categorized as inotropes. In addition to the use of each separate vasoactive drug, we analyzed the simultaneous use of different vasopressors and the combinations of vasopressors and inotropes. Serial blood samples were collected in 178 patients at 0 h, 12 h, 24 h, 48 h, 72 h and 96 h; plasma was immediately frozen and stored at −80 °C. Creatinine, high-sensitivity troponin T (hsTnT), and N-terminal pro-B-type natriuretic peptide (NT-proBNP) (Roche Diagnostics, Basel, Switzerland) were analyzed centrally from these samples. The endpoint of interest was 90-day all-cause mortality; three patients had missing follow-up data after hospital discharge. The CardShock study was approved by local ethics committees at the participating centers (see “Acknowledgements”) and conducted in accordance with the Declaration of Helsinki.

Patient characteristics are shown in Table 1. A comprehensive description of the study population has been published previously [14]. Briefly, the mean age of patients was 67 (12) years and 26 % were women. On average, SBP was 78 (14) mmHg, mean arterial pressure 57 (11) mmHg, and heart rate 90 (28) beats/minute. LVEF was markedly reduced, 33 (14) %. ACS was the most common etiological cause of CS (81 % patients), and 84 % of these patients had STEMI. The characteristics of patients with CS of ACS and non-ACS etiology have been described recently [14]. Overall 90-day mortality was 41 %. Compared with survivors, non-survivors were older, more frequently had previous medical history of, for example, myocardial infarction, CABG, and renal insufficiency, and had lower LVEF, a worse biomarker profile, and less frequently sinus rhythm at baseline (Additional file 1). Furthermore, 60 patients (28 %) were resuscitated prior to inclusion.

This study describes the contemporary use of vasoactive medications and their association with 90-day survival in an unselected patient population with CS. Combined use of vasopressors and inotropes was common, mostly noradrenaline with either dobutamine or levosimendan. Adrenaline, regardless of the maximum infusion rate, was consistently associated with worse outcome. In addition, despite reaching similar hemodynamic stabilization as in patients treated with other vasopressors, patients treated with adrenaline had significant worsening in cardiac and renal biomarker profiles. We observed no difference in 90-day mortality between patients who received combination of dobutamine and noradrenaline and those who received levosimendan in combination with noradrenaline.

In light of previous data suggesting that adrenaline is less safe and has more side effects than other vasoactive medications, the use of adrenaline in the present study was unexpectedly frequent [13, 20, 21]. While adrenaline is recommended for resuscitation during cardiac arrest, most patients treated with adrenaline in the CardShock study, however, had not been resuscitated. More importantly, our study shows that adrenaline is associated with 90-day mortality independent of prior cardiac arrest, and even after further multivariable adjustment or using propensity score methods. Furthermore, compared with use of other vasopressors, adrenaline is associated with marked aggravation of cardiac stress, myocardial injury and kidney dysfunction during the 4 days following detection of shock. Strikingly, these associations remained consistent in the subgroup of patients with no prior cardiac arrest. This may reflect an increase in myocardial oxygen consumption, excessive vasoconstriction and/or direct organ (cardiac, kidney, or other) toxic effects due to intense adrenergic stimulation [22, 23]. Based on these findings, alternative treatment strategies seem preferable. These would include the use of other vasopressors and/or inotropes, or more advanced therapy, such as mechanical circulatory support, in CS patients requiring potent hemodynamic stabilization. The benefit of such strategies also needs to be promptly and properly investigated.

Noradrenaline was the most commonly used vasopressor; a finding in line with the current recommendations [7, 8, 22, 24]. Dopamine was also given to one fourth of patients. Although commonly classified as a vasopressor [8, 22, 25], most patients actually received it at low-intermediate, or “renal-inotropic” doses [26]. However, the drug has failed to show beneficial effects on outcome [27, 28]. Moreover, compared to noradrenaline it has a weaker vasopressor effect and more adverse effects [29, 30]. A randomized study comparing dopamine with noradrenaline in shock showed that arrhythmia was more frequent in the dopamine group, the drug was discontinued more often due to severe arrhythmia, and the outcome was worse in the subgroup with CS [11]. As the present study was observational, the management of patients was at the discretion of the local physicians in charge. Lack of randomized data and evidence on the benefits of different vasoactive medications, and concern about possible adverse effects make the choice between different agents challenging [20, 31]. In addition, local practices in the choice of vasoactive medication may vary, as has been reported in studies of acute heart failure [32, 33].

The paucity of scientific data on the use of inotropes and their effect on mortality in AMI complicated by CS is highlighted by a recent Cochrane review [34]. While vasopressors are indicated for correcting low perfusion pressure, inotropes are used to increase cardiac output in CS. A large retrospective analysis of three cohorts with acute heart failure suggested that combining a vasopressor with an inodilator (i.e., dobutamine, levosimendan, or PDE3i) in CS was associated with lower mortality compared to vasopressors alone [35]. In turn, our study showed that the combination of noradrenaline with levosimendan was not associated with excessive mortality. Then again, while the combination of dobutamine and noradrenaline was associated with increased mortality in unadjusted analysis, the adjusted analyses showed a similar outcome for the combination of noradrenaline with either dobutamine or levosimendan. More importantly, both achieved similar hemodynamic stabilization with no clinically relevant differences in serial cardiac or renal biomarkers. Altogether, levosimendan and dobutamine appeared to be equally useful alternatives to be combined with noradrenaline in the setting of unselected patients with CS. Considering the deleterious outcome related to adrenaline, a combination of dobutamine or levosimendan with noradrenaline could be the preferred choice in patients needing inotropic support.

There are some limitations to be acknowledged. First, there was no formal standardization of management in the CardShock study. However, the primary goal was to describe the current use of vasopressors and inotropes in CS and data on vasoactive treatments were prospectively collected. Second, the total dose of vasoactive medications, and duration of the maximum dose might have given further depth to the interpretation of data. However, these details were not registered. Third, the numbers of patients in the treatment groups including adrenaline or levosimendan were limited, and caution in the interpretation of the results is advocated. As the study lacks randomization, confounding by indication is a possible bias when assessing possible effect of adrenaline on mortality. Propensity score methods were used to minimize this bias; however, these methods allowed us only to account for the measured variables and the estimates of treatment effect may be susceptible to bias due to unknown and unmeasured confounding variables. Nevertheless, the association between adrenaline use and poor outcome seemed consistent. Finally, we classified dopamine as a vasopressor, although actual doses used and combining with other vasopressors might suggest a pursuit of “renal-preserving” or inotropic effect.

Vasopressors and/or inotropes are almost invariably used in the treatment of CS and were initiated soon after detection of shock. Adrenaline, used alone or in combination with other vasoactive medication, was associated with worsening of cardiac and renal injury and increased mortality, raising questions about the safety of this treatment. In turn, levosimendan in combination with noradrenaline, was associated with a more positive outcome, and dobutamine with noradrenaline appeared to be a prognostically equal alternative; these combinations should be favored in the early management of CS. Because our study does not prove causality, our findings underscore the need for randomized controlled trials of adrenaline versus noradrenaline in CS.

ACS, acute coronary syndrome, AMI, acute myocardial infarction, CABG, coronary artery bypass graft; CS, cardiogenic shock, hsTnT, high-sensitivity troponin T, IABP, intra-aortic balloon pump, LVEF, left ventricular ejection fraction, non-STEMI myocardial infarction without ST elevation; NT-proBNP, N-terminal pro-B-type natriuretic peptide, PCI, percutaneous coronary intervention, PDE3i, phosphodiesterase 3 inhibitor, SBP, systolic blood pressure; STEMI, ST elevation myocardial infarction