OUP user menu

Employing vasopressin during cardiopulmonary resuscitation and vasodilatory shock as a lifesaving vasopressor

Volker Wenzel, Karl H. Lindner
DOI: http://dx.doi.org/10.1016/S0008-6363(01)00262-0 529-541 First published online: 15 August 2001


Epinephrine during cardiopulmonary resuscitation (CPR) is being discussed controversially due to its β-receptor mediated adverse effects such as increased myocardial oxygen consumption, ventricular arrhythmias, ventilation-perfusion defect, postresuscitation myocardial dysfunction, ventricular arrhythmias and cardiac failure. In the CPR laboratory simulating adult pigs with ventricular fibrillation or postcountershock pulseless electrical activity, vasopressin improved vital organ blood flow, cerebral oxygen delivery, resuscitability, and neurological recovery better than did epinephrine. In paediatric preparations with asphyxia, epinephrine was superior to vasopressin, whereas in both paediatric pigs with ventricular fibrillation, and adult porcine models with asphyxia, combinations of vasopressin and epinephrine proved to be highly effective. This may suggest that a different efficiency of vasopressors in paediatric vs. adult preparations; and different effects of dysrhythmic vs. asphyxial cardiac arrest on vasopressor efficiency may be of significant importance. Whether these theories can be extrapolated to humans is unknown at this point in time. In patients with out-of-hospital ventricular fibrillation, a larger proportion of patients treated with vasopressin survived 24 h compared with patients treated with epinephrine; during in-hospital CPR, comparable short-term survival was found in groups treated with either vasopressin or epinephrine. Currently, a large trial of out-of-hospital cardiac arrest patients being treated with vasopressin vs. epinephrine is ongoing in Germany, Austria and Switzerland. The new CPR guidelines of both the American Heart Association, and European Resuscitation Council recommend 40 U vasopressin intravenously, and 1 mg epinephrine intravenously as equally effective for the treatment of adult patients in ventricular fibrillation; however, no recommendation for vasopressin was made to date for adult patients with asystole and pulseless electrical activity, and paediatrics due to lack of clinical data. When adrenergic vasopressors were unable to maintain arterial blood pressure in patients with vasodilatory shock, continuous infusions of vasopressin (∼0.04 to ∼0.1 U/min) stabilised cardiocirculatory parameters, and even ensured weaning from catecholamines.

  • Adrenergic (ant)agonists
  • Defibrillation

Time for primary review 21 days.

1 Epinephrine's 100 year history for CPR turns out to be controversial

Morbidity and mortality after successful resuscitation from cardiac arrest largely depends on recovery of neurologic function [1]. Accordingly, efficiency of an intervention during cardiopulmonary resuscitation (CPR) has to be judged with regard to its effects on neurologic outcome [2]. The American Heart Association [3] and the European Resuscitation Council [4] have acknowledged a lack of clinical evidence to recommend epinephrine during CPR in previously published guidelines; however, no change was made in the 1992 and 1998 recommendations due to a lack of proven alternative drugs or interventions.

Laboratory studies employing epinephrine during CPR showed increased β-receptor-mediated myocardial oxygen consumption [5], ventricular arrhythmias [6], a ventilation-perfusion defect [7] that was possibly caused by attenuated hypoxic pulmonary vasoconstriction [8], postresuscitation myocardial dysfunction, ventricular arrhythmias, and cardiac failure [9]. Another explanation for the adverse effects of epinephrine during CPR is a significant reduction of the myocardial action potentials; it was shown that epinephrine increased the probability that re-entry action potentials hit excitable tissue, which may provoke or even stabilise ventricular fibrillation [10].

In one clinical study from Australia, epinephrine did not result in better short-term survival or hospital discharge rate compared with saline placebo [11]. Further, high-dose epinephrine did not improve hospital discharge rates in several large multicenter clinical trials [12–17]. In a study in the Vienna General Hospital [18], patients with an unfavourable vs. favourable cerebral performance category received a cumulative dose of 4 vs. 1 mg epinephrine during CPR; this finding persisted after stratification by duration of resuscitation. Accordingly, it is disappointing that a drug such as epinephrine, which is being used for resuscitation for more than 100 years [19,20], has controversial safety and efficacy. Given these potential adverse epinephrine-mediated effects, an American researcher asked in an editorial [21] “have we ridden the adrenergic horse as far as it will take us?”

2 Vasopressin, an endogenous stress hormone

Vasopressin, a hormone that is called an antidiuretic hormone as well, has a long evolutionary history. With the emergence of life on land, vasopressin became the mediator of a remarkable regulatory system for the conservation of water. Natural vasopressin forms are nonapeptides with two cysteine residues forming a bridge between positions one and six. The integrity of this disulfide bond is essential for its biological activity, and amino acid substitutions dictate physiologic actions such as alterations of antidiuretic, or vasopressor function. Interestingly, vasopressor effects were observed as early as 1895; vasopressin acts directly via V1-receptors on contractile elements — an effect that cannot be reversed by adrenergic blockade or denervation [22] (Table 1). Primary indication of vasopressin and its analogues so far is management of hypothalamic diabetes insipidus, and treating bleeding esophageal varices in some cases.

View this table:
Table 1

Chemical structures and activities of antidiuretic peptides/vasopressin analoguesa

PeptideSimplified amino acid structureActivity in relation toComment
arginine vasopressin
Arginine vasopressinCys–Tyr–Phe–Glu–Asp–Cys–Pro–l-Arg–Gly–(NH2)100100ADH/AVP (Pitressin®), present in mammals
Lysine vasopressinCys–Tyr–Phe–Glu–Asp–Cys–Pro–Lys–Gly–(NH2)8060LVP (Lypressin®), present in swine
OxytocinCys–Tyr–Ile–Glu–Asp–Cys–Pro–Leu–Gly–(NH2)11Induces uterine contraction
Ornithine vasopressinCys–Tyr–Phe–Glu–Asp–Cys–Pro–Orn–Gly–(NH2)2290POR 8®, treatment of esophageal varices
DDAVPbCys–Tyr–Phe–Glu–Asp–Cys–Pro–d-Arg–Gly–(NH2)12000.39Desmopressin®, increases factor VIII activity
  • a All peptides shown have disulfide bonds between amino acids at positions one and six.

  • b Deaminated on position 1 (synthetic peptide). Primary indication of vasopressin and its analogues is management of hypothalamic diabetes insipidus, and treating bleeding esophageal varices in some cases; 400 U arginine vasopressin=1 mg; unless otherwise noted, vasopressin discussed in this article reflects arginine vasopressin.

A number of fundamental endocrine responses of the human body to cardiac arrest and CPR have been investigated in the past 10 years [23–25]. Circulating endogenous vasopressin concentrations were high in patients undergoing CPR, and levels in successfully resuscitated patients have been shown to be significantly higher than in patients who died (Table 2) [26]. This may indicate that the human body discharges vasopressin as an adjunct vasopressor to epinephrine in life-threatening situations such as cardiac arrest in order to preserve homeostasis. In a clinical study of 60 out-of-hospital cardiac arrest patients, parallel increases in plasma vasopressin and endothelin during CPR were found only in surviving patients. Both before and after epinephrine administration, plasma epinephrine and norepinephrine concentrations were significantly higher in patients who died when compared with surviving cardiac arrest victims [27]. Thus, plasma concentrations of vasopressin may have a more important effect on CPR outcome than previously thought; and prompted several investigations to assess its role for possible CPR management in order to improve CPR management.

View this table:
Table 2

Endogenous vasopressors during basic life support CPR, and advanced cardiac life support CPRa

HormoneBefore epinephrine by EMS physicianAfter epinephrine by EMS physician
Survivors (n=12)Nonsurvivors (n=14)Survivors (n=24)Nonsurvivors (n=36)
Endothelin (0.6–1.2 pg/ml)4.3±0.95.5±0.44.0±1.05.3±0.5
Epinephrine (0.05–0.2 ng/ml)14.1±2.025.3±3.6**145±16‡201±21*,‡
Norepinephrine (0.25–0.5 ng/ml)5.0±0.98.4±1.1*3.9±0.98.3±1.1**
Vasopressin (1–3 pg/ml)193±2870±9***177±2758±9***
Adrenocorticotropin (25–50 pg/ml)128±3457±6*234±92†85±9***
Cortisol (5–25 μg/dl)18±315±217±213±2
  • a Values are given as mean±S.E.M.; normal ranges of hormones are given in parentheses. Note that both before and after injection of epinephrine by the emergency medical service physician, vasopressin plasma levels in surviving patients were significantly higher than in non-survivors. Adapted from Lindner et al., Heart 1996;75:145–150. *, P<0.05; **, P<0.01; ***, P<0.001 for comparisons between survivors and nonsurvivors; P, <0.01; , P<0.001 before and after epinephrine.

3 Vasopressin during CPR in laboratory models: assessment of individual organ function

During ventricular fibrillation, a dose–response investigation of three vasopressin dosages (0.2, 0.4, 0.8 U/kg) compared with the maximum effective dose of 200 μg/kg epinephrine showed that 0.8 U/kg vasopressin was the most effective drug in regards of increasing vital organ blood flow [28] (Fig. 1). Also, vasopressin significantly improved cerebral oxygen delivery, and ventricular fibrillation mean frequency during CPR when compared with a maximum dose of epinephrine [29] (Figs. 2–4). Furthermore, effects of vasopressin on vital organ blood flow lasted longer after vasopressin than after epinephrine (∼4 vs. ∼1.5 min); significantly more vasopressin animals could be resuscitated; and vasopressin did not result in bradycardia after return of spontaneous circulation (Fig. 5) [30]. Interestingly, the combination of vasopressin and epinephrine vs. vasopressin only resulted in comparable left ventricular myocardial blood flow, but in significantly decreased cerebral perfusion [31]. The binding of both vasopressin and epinephrine to its receptors causes characteristic changes such as intracellular concentration of phosphatidylinositol and calcium [32,33]. In fact, a rodent study evaluating administration of vasopressin, norepinephrine, and a combination of vasopressin and norepinephrine showed that V1- and α-adrenergic receptors saturated the same intracellular transduction pathway [34]. Although speculative, this mechanism may have hampered nitric oxide release in the cerebral vasculature induced by vasopressin, and therefore, suppressed cerebral perfusion in our animals receiving a combination of vasopressin and epinephrine. These results are striking, since epinephrine selectively spares the cerebral circulation from vasoconstriction when administered during CPR alone.

Fig. 5

Porcine laboratory model of prolonged cardiac arrest of 15 min with postcountershock pulseless electrical activity. Left ventricular myocardial, and total cerebral blood flow before, 90 s, and 5 min after vasopressin (■) or epinephrine (□) administration during CPR. Each box plot represents median, 25th and 75th percentiles, and minimum and maximum values of eight measurements in both groups; *, P<0.05 vs. vasopressin. Note that vasopressin during CPR may increase myocardial blood flow to ∼60% of prearrest values (normal left ventricular myocardial blood flow in pigs, ∼200 ml/min/100 g), but brain perfusion even to ∼150% of prearrest values (normal brain blood flow in pigs, ∼50 ml/min/100 g). Reprinted with permission from Wenzel et al., Crit Care Med 1999;27:486–492. Copyright Lippincott Williams & Wilkins 1999.

Fig. 4

Power spectrum and 10 s ECG ventricular fibrillation pattern of a laboratory pig during basic life support CPR (before vasopressin), and during advanced cardiac life support CPR 90 s after vasopressin. Note that vasopressin induces an increase of ventricular fibrillation mean frequency, which is validated by the rightward shift of the power spectrum (a, c); and increase in ECG amplitude (b, d). Reprinted with permission from Achleitner et al., Anesth Analg 2000;90:1067–1075. Copyright International Anesthesia Research Foundation 2000.

Fig. 3

Cerebral oxygen delivery and consumption indices, and oxygen extraction ratio during cardiopulmonary resuscitation before, and 90 s and 5 min after intravenous administration of 200 μg/kg epinephrine (maximum effective dosage in pigs; □) and 0.4 U/kg vasopressin (medium dosage in pigs; ■). Each box plot represents median, 25th and 75th percentiles, and minimum and maximum values. *, P<0.05 vs. epinephrine; , P<0.01 vs. epinephrine; , P<0.05 vs. before drug administration (DA). Note that the superior cerebral blood flow with vasopressin lead to a greater cerebral oxygen consumption, but this was not the result of an overstimulation of metabolism because cerebral oxygen extraction fell, and oxygen uptake became independent of oxygen delivery. Reprinted with permission from Prengel et al., Stroke 1996;27:1241–1248. Copyright 1996 American Heart Association.

Fig. 2

Cerebral perfusion pressure, and total cerebral blood flow during cardiopulmonary resuscitation before and 90 s and 5 min after intravenous administration of 200 μg/kg epinephrine (maximum effective dosage in pigs; □) and 0.4 U/kg vasopressin (medium dosage in pigs; ■). Each box plot represents median, 25th and 75th percentiles, and minimum and maximum values. *, P<0.05 vs. epinephrine; , P<0.01 vs. epinephrine; , P<0.05 vs. before drug administration (DA). Note that even the medium vasopressin dose induced a significant increase in cerebral blood flow, which is most likely induced by a nitric oxide-mediated cerebral vasodilatation. Reprinted with permission from Prengel et al., Stroke 1996;27:1241–1248. Copyright American Heart Association 1996.

Fig. 1

Systemic vascular resistance during closed-chest cardiopulmonary resuscitation (CPR) in pigs before, 90 s after, and 5 min after administration of 200 μg/kg epinephrine, 0.2 U/kg vasopressin (low dose), 0.4 U/kg vasopressin (medium dose), and 0.8 U/kg vasopressin (high dose), and in the 1-h postresuscitation phase. Each bar represents the mean±S.E.M. of seven observations. **, P<0.01, ***, P<0.001 between epinephrine and 0.8 U/kg vasopressin groups. VF, ventricular fibrillation; Defib, defibrillation; and ROSC, restoration of spontaneous circulation. Note that 0.8 U/kg vasopressin increases systemic vascular resistance during CPR fundamentally to levels of ∼10 000 dyness−1cm−5, which shifts blood flow from muscle, skin and gut towards the heart and brain. Reprinted with permission from Lindner et al., Circulation 1995;91:215–221. Copyright 1995 American Heart Association.

Administration of endobronchial drugs during CPR may be a simple and rapid alternative, when intubation is performed before intravenous cannulation [35], when the time interval for intravenous access is prolonged, or when attempts for intravenous access are simply unsuccessful, such as in young children with poor peripheral perfusion. A laboratory model showed that the same dose of intravenous and endobronchial vasopressin resulted into the same coronary perfusion pressure four min after drug administration [36] (Fig. 6). In contrast, the equipotent endobronchial epinephrine dose is approximately ten times higher than the intravenous epinephrine dose during CPR [37]. This investigation showed that endobronchial vasopressin is absorbed during CPR, increased coronary perfusion pressure significantly within a very short period, and increased the chance of successful resuscitation [39]. However, endobronchial drug administration may be less appropriate in children, who suffer cardiac arrest due to respiratory disorders. For example, in children suffering cardiac arrest mostly due to severe pneumonia, endobronchial drug delivery is not likely to result in adequate drug absorption, and therefore, may be rather ineffective. In such cases, the endobronchial drug delivery route may render drug absorption erratic due to pulmonary oedema and capillary leak, and therefore, may further compromise oxygenation and ventilation in children [38]. Additionally, hypovolemia can be rapidly corrected with fluids via an intraosseous catheter, but not via the endobronchial route. Accordingly, the intraosseous route has been recommended for paediatric emergency situations, and is widely taught both in the Americas and internationally by the Paediatric Advanced Life Support and Advanced Trauma Life Support courses [39]. In a porcine investigation, intraosseous vs. intravenous vasopressin resulted in comparable vasopressin plasma levels, hemodynamic variables, coronary perfusion pressure, and return of spontaneous circulation rates [40] (Fig. 7). Therefore, intraosseous vasopressin may be a valuable alternative for vasopressor administration during CPR, when intravenous access is delayed, or not available. Accordingly, our laboratory studies indicate that the same vasopressin dosage may be administered intravenously, endobronchially, and intraosseously; rendering usage of this vasopressor during CPR simple, rapid and inexpensive.

Fig. 7

Individual coronary perfusion pressure tracings after intravenous (▲) and intraosseous (■) vasopressin administration during CPR. No statistical analysis was performed, and some tracings may be superimposed on each other; defibrillation was performed approximately at 5 min and 15 s after drug administration. Intraosseous vs. intravenous vasopressin resulted at all points of the experiment in comparable coronary perfusion pressure during CPR; also, note that coronary perfusion pressure after either intravenous or intraosseous vasopressin is elevated for about 5 min, which allows ample time for successful defibrillation. Reprinted with permission from Wenzel et al., Crit Care Med 1999;27:1565–1569. Copyright Lippincott Williams & Wilkins 1999.

Fig. 6

Coronary perfusion pressure before and after endobronchial drug administration during CPR. All variables are given as mean±S.E.M.; DA, drug administration; CPP, coronary perfusion pressure; intravenous vasopressin (●); endobronchial vasopressin (▲); endobronchial saline placebo (○); *, P<0.05 compared with endobronchial saline placebo; , P<0.05 compared with endobronchial vasopressin. Note that seven of seven intravenous vasopressin, and nine of nine endobronchial vasopressin swine survived, but only two of five endobronchial saline placebo pigs (P<0.05). Reprinted with permission from Wenzel et al., Anesthesiology 1997;86:1375–1381. Copyright American Society of Anesthesiologists 1997.

Both the American Heart Association and European Resuscitation Council continue to recommend repeated administration of epinephrine during advanced cardiac life support [56,57], although it is not proven whether repeated epinephrine given during CPR may be effective, or if this strategy may even result in inadvertent catecholamine toxicity. After repeated dosages of vasopressin vs. epinephrine were administered, coronary perfusion pressure increased only after the first of three epinephrine injection, but increased after each of three vasopressin injections; accordingly, all vasopressin animals survived; whereas all pigs resuscitated with epinephrine died [41]. Long-term survival after cardiac arrest may be determined by the ability to ensure adequate organ perfusion during CPR, and in the postresuscitation phase. In the early postresuscitation phase, vasopressin administration resulted in higher arterial blood pressure, but a lower cardiac index; a reversible depressant effect on myocardial function of the vasopressin pigs was observed when compared with epinephrine. However, overall cardiovascular function was not irreversibly or critically impaired after the administration of vasopressin [42]. Renal and splanchnic perfusion may be critically impaired during [43,44] and after [45] successful resuscitation from cardiac arrest. For example, 30 min after return of spontaneous circulation, renal and adrenal blood flow were significantly lower in the vasopressin pigs as compared with the epinephrine group; pancreatic, intestinal, and hepatic blood flow were not significantly different in animals after receiving epinephrine or vasopressin [46]. It is unknown whether the vasopressin-mediated pronounced blood shift during CPR from the muscle, skin and gut towards the myocardium and brain might be deleterious for splanchnic organs, and whether this may contribute to multiorgan failure after return of spontaneous circulation. Furthermore, it is unknown whether a high bolus of vasopressin administered during CPR may result in oliguria or anuria due to its antidiuretic effects in the postresuscitation phase. In a porcine CPR investigation, vasopressin impaired cephalic mesenteric blood flow during CPR and in the early postresuscitation phase, but did not result in an anti diuretic response. Neither renal blood flow, nor renal function was influenced by vasopressin or epinephrine in this investigation [47]. Continuous infusion of low-dose dopamine mediated a significant increase in superior mesenteric blood flow after successful CPR with vasopressin by selective vasodilatation of intestinal vessels. Accordingly, administering dopamine may improve gut perfusion, and therefore, may improve gut function in the postresuscitation period [48].

In laboratory investigations of short and prolonged cardiac arrest, vasopressin resulted in significantly higher vital organ blood flow and cerebral oxygen delivery than did epinephrine [28–31]. This implies that vasopressin dilated cerebral arterioles and subsequently, resulted in superior brain perfusion throughout the neuroaxis. In the vasopressin animals, the superior cerebral blood flow lead to a greater cerebral oxygen consumption, but was not the result of an overstimulation of metabolism because cerebral oxygen extraction fell, and oxygen uptake became independent of oxygen delivery [29]. The question arises, if increased cerebral blood flow during CPR with vasopressin is beneficial in regards of neurological recovery, or detrimental due to fatal complications such as cerebral oedema after return of spontaneous circulation. Although a Glasgow-Coma Score rating was performed after successful CPR with vasopressin vs. epinephrine in preliminary clinical studies, the results could unfortunately not be attributed to a single CPR-intervention due to many confounding variables. In a porcine model simulating prolonged (22 min) of advanced cardiac life support, all vasopressin animals had return of spontaneous circulation, whereas all pigs in the epinephrine and saline placebo group died. After 24 h of return of spontaneous circulation, the only neurologic deficit of all vasopressin pigs was an unsteady gait, which disappeared within another 3 days [49]. This observation confirms that in order to achieve full recovery after cardiac arrest, both excellent management of basic and advanced cardiac life support, and careful optimisation of organ function in the postresuscitation phase are of fundamental importance. (Fig. 8) [50] Since neurodiagnostic tests such as evoked potentials, electroencephalogram, or positron emission tomography may not be able to accurately detect pathology [51], we employed cerebral magnetic resonance imaging after successful CPR. Since imaging was performed 4 days after the experiment to ensure that cerebral ischemic regions due to extracellular edema would be fully developed, the absence of cerebral cortical and subcortical edema, intraparenchymal haemorrhage, ischemic brain lesions, or cerebral infarction confirmed by T2-weighted magnetic resonance imaging indicates that the vasopressin pigs fully recovered from cardiac arrest in regards of both anatomical and physiological terms. Thus, laboratory evidence suggests that vasopressin given during CPR may be a superior drug when compared with epinephrine in order to ensure both return of spontaneous circulation and neurological outcome.

Fig. 8

Effects of vasopressin, epinephrine, and saline placebo on diastolic aortic pressure during prolonged advanced cardiac life support CPR. Repeated doses of vasopressin (♦), but not epinephrine (■) or saline placebo (▲), maintained mean±S.E.M. aortic diastolic arterial blood pressure above a threshold of about 30 to 40 mmHg (horizontal dashed lines) that renders successful defibrillation likely. DA 1 indicates drug administration of 0.4 U/kg vasopressin vs. 45 μg/kg epinephrine vs. saline placebo; DA 2, drug administration of 0.4 U/kg vasopressin vs. 45 μg/kg epinephrine vs. saline placebo; DA 3, drug administration of 0.8 U/kg vasopressin vs. 200 μg/kg epinephrine vs. saline placebo; *, P<0.05 vs. epinephrine and saline placebo, respectively; §, P<0.05 vs. saline placebo; ACLS, advanced cardiac life support; BLS, basic life support; VF, ventricular fibrillation; time is given in min (′) and seconds (″). Note that only the first of three epinephrine injections, but three of three vasopressin injections maintained coronary perfusion pressure at a level that ensured return of spontaneous circulation: Six of six vasopressin pigs survived with full neurological recovery, whereas both six of six epinephrine- and five of five saline placebo pigs died. Reprinted with permission from Wenzel et al., J Am Coll Cardiol 2000;35:527–533. Copyright American College of Cardiology 2000.

4 Clinical experience with vasopressin during cardiac arrest

In patients with refractory cardiac arrest, intravenous vasopressin induced an increase in arterial blood pressure, and in some cases, return of spontaneous circulation, where standard therapy with chest compressions, ventilation, defibrillation, and epinephrine had failed [52]. In a small (n=40) prospective, randomised investigation of patients with out-of-hospital ventricular fibrillation, a significantly larger proportion of patients treated with vasopressin were successfully resuscitated and survived 24 h compared with patients treated with epinephrine [53]. This was not designed as a mortality study, but even with such small numbers, there was a non-significant trend towards an improvement in hospital discharge rate (P=0.16). In a large (n=200) in-hospital CPR trial from Ottawa, Canada, comparable short-term survival was found in both groups treated with either vasopressin or epinephrine, indicating that these drugs may be equipotent when response times of rescuers are short [54]. In another clinical evaluation in Detroit, Michigan, four of ten patients responded to vasopressin administration after approximately 40 min of unsuccessful advanced cardiac life support, and had a mean increase in coronary perfusion pressure of 28 mmHg [55]. This is surprising, since an arterial blood pressure increase with any drug after such a long period of non-effective CPR management would be normally not expected. In this study, vasopressin caused an increase in coronary perfusion pressure, and a decrease of epinephrine plasma levels, which is similar to an animal study with increased vital organ blood flow, but decreased catecholamine plasma levels after vasopressin [56]. Although speculative, it is possible that baroreceptors in the arterial vasculature registered increased organ perfusion after vasopressin, and subsequently down-regulated catecholamine secretion — whether this may decrease epinephrine tachyphylaxis and therefore, improve subsequent catecholamine effects, remains under investigation.

The authors of the present article are currently co-ordinating a multicenter clinical study to determine the effects of 40 U vasopressin IV vs. 1 mg epinephrine IV given up to two times in out-of-hospital cardiac arrest patients with ventricular fibrillation, asystole, and pulseless electrical activity (Fig. 9). This clinical trial is conducted under the aegis of the European Resuscitation Council, and has randomised per March, 2001 a total of 770 patients in 42 EMS systems (37 ground, and 5 rotor-winged EMS programs) in Germany, Switzerland, and Austria. A preliminary analysis has revealed that the study is safe, randomisation is working properly, no adverse results were reported, and that there is a good chance to find significant results after randomising 1500 patients as planned. The new international CPR guidelines of both the American Heart Association [57] and European Resuscitation Council [58] recommend 40 U vasopressin IV and 1 mg epinephrine IV as equally effective for the treatment of patients in ventricular fibrillation; however, no recommendation was made to date for patients with asystole and pulseless electrical activity, and paediatrics due to lack of clinical data. When the aforementioned large clinical study may be concluded in Autumn 2001, we may be able to better determine the role of vasopressin vs. epinephrine for management of cardiac arrest patients presenting with ventricular fibrillation, asystole, or pulseless electrical activity [59].

Fig. 9

Study algorithm of a multicenter, blockrandomized, doubleblind, randomized controlled trial in Europe: comparison of vasopressin vs. epinephrine during cardiac arrest. A study conducted under the aegis of the European Resuscitation Council. Reprinted with permission from Krismer et al. (article in German) Notfall Rettungsmed 1999;2:478–485. Copyright Springer Verlag, Berlin, Heidelberg, New York, 1999.

Efficiency of a CPR intervention is judged with regards to its effects on hospital discharge rate, long-term survival, and especially, neurologic outcome. However, in order to determine effects of a drug given during CPR on neurological recovery is extremely difficult due to confounding variables. For example, a given patient population differs significantly due to past medical history, standard of life, age, race, cardiac rhythm at collapse, and location of cardiac arrest [60,61]. Second, the healthcare system itself may affect CPR outcome fundamentally due to differences in emergency medical services, response times, intensive care unit management, and access to diagnostic technology [62]. In order to detect a significant increase in hospital discharge rates and/or neurological recovery employing a new pharmacological CPR intervention, ∼20 000 patients, several years, and millions of US Dollars or Euros would be necessary — a trial that would be significantly bigger than large Framingham Heart studies with typically ∼3000 [63] to ∼5000 patients [64]. Thus, initiating such as study seems to be almost next-to-impossible, since healthcare and research funding is decreasing throughout the world; further, commercial manufacturers of vasopressin have not much interest in this drug since the patent of vasopressin has expired ∼50 years ago. As such, the primary endpoint of the European vasopressor study (anticipated number of patients, 1500) are effects of vasopressin vs. epinephrine on hospital admission — not the optimum, but a realistic solution.

5 Where is the future of CPR management?

We have shown beneficial effects of vasopressin vs. epinephrine in special CPR situations such as epidural anaesthesia [65], hypothermia [66], and hypovolemic shock [67] in the laboratory. In the clinical setting, we observed positive effects of vasopressin infusions in some patients with life-threatening haemorrhagic shock with collapsing arterial blood pressure, who did not respond anymore to adrenergic catecholamines. Although these observations in the laboratory and in individual patients are very promising, they have to be yet confirmed in prospective, randomised clinical trials; however, due to the relatively rare occurrence of these special CPR situations, it is unlikely that vasopressin can be studied systematically under these circumstances.

Another area that is extremely complicated to investigate, but definitely needs to be addressed, is the paediatric cardiac arrest population. Due to the lack of randomised clinical trials in paediatric emergency cardiac care, some CPR recommendations have been extrapolated from adult or laboratory studies. In a porcine preparation, epinephrine was superior to vasopressin in a paediatric CPR model of asphyxia cardiac arrest [68]. We have shown in this paediatric model that a combination of vasopressin and epinephrine was superior to either vasopressin or epinephrine alone; however, this may only apply to paediatric but not adult settings, and definitely warrants especially clinical investigations. Upon repeating the aforementioned experiment of asphyxia cardiac arrest in adult swine to determine effects of age vs. underlying cause of cardiac arrest, we confirmed that a combination of vasopressin with epinephrine was superior to either vasopressin alone, or epinephrine alone. This observation is entirely new and surprising, and may suggest that the usual approach of pharmacological CPR management to administer identical drugs and dosages for patients presenting with cardiac arrest due to dysrhythmia or asphyxia may have to be reconsidered. In fact, it is possible that when the degree of ischemia is fundamental such as during asphyxia, or advanced cardiac life support is prolonged, a combination of vasopressin with epinephrine could be beneficial. The laboratory experience of this phenomenon may be confirmed by observations from CPR efforts in eight patients who were unsuccessfully resuscitated with epinephrine for ∼10–20 min, subsequently received 40 U of vasopressin, and subsequently, all patients had return of spontaneous circulation, with three of eight patients being even discharged alive from the hospital [52]. This explanation would be in agreement with a Canadian study [54] indicating comparable outcome with vasopressin vs. epinephrine in in-hospital cardiac arrest, with a short time of ischemia.

Thus, our usual strategy according to the current CPR guidelines of employing vasopressors during CPR may not sufficiently address the cause of cardiac arrest, and degree of ischemia. Accordingly, it is possible that the ideal vasopressor and especially, the ideal dosing regimen for CPR is yet to be discovered, rendering a combination of agents possibly necessary; however, development of a CPR ‘cocktail’ may be extremely difficult due to multiple potential permutations of different drugs and dosages whenever a combination therapy is employed [69]. Laboratory experiments were performed with a combination of vasopressin, epinephrine and nitroglycerine; however, determining optimal dose–response effects is most likely difficult, and clinical experience is pending [70]. Further, vasopressors that improve initial resuscitation may not be the best drugs to improve survival in the post resuscitation phase. The ideal vasopressor for CPR is a drug that significantly increases myocardial and cerebral perfusion during CPR, and yet, if necessary, can be rapidly, but titratibly reversed in the immediate post-resuscitation phase [71]. Although the case of vasopressin looks promising at this point in time, the large clinical study in Europe would have to come out significantly in favour of vasopressin in order to further change international CPR guidelines.

The exciting news about vasopressin studies cannot cover the fact that many issues have not been addressed as of today, at least partially because commercial interest is lacking due to an expired patent from ∼50 years ago. For example, the only dose–response investigation with vasopressin stems from young healthy pigs; vasopressin experience in humans is solely based on injections of 40 U. Second, preliminary laboratory evidence suggests beneficial effects of a combination of vasopressin with epinephrine in asphyxia cardiac arrest; however, this observation is inconsistent with a report in a postcountershock pulseless electrical activity preparation — further research is necessary which drug plays which role in what situation. Third, vasopressin receptors in pigs (lysine vasopressin) and humans (arginine vasopressin) are different, which may result in a different hemodynamic response to exogenously administered arginine vasopressin. However, the circulatory effects of arginine vasopressin, as administered in all laboratory investigations, may be even greater in humans with arginine vasopressin receptors when compared with pigs with lysine vasopressin receptors. Furthermore, extrapolating experience with vasopressin from the adult to the paediatric setting seems to be difficult, and needs further significant investigation. Also, young healthy pigs that were free of atherosclerotic disease were employed in animal studies; it is therefore difficult to determine whether all laboratory results can be safely extrapolated to humans with sometimes severe underlying cardiovascular pathology upon suffering sudden cardiac arrest. Although preliminary evidence suggests slight coronary vasodilatation after vasopressin administration, this model of extremely low cardiac output such as shock or CPR had the limitation to have a pre- and afterload that may be different from normal cardiocirculatory circumstances due to an open-chest preparation [72]. Lastly, There is no doubt that a CPR attempt employing vasopressin is more expensive than epinephrine (cost of 40 U vasopressin vs. 1 mg epinephrine, ∼15 vs. ∼1 US Dollar/Euro); however, this is not as expensive as a day on the intensive care unit on powerful antibiotics. In summary, laboratory studies employing vasopressin administration may have to detect the best time point, dosage, and exact indication for shock states — a lot of time and resource-consuming work is down the road, but as an ‘orphan’ drug, vasopressin has come a long way already.

6 Vasopressin as a continuous infusion during vasodilatory shock

The standard treatment of vasodilatory septic shock includes antibiotics, extracellular volume expansion, vasopressors, and drugs that increase myocardial contractility. Adrenergic catecholamines employed in this setting such as norepinephrine often have a diminished vasopressor action in vasodilatory shock; therefore, alternatives may be very useful. First reports with vasopressin in septic shock described a patient who was successfully resuscitated from cardiac arrest, but subsequently developed bacterial septicaemia, and needed a continuous infusion of both epinephrine and norepinephrine to maintain systolic arterial pressure of ∼90 mmHg (Fig. 10). Continuous infusion of vasopressin (0.04 U/min) then easily improved mean arterial blood pressure, cardiac output remained stable, and urine output increased from ∼6 to >50 ml/h. Stepwise withdrawal of vasopressin subsequently resulted in a decrease in arterial pressure, which again required norepinephrine to maintain systolic arterial pressure at 90–100 mmHg; urine output decreased to 30 ml/h. Six hours later, systolic arterial pressure was maintained at ∼105 mmHg on vasopressin alone [73]. In patients with vasodilatory shock after left ventricular assist device placement or postbypass vasodilatory shock, vasopressin (0.1 U/min) increased mean arterial pressure, while norepinephrine was decreased [74,75]. In the latter study, inappropriately low serum vasopressin concentrations were found before vasopressin administration, indicating endogenous vasopressin deficiency. Thus, it is possible that endogenous vasopressin stores in these patients were simply depleted, or not able to meet the demand to maintain cardiocirculatory homeostasis. Another investigation studying patients on amiodarone and angiotensin-converting enzyme inhibitors with refractory vasodilatation after cardiopulmonary bypass revealed a beneficial potential for continuous vasopressin infusion [76]. Similar observations were made in patients with milrinone-induced hypotension, when 0.03–0.07 U/min vasopressin caused an increase in systolic artery pressure of from ∼90 to ∼130 mmHg [77].

Fig. 10

Arterial blood pressure, urine output, and vasopressor management of a patient who was successfully resuscitated, but developed septic vasodilatory shock that was refractory to catecholamines. Note that an infusion of vasopressin increased arterial blood pressure, while both epinephrine and norepinephrine could be weaned; when vasopressin was discontinued, epinephrine and norepinephrine were again necessary to maintain arterial blood pressure. Reprinted with permission from Landry et al., Crit Care Med 1997;25:1279–1282. Copyright 1997 Lippincott Williams & Wilkins, Baltimore.

These observations may be similar to first experiences in cardiac arrest patients, when high endogenous vasopressin levels correlated with subsequent survival [27]. As such, it is likely that vasopressin in addition to the well established endogenous stress hormone epinephrine plays an important role in regards of arterial blood pressure regulation during shock states such as septicaemia, and CPR. Currently, we do not exactly know whether vasopressin acts simply as a back-up of the back-up vasopressor epinephrine during life-threatening shock states, whether vasopressin or epinephrine alone is better in certain situations, or whether these two hormones have unique adjunct features that we are only starting to understand. A better knowledge of these underlying mechanisms most likely would save many patients; patients who at this point in time have a relatively small chance of survival.

7 Practice points: management of cardiac arrest and vasodilatory shock patients

In adult patients with shock-refractory ventricular fibrillation, 40 U vasopressin, or 1 mg epinephrine should be administered intravenously; vasopressin is currently a single dosage, epinephrine can be repeated every 2–5 min. No recommendation for vasopressin was made to date for adult patients with asystole and pulseless electrical activity, and paediatrics due to lack of clinical data. Patients with vasodilatory shock refractory to adrenergic vasopressors should receive continuous infusions of vasopressin early (∼0.04 to ∼0.1 U/min) in order to stabilise cardiocirculatory function.


Supported, in part, by the Austrian Science Foundation grant P14169-MED, and the Austrian National Bank, both of Vienna, Austria; a Founder's Grant of the Society of Critical Care Medicine, Anaheim, California; and the Laerdal Foundation for Acute Medicine, Stavanger, Norway.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
View Abstract