Inotropes and Vasopressors in Acute Heart Failure, when the Devil Dresses as an Angel

I have more respect for a man who lets me know where he stands, even if he's wrong, than the one who comes up like an angel and is nothing but a devil. (Malcolm X). Treatment with inotropes for acute heart failure (HF) remains controversial. However, the 2016 European guidelines1 confirm that inotropes and vasopressors are in class III (not to be used) for patients with systolic blood pressure (SBP) > 90 mmHg, or who are not in cardiogenic shock. Nevertheless, many registries have shown that these drugs continue to be used in 15–40% of patients with HF who have no signs of hypoperfusion and often a SBP ≥ 110 mmHg.2-5 Thus the first question is: ‘Why does this occur?’ The international guidelines suggest not to use this class of drugs in patients with acute HF who do not have clinical signs of hypoperfusion or cardiogenic shock. Although these guideline indications are not based on the results of randomized clinical trials, they come from evidence derived from several registries, meta-analyses, retrospective studies, and national or multinational databases. Possibly, in clinical practice, physicians do not consider these investigations sufficiently convincing, or believe they do not have alternative drugs available to treat patients with acute HF, when diuretics are ineffective. Moreover, inotropes and vasopressors have been used for more than 30 years in emergency care units, and both nurses or physicians are familiar with dosages, clinical responses and possible complications of these medications. Indeed, the complex relationship between severity of the patient haemodynamic profile, cardiovascular mortality, and effects of inotropic drugs continues to represent a challenge in the statistical interpretation of the possible cause–effect relationship between these medications and clinical results. However, all the available data in the recent literature go in the same direction and show that, particularly when patients in cardiogenic shock are excluded, inotropes and vasopressors are harmful in the short term, with a direct increase in in-hospital mortality, as well as a long-term increase in both rehospitalization and total mortality rates. In the present study, Mebazaa et al.6 report the results of an investigation that included 6926 patients admitted for unscheduled hospitalization for acute HF. This large study population was enrolled in 21 countries, as part of the European Society of Cardiology Heart Failure Long-Term (ESC-HF-LT) Registry. Of these 6926 patients, 1304 (19%) were treated with a combination of intravenous (i.v.) vasodilators and diuretics, and 833 (12%) with i.v. inotropes and/or vasopressors. The primary endpoint was long-term all-cause mortality. The main secondary endpoints were in-hospital and post-discharge mortality. The authors concluded that vasodilators did not demonstrate any association with long-term clinical outcomes, while inotropes and/or vasopressors were associated with an increased risk of all-cause mortality, mainly related to an excess in in-hospital mortality. It is of note that these results were confirmed in each of the countries participating in the registry.6 To minimize clinical differences amongst patients treated with or without i.v. drugs, a propensity-based matching statistical approach was used in this study. More specifically, two separate matched samples were created to compare treatment with i.v. vasodilators combined with i.v. diuretics vs. i.v. diuretics alone, and treatment with i.v. inotropes and/or vasopressors vs. treatment without i.v. inotropes and/or vasopressors. Propensity score (PSc) was estimated using logistic regression, with the treatment under study as the dependent variable.6 The main criticism to this approach is that, although this statistical analysis is crucial in reducing the impact of clinical differences at baseline, this method cannot replace a randomized study. However, in the impossibility of performing randomized trials, the available statistical methods have to be used and accepted. In the reassessment of the data of the Italian IN-HF Outcome registry,7 the effect of inotrope treatment on mortality was also tested after a PSc adjustment using logistic regression. The practical implementation of PSc involved two stages. The first stage developed the PSc model, which calculates the PSc as the conditional probability of a patient receiving the treatment assignment given pre-treatment covariates. The second stage analysed the outcome variable, and three robust analytic approaches considering mortality as the dependent variable and inotropes as the predictive variable.7 In all statistical analyses, inotropes were significant predictors of in-hospital and 1-year mortality independently of baseline blood pressure at entry.7 The second and more complex question is: ‘Why are inotropes harmful?’ In the present paper, Mebazaa et al.6 do not comment about the possible mechanisms that could explain the detrimental effect of inotropes in patients with acute HF.6 Previous studies have suggested that inotropes and vasopressors, as mediators of sympathetic activation, resemble the negative effect of catecholamine stimulation on the myocardium. The cardiotoxicity induced by catecholamines has been known since the 1990s. Several hypotheses have been formulated regarding the mechanisms responsible for the toxic effects of vasopressors, including relative hypoxia, increased sarcolemma permeability, calcium overload, elevation of cAMP, activation of α-adrenergic receptors, and the formation of oxidative catecholamine metabolites.8 In the animal model, it was demonstrated that a 24-h exposure to noradrenaline stimulates apoptosis in adult rat cardiac myocytes by activation of the β-adrenergic pathway.9 Apoptosis is a genetically orchestrated, energy-dependent mechanism for the programmed removal of cells and is present in the myocardium in a variety of pathological circumstances, including HF, pressure overload, and myocardial infarction. The stimuli that modulate apoptosis in the myocardium remain to be elucidated. However, it has been shown that catecholamine stimulation, such as angiotensin II, tumour necrosis factor-α, mechanical stretch, hypoxia, and atrial natriuretic peptide can contribute to apoptosis and to the loss of myocytes, thus leading to the progression of myocardial dysfunction.10 The toxic effect of adrenergic stimulation can determine structural damage of the myocardium, which may be responsible for the long-lasting effect of inotropes on mortality. A possible limitation in the interpretation of results from registries can be the dosage of vasopressors or inotropes that can induce a toxic effect on the myocardium. High concentrations of catecholamines were required in the animal model, and no comparable doses can be translated in humans. Indeed, in most of the registries on acute HF, no information is available regarding the doses of inotropes or vasopressors used and for how long these drugs were administered to patients. For example, when it is stated that adrenaline and dopamine are associated with higher in-hospital mortality, we cannot say whether this effect is either due to excessively higher dosages, or standard dosages, or even to low dosages. We simply do not have an answer to these questions. Adrenaline is recommended to be infused at 0.01–0.02 μg/kg/min, while in many studies it was administered at a dose 10 times higher (0.2–0.3 μg/kg/min). The dose of adrenaline and concomitant use of other vasopressors and inotropes may indeed be subject to variation according to local experience and practices. However, no-one knows for certain whether a lower dose of a drug administered for a shorter period of time is less or more detrimental to patients with acute HF than a higher dosage of the same drug associated with other inotropic agents. In conclusion, the ESC-HF-LT Registry reported by Mebazaa and co-workers6 reinforces once more the concept that inotrope and vasopressor medications are strongly associated with an unfavourable outcome. This class of drug, particularly when used in patients who have been admitted for acute HF but do not have signs and symptoms of cardiogenic shock, has a negative influence on prognosis despite its initial haemodynamic support. Nevertheless, inotropes are still administered to a large percentage of patients with acute HF, SBP > 90 mmHg and cardiac output presumably preserved. The finding that the ominous effects of these drugs are amplified in such patients suggests that therapeutic strategies different from i.v. inotropes or vasopressors should be selected when attempts to reduce both central and peripheral congestion with diuretics are ineffective. The data available in the literature, including those obtained from the ESC-HF-LT Registry, are not powerful enough to suggest possible differences amongst inotropic or vasopressor drugs, and important uncertainties regarding dosages still persist. However, data already available from registries and uncontrolled, non-randomized studies are convincing enough regarding the detrimental effects of inotropes and vasopressors used improperly in patients without systolic arterial hypotension or cardiogenic shock. Therefore, all inotropes and vasopressors should be administered with caution in the belief that those initial benefits with an haemodynamic improvement that ‘seem to be due to an angel’, may later be paid as the ‘unfavourable effects of the devil’. Conflict of interest: none declared.