The use of vasopressin for treating vasodilatory shock and cardiopulmonary arrest

Abstract

Objective – To discuss 3 potential mechanisms for loss of peripheral vasomotor tone during vasodilatory shock; review vasopressin physiology; review the available animal experimental and human clinical studies of vasopressin in vasodilatory shock and cardiopulmonary arrest; and make recommendations based on review of the data for the use of vasopressin in vasodilatory shock and cardiopulmonary arrest.

Data Sources – Human clinical studies, veterinary experimental studies, forum proceedings, book chapters, and American Heart Association guidelines.

Human and Veterinary Data Synthesis – Septic shock is the most common form of vasodilatory shock. The exogenous administration of vasopressin in animal models of fluid-resuscitated septic and hemorrhagic shock significantly increases mean arterial pressure and improves survival. The effect of vasopressin on return to spontaneous circulation, initial cardiac rhythm, and survival compared with epinephrine is mixed. Improved survival in human patients with ventricular fibrillation, pulseless ventricular tachycardia, and nonspecific cardiopulmonary arrest has been observed in 4 small studies of vasopressin versus epinephrine. Three large studies, though, did not find a significant difference between vasopressin and epinephrine in patients with cardiopulmonary arrest regardless of initial cardiac rhythm. No veterinary clinical trials have been performed using vasopressin in cardiopulmonary arrest.

Conclusion – Vasopressin (0.01–0.04 U/min, IV) should be considered in small animal veterinary patients with vasodilatory shock that is unresponsive to fluid resuscitation and catecholamine (dobutamine, dopamine, and norepinephrine) administration. Vasopressin (0.2–0.8 U/kg, IV once) administration during cardiopulmonary resuscitation in small animal veterinary patients with pulseless electrical activity or ventricular asystole may be beneficial for myocardial and cerebral blood flow.

Introduction

Systemic hypotension is a medical condition in which low arterial pressure causes inadequate tissue perfusion. Irreversible systemic hypotension can result in end-organ failure and subsequent death.¹ The most common cause of severe hypotension is systemic vasodilation, often referred to as vasodilatory shock.¹ Sepsis is the most frequent cause of vasodilatory shock although other conditions such as hemorrhagic shock, cardiogenic shock, carbon monoxide intoxication, and anaphylaxis can cause vasodilatory shock.² In cats, hypothermia may be a significant cause of vasodilatory shock. The underlying mechanism for the loss of peripheral vasomotor tone is not fully understood. The present evidence suggests that at least 3 pathways are involved: activation of ATP-sensitive potassium channels (K_ATP), increased synthesis of nitric oxide, and arginine vasopressin (AVP) deficiency (Figure 1).²

Figure 1.  The underlying mechanisms of vasodilatory shock. Sepsis, hemorrhagic shock, cardiogenic shock, or any state of prolonged shock can cause tissue hypoxia and lactic acidosis that in turn increase nitric oxide synthesis, activate ATP sensitive potassium channels (K_ATP) and lead to depletion of vasopressin. Decreased vascular smooth muscle ATP, increased hydrogen ion (H⁺) and lactate concentration can activate K_ATP. Activation of these channels can result in lower cytoplasmic calcium concentration and subsequent vasodilation. Cyclic guanosine monophosphate, cGMP. (Adapted from Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. N Engl J Med 2001; 345(8):558–595. Reprinted with permission.)

Vascular smooth muscle cells contain receptors for potent vasoconstricting substances (angiotensin II, norepinephrine, and endothelin). These substances bind to and activate vascular smooth muscle cells by increasing the cytosolic concentration of calcium via a second messenger (ie, inositol triphosphate [IP3] and diacylglycerol) (Figure 2).² The maintenance of membrane potential of vascular smooth muscle cells is integral to vasomotor tone. Hyperpolarization can close plasma membrane voltage-dependent calcium channels, resulting in vascular relaxation.³ K_ATP are located within the plasma membrane of vascular smooth muscle cells and are sensitive to decreases in cellular ATP.¹ Activation of these channels results in an efflux of intracellular potassium and subsequent plasma membrane hyperpolarization.² The opening of K_ATP can also be caused by an increase in cellular hydrogen ion and lactate concentration (Figure 3).¹

Figure 2.  The effect of hormonal or neuronal ligands upon a vascular smooth muscle cell. Angiotensin II and norepinephrine cause vasoconstriction via second messenger inositol triphosphate (IP3) and diacylglycerol (DAG). These messengers increase the cytosolic calcium concentration (Ca²⁺) via membrane calcium channels and intracellular stores. Increased cytoplasmic calcium activates a kinase that phophorylates myosin and results in vascular smooth muscle contraction. Atrial natriuretic peptide and nitric oxide activate a kinase that activates myosin phosphatase via the formation of cyclic guanosine monophosphate (cGMP) with subsequent dephosphorylation of myosin and vascular smooth muscle relaxation. (Adapted from Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. N Engl J Med 2001; 345(8):558-595. Reprinted with permission.)

Figure 3.  Resting potential and hyperpolarization of the vascular smooth muscle cell. Norepinephrine and angiotensin II cause vasoconstriction by opening voltage-gated calcium channels. Under normal resting conditions the ATP sensitive potassium channels (K_ATP) remain closed with no loss of cellular potassium (K⁺). Decreased cellular ATP, increased hydrogen ion (H⁺) and lactate concentration can activate K_ATP. Activation of these channels causes cellular potassium loss, membrane hyperpolarization, and closure of the voltage-gated calcium channels. Decreased cellular calcium prevents response to norepinephrine or angiotensin II. (Adapted from Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. N Engl J Med 2001; 345(8):558–595. Reprinted with permission.)

Amplification of nitric oxide (NO) by the increased expression of inducible nitric oxide synthase is an important cause of arterial hypotension.^2,4,5 Inflammatory cytokines, bacterial endotoxins, or both can enhance NO production by the vascular endothelium and vascular smooth muscle.⁵ Increased synthesis of NO has been observed in septic shock and decompensated hemorrhagic shock.² NO and its metabolite peroxynitrite (ONOO⁻¹) have also been implicated in down regulation of α₁-adrenoreceptors and norepinephrine.⁵ The vasodilating effect of NO is mediated by the activation of calcium-sensitive potassium channels (K_Ca) within the plasma membrane of vascular smooth muscle cells.² The mechanism of activation is direct nitrosylation of the channel, activation of cyclic guanosine monophosphate (cGMP), or both.⁶ Similarly, K_ATP can be activated by NO. Regardless of the mechanism, activation of these channels results in potassium efflux from the cells and subsequent plasma membrane hyperpolarization.²

AVP deficiency has also been documented in hemorrhagic shock.⁷ Vasopressin, also known as antidiuretic hormone, and oxytocin are nonapeptides composed of a 6-membered disulfide ring and 3-membered tail on which the terminal carboxyl group is amidated.⁸ The removal of a terminal amino group from cysteine will yield 1-desamino-8d arginine vasopressin, commercially available as desmopressin acetate for antidiuretic activity.⁸

AVP is synthesized as a large prohormone in the magnocellular neurons within the supraoptic and paraventricular nuclei of the hypothalamus.^9,10 The rate of release of AVP increases as the frequency of action potentials upon the magnocellular neurons of the hypothalamus increases.^9,11,12 Ten to 20% of the total hormonal pool within the posterior lobe of the pituitary can be released at one time.¹³ AVP is rapidly metabolized in the liver and kidney.^11,14

The regulation of AVP release can be divided into osmotic and nonosmotic stimuli.^9,13,15 Osmotic control of AVP production and release is mediated by peripheral and central osmoreceptors (Figure 4).^9,13,15,16 Peripheral osmoreceptors are located in the mesenteric and portal vasculature, whereas central receptors are located in the anteroventral region of the third ventricle.^9,11,17,18 Afferent nerve impulses from peripheral osmoreceptors travel via the vagus nerve to the nucleus tractus solitarus and ventrolateral medulla onto the paraventricular and supraoptic nuclei.^9,13 Severe hypovolemia and hypotension are the most potent nonosmotic stimuli for AVP release.^9,13,15 In contrast, increases in blood pressure and intravascular volume decrease AVP release.¹⁹ Other stimuli such as nausea, hypoglycemia, hypoxemia, hypercapnia, and intracranial hypertension can also cause AVP release.^15,16,20 Acetylcholine, angiotensin II, histamine, dopamine, and catacholamines stimulate AVP release.²¹ Inhibitors of AVP release include opioids, γ-aminobutyric acid, and atrial natriuretic peptide.²² Neurohumoral inhibition of AVP release is mediated by NO via cGMP.²²

Figure 4.  Peripheral osmoreceptors, central osmoreceptors, volume, and pressure receptors regulate vasopressin release. Peripheral osmoreceptors are located in the mesenteric (MV) and portal vasculature (PV). Central osmoreceptors, subfornical organ (SFO), and organum vasculosum lamina terminalis (OVLT) are located outside the blood brain barrier. Volume receptors are located in the atria (A) and ventricles (V). Pressure receptors are located in the aortic arch (AA) and carotid sinus (CS). Signals from the MV, PV, A, AA, and CS are sent via the vagus nerve to the nucleus tractus solitarus (NTS), area postrema (AP) and ventrolateral medulla (VLM). These cells then project to the supraoptic nuclei (SON) and paraventricular nuclei (PVN). The SFO and OVLT detect changes in systemic osmolality. Signals are sent to the SON and PVN. The median preoptic nucleus (MPN) has mutual connections with SFO, OVLT, SON, and PVN. The final common pathway for vasopressin release is synthesis in the magnocellular neurons (SON, PVN) and subsequent transport to the posterior pituitary (PP). (Adapted from Holmes CL, Patel BM, Russell JA et al. Physiology of vasopressin relevant to management of septic shock. Chest 2001; 120:989–1002. Reprinted with permission.)

Inappropriate release of AVP can result in hyponatremia and impaired water excretion.¹⁵ In canine patients, this condition-syndrome of inappropriate secretion of antidiuretic hormone (SIADH) has been reported to be caused by heartworm disease and neoplasia.^8,23 In human patients, malignant neoplasia, pulmonary diseases, CNS diseases, endocrine diseases, and certain drugs cause SIADH.^8,15,23

AVP and its analog terlipressin exert their effects through 5 different G-protein-coupled receptors: V₁, V₂, V₃, the oxytocin receptor, and the purinergic P₂ receptor.^24,25 Terlipressin has a higher vascular affinity V1:V2 ratio than vasopressin (2.2 versus 1, respectively). The pressor effect of vasopressin is mediated by stimulation of V₁ receptors located in high density on vascular smooth muscle.²⁵ V₁ receptors are also present in the brain, testis, liver, and renal medulla.²⁵ Activation of the V₁ receptor involves G proteins (G_q/11family) and subsequent production of phospholipase C and intracellular messengers IP3 and diacylglycerol.^25–27 The intracellular messengers activate protein kinase C and elevate intracellular calcium.^25–27 The increased intracellular release of calcium from the sarcoplasm and increased extracellular calcium influx result in muscle contraction and peripheral vasoconstriction (Figure 5).^25,26,28 Activation of the V₂ receptor, located in the renal collecting duct, increases intracellular cAMP. This triggers increased water reabsorption via the fusion of aquaporin-2-bearing vesicles with the apical plasma membrane of the renal collecting duct cells.^25,26,28 Activation of V₃ receptor, located in the anterior pituitary cells, causes secretion of ACTH.²⁵ Stimulation of the oxytocin receptor may cause smooth muscle contraction of the myometrial and mammary myoepithelial cells or vasodilation of certain vascular beds.²⁵ Purinoreceptors have been located on the cardiac endothelium.²⁵ Activation of this receptor in isolated perfused guinea pig hearts resulted in coronary vasoconstriction and negative inotropy, but in isolated perfused rat hearts positive inotropy was observed.²⁵

Figure 5.  The V₁ receptor of a vascular smooth muscle cell is coupled with G-proteins (G_q/11). Vasopressin acts upon the V₁ receptor causing the formation of guanosine diphosphate (GDP) from guanosine triphosphate (GTP). Subsequent activation of phospholipase C (PL-C) causes the formation of inositol triphosphate (IP3) and diacylglycerol (DAG). Activation of protein kinase C (PK-C) stimulates the release of calcium from the sarcoplasmic reticulum (SR). Increased intracellular calcium activates a kinase capable of phosphorylating myosin that in turn causes vasoconstriction. (Adapted from Holmes CL, Landry DW, Granton JT. Science review: vasopressin and the cardiovascular system part 1-receptor physiology. Critical Care 2003; 7:427–434. Reprinted with permission.)

Hemodynamic Effects of AVP or Terlipressin

Animal studies

The effects of V₁ agonists (AVP and terlipressin) on mean arterial pressure (MAP), cardiac output (CO), and systemic vascular resistance are similar although differences in MAP and CO were observed in hypodynamic versus hyperdynamic endotoxic shock models. In a study of fluid-challenged endotoxic rats terlipressin (6 μg/kg, IV) markedly increased MAP.²⁹ The same study observed a nonsignificant change in MAP without fluid challenge when given the same dose of terlipressin. Significant increases in MAP via V₁ systemic receptor stimulation have been observed in porcine^30–33 and ovine^4,34–36 endotoxic shock models. The use of AVP in 2 studies of canine hemorrhagic shock also resulted in a marked increase in MAP.^7,37 A recent case series of dopamine resistant hypotension and vasodilatory shock in dogs showed an increase in MAP after AVP (0.5–1.25 mU/kg/min) constant rate infusion.³⁸ Decreased CO was observed in studies of porcine endotoxemia,^30,32 ovine endotoxemia,^4,34,35 canine hemorrhagic shock,^7,35 and was unchanged in another study of porcine septic shock.³¹ The authors of this study speculated that the absence of change in CO was a result of fluid resuscitation and thus a hyperdynamic endotoxic state. The same study, though, demonstrated that a high-dose infusion of AVP dramatically decreased renal blood flow. It is believed that AVP at low doses (0.01–0.04 U/min) does not cause preferential vasoconstriction to any organ but at high does becomes markedly selective.²¹ In studies that measured vascular resistance a significant increase was noted with the use of V₁ agonists.^4,30,37,39 In addition to improved hemodynamic parameters, improved survival was observed with the use of low-dose V₁ agonists in endotoxic rats.²⁹

Human studies

In all septic shock studies, AVP or terlipressin increased MAP.^18,39–49 In a prospective open-label study of 17 catecholamine resistant septic shock patients 1 or 2 boluses of terlipressin (1 mg, IV) moderately increased MAP.⁴⁹ In a prospective, randomized, controlled study of 48 cardiovascular surgical patients with advanced vasodilatory shock a continuous infusion of AVP (0.066 U/min) was compared with continuous infusion of norepinephrine (0.51–2.25 μg/kg/min). Results showed mild to moderate increases in MAP with AVP.⁴⁸

Liver function: V₁ agonists significantly increased total bilirubin, alanine transferase, and aspartate aminotransferase in 2 studies.^48,49 The change in hepatic enzymes and liver function may be due to an effect upon hepatic blood flow or a direct effect on hepatocellular function.^1,48,50

Platelet count: In all studies of AVP in advanced vasodilatory shock platelet counts decreased significantly.^48,51,52 Thrombocytopenia, though, was not associated with increased mortality.^51,52 Thromboelastography was performed in one study revealing no difference in clot stability or platelet function between AVP or norepinephrine group.⁵² A possible explanation for decreased platelet counts after AVP infusion may be activation of V₁ receptors on platelet membranes. Stimulation of these receptors activates the phosphatidyl-inositol-cascade leading to an increase in cytoplasmic calcium with subsequent platelet aggregation.^51,53

Urine production: Urine production was increased in many studies.^43,44,46,49 This effect may be due to natriuretic effect of AVP at high serum concentrations (≥100 pg/mL),⁴⁶ or glomerular efferent arteriole vasoconstriction,³⁶ or both.

Splanchnic Circulation Effects of AVP or Terlipressin

Animal studies

The effects of V₁ agonists on splanchnic circulation are varied depending upon the species and model investigated. In a study of mongrel anesthetized dogs incremental doses of AVP caused significant reductions in mesenteric, renal, and iliac blood flow.⁵⁴ Likewise, coronary vasoconstriction but not pulmonary vasoconstriction was observed in response to AVP.⁵⁵ These studies support different vascular beds have a unique sensitivity to AVP. The selective effect of AVP, therefore, causes redistribution of blood flow. Possible explanations for these differences include receptor type, location density, and mechanism of action. The effects of AVP on various models of gastric mucosal damage have been studied in rats.⁵⁶ It was found that an AVP receptor antagonist reduced gastric lesions in a dose-dependent manner.⁵⁶ In another study of acute endotoxemia in rats an AVP receptor antagonist improved intestinal mucosal integrity.⁵⁷ Macroscopic mucosal injury and increased plasma leakage from the ileum and colon were observed without the AVP receptor antagonist.⁵⁷ These studies suggest AVP is involved in the pathogenesis of gastric and intestinal lesions in hypotensive models without fluid administration. Decreased splanchnic blood flow was found in a study of endotoxic non–fluid-resuscitated rats after infusion of terlipressin.²⁹ This finding is in contrast to endotoxic fluid-resuscitated rats that were administered terlipressin.²⁹ Increased mesenteric blood flow as well as increased ileal microcirculation were observed in this hyperdynamic state.⁴⁵ In a study of porcine endotoxemia, vasopressin caused significant visceral dysoxia when used in the hypodynamic state.²⁹ Decreased mesenteric and portal vein blood flow were measured along with increased jejunal lactate lumen release.³² In contrast, in fluid-resuscitated endotoxic pigs that were administered AVP (approx 0.05 U/min) no detrimental effects on splanchnic circulation were seen.³¹ In animal models fluid resuscitation before the administration of V₁ agonists in systemic hypotensive states appears important to splanchnic circulation and intestinal integrity.

Cardiopulmonary Resuscitation and AVP

Animal studies

The finding of increased endogenous AVP concentrations in successfully resuscitated humans following in-hospital cardiopulmonary arrest stimulated many animal studies.⁵⁸ The porcine model of ventricular fibrillation and subsequent cardiopulmonary resuscitation (CPR) is commonly used to study human cardiopulmonary arrest.

Coronary perfusion pressure: In 2 studies using a porcine cardiopulmonary arrest and resuscitation model and performed by the same investigator significant increases in coronary perfusion pressure (CoPP) and myocardial blood flow were observed.^59,60 In the first experiment, AVP (0.8 U/kg, IV) was administered after cardiopulmonary arrest. Moderate and mild significant increases in CoPP were measured at 90 seconds and 5 minutes, respectively.⁵⁹ Moderate and marked significant increases in myocardial blood flow were also observed at 90 seconds and 5 minutes, respectively.⁶⁰ Significant increases in CoPP have also been observed after endobronchial or intraosseous administration of AVP.^61,62 Repeated administration of AVP was found to maintain CoPP threshold (20–30 mm Hg) for successful cardiac defibrillation or return to spontaneous circulation (ROSC) while epinephrine was not.⁶³ A possible explanation for this finding is that repeated administration of epinephrine increases myocardial oxygen consumption, which may result in a severe mismatch of cardiac oxygen delivery versus oxygen consumption during CPR.⁶³

Cardiopulmonary arrest: In a clinically relevant porcine model of pulseless electrical activity (PEA), AVP (0.8 U/kg), epinephrine (0.02 mg/kg), and a combination of AVP (0.4 U/kg) and epinephrine (0.01 mg/kg) were compared.⁶⁴ CoPP that predict ROSC were exceeded in the AVP group and the AVP and epinephrine group.⁶⁴ In addition to myocardial oxygen consumption and demand mismatch, another reason for the findings in the epinephrine group may be predominant beta vasodilator effects in this model.⁶⁴

Cerebral effects: In a porcine model of cardiopulmonary arrest significantly increased cerebral blood flow was observed in AVP-treated animals compared with epinephrine.⁶⁵ In yet another porcine model the effects of AVP (0.4 U/kg), epinephrine (0.2 mg/kg), and combined AVP (0.4 U/kg), epinephrine (0.045 mg/kg), and norepinephrine (0.045 mg/kg) on cerebral blood flow were compared. Significant increases in cerebral blood flow were observed in all AVP groups compared with epinephrine.⁶⁶ Two studies have assessed the effect of AVP and epinephrine upon cerebral pathology and survival postCPR.^67,68 In 1 study of rats, no difference in neurological deficit score (physical exam) or histological quantitative analysis of viable and apoptotic neurons was observed between AVP or epinephrine group.⁶⁸ In another study of pigs, brain magnetic resonance imaging (MRI) was performed 96 hours after prolonged CPR with either AVP or epinephrine.⁶⁷ No brain MRI abnormalities were observed in AVP group.⁶⁷ Interestingly, no pigs in the epinephrine group survived to 96 hours.⁶⁷

Canine asystole: One successful case of canine CPR with AVP administration has been reported.⁶⁹ Asystole was observed approximately 1 hour into surgery. CPR was performed and 2 doses of AVP (0.8 U/kg, IV) were administered. ROSC with transient second-degree heart block was observed. No cardiac or neurologic dysfunction was noted during the recovery period. The patient was discharge 3 days later.

Human studies

AVP administration in refractory cardiopulmonary arrest was first reported in 1996 by Lindner et al.⁷⁰ Standard American Heart Association advanced cardiac life support therapy was performed on 8 patients with in-hospital cardiopulmonary arrest. Each patient received at least 1 mg of epinephrine (range 1–13 mg) during standard CPR with no response. After at least 12 minutes of resuscitation AVP (40 U) was administered with subsequent ROSC. Three of 8 patients did survive to discharge.⁷⁰ Subsequent medical investigations have focused on the importance of CoPP and ROSC from cardiopulmonary arrest.

CoPP: In a prospective open-label trial of AVP versus epinephrine CoPP was measured in 10 patients with cardiopulmonary arrest.⁷¹ When resuscitative measures according to advanced cardiac life support failed, patients were administered epinephrine followed 5 minutes later by AVP. AVP administration resulted in a significant increase in CoPP in 4 of 10 patients.⁷¹

Cardiopulmonary arrest: A prospective, randomized, double-blinded study of epinephrine versus AVP in 40 patients with out-of-hospital ventricular fibrillation was performed in 1997.⁷² A significant increase in survival at 24 hours was observed in the AVP group.⁷² In 1999, Li et al.⁷³ studied 2 different dosages of epinephrine and AVP in 83 patients with in-hospital cardiopulmonary arrest. ROSC and survival was significantly increased in the high dose AVP (1 U/kg) group compared with epinephrine (1 and 5 mg) groups.⁷³

Four studies, because of study design and patient numbers, provide the most accurate information regarding the effect of AVP on cardiopulmonary arrest. A randomized, placebo-controlled study of 200 patients with in-house cardiopulmonary arrest was used to compare AVP and epinephrine.⁷⁴ In that study, patients were assigned to either receive AVP (40 U) or epinephrine (1 mg). Patients who did not respond to the first injection were then given epinephrine as a rescue medication. No significant differences in survival to hospital discharge, 1-hour survival, or mental state examination scores were identified.⁷⁴ Possible explanation for the lack of significant differences could be a greater comorbidity of various chronic illness⁷⁴ within this patient population or that epinephrine and AVP are equipotent when rescue times are short.⁷⁵ In a prospective, double blinded, controlled, multicenter European study, 1,186 patients were randomized to receive either 2 injections of AVP (40 U) or epinephrine (1 mg) for out-of-hospital cardiopulmonary arrest.⁷⁶ No significant differences to hospital admission were observed among patients with ventricular fibrillation or PEA.⁷⁶ However, a significant increase in hospital admission and hospital discharge was observed in patients with asystole that were administered AVP. Interestingly, for those 732 patients that did not have ROSC after 2 injections of either AVP or epinephrine, repeated doses of epinephrine in the AVP group but not the epinephrine group resulted in significant improvements in hospital admission and hospital discharge.⁷⁶ This finding was different from a previous study.⁷⁴ In this 2001 study, patients with in-hospital cardiac arrest were compared. No difference in survival, mental state examination score or cerebral performance score was observed between the AVP or epinephrine group. In 2005, Aung and Htay⁷⁷ published a meta-analysis of 1,519 patients with cardiopulmonary arrest comparing AVP and epinephrine. Five randomized controlled trials were identified. No significant advantage of AVP over epinephrine was identified with regard to in-hospital arrest, out-of-hospital arrest, or any subgroup analysis (initial cardiac rhythm).⁷⁷ Combined effect of AVP and epinephrine, though, could not be analyzed in this study. The effect of AVP and epinephrine during out-of-hospital cardiopulmonary arrest was evaluated by Callaway et al.⁷⁸ Three hundred and twenty-five adult subjects with out-of-hospital arrest were randomly assigned to receive either AVP (40 U) or placebo after 1 injection of epinephrine during CPR. No significant differences in rates of ROSC, presence of pulses at hospital arrival, or initial cardiac rhythm were observed.⁷⁸ The effect of AVP or epinephrine upon initial cardiac rhythm (ventricular fibrillation or pulseless ventricular tachycardia) and subsequent survival was specifically studied in 109 patients by Grmec and Mally.⁷⁹ The design of the study was a prospective observational cohort with a retrospective control group. Significant improvement in survival was observed in the vasopressin groups versus the epinephrine group.⁷⁹ Results of the study should be interpreted cautiously because of the number of patients and design of the study.

Colonic ischemia: The use of vasopressin may have side effects in prolonged CPR. A case report of a woman with sudden cardiopulmonary arrest secondary to ventricular fibrillation was resuscitated with vasopressin.⁸⁰ The postresuscitation period was complicated by massive colonic ischemia and necrosis necessitating subtotal colectomy.⁸⁰ The patient did survive the complication.

Summary

The mechanism of action for AVP in advanced vasodilatory shock may be its ability to block K_ATP in vascular smooth muscle and interfere with NO signaling. Improved survival and lower frequency of new onset tachyarrhythmias are favorable characteristics of V₁ agonists in conjunction with low-dose catecholamines in septic shock.^{18,39,41,44–49} The effects of V₁ agonists on splanchnic circulation, though, are mixed depending upon whether fluid resuscitation has been administered or not. Improved splanchnic blood flow with fluid resuscitation has been observed in endotoxemic rat models.²⁹ If, on the other hand, fluids are not administered, decreased mesenteric and portal blood flow with significant visceral dysoxia have been found in porcine endotoxemic models.³² The side effects of V₁ agonists cannot be overlooked and include mild to moderate increases in hepatic enzymes,⁴⁹ total bilirubin,^48,49 urine production,^36,46 decreased CO,^{30,32,40,43,49} heart rate,⁴⁸ platelet count,⁴⁹ and skin necrosis.⁸¹ The use of V₁ agonists in systemic hypotension unresponsive to fluid resuscitation and catecholamine administration should be considered.

CPR with V₁ agonists, specifically AVP, has been studied predominately in porcine models and human clinical trials. Intravenous, intraosseous, or endobronchial administration of AVP can increase CoPP and myocardial blood flow during CPR.^59–62,71 Repeated administration of AVP but not epinephrine has been found to maintain a CoPP threshold (20–30 mm Hg) for successful cardiac defibrillation.⁶³ Two experimental porcine studies have shown increased cerebral blood flow when using AVP compared with epinephrine in CPR.^65,66 Yet no difference in neurologic exam or histological analysis of brain tissue was found.⁶⁸ Four studies have shown an increased ROSC and survival in human patients with ventricular fibrillation, pulseless ventricular tachycardia, and nonspecific cardiopulmonary arrest when using AVP versus epinephrine.^70,72,73,79 Three large studies, though, have not identified a significant difference between AVP and epinephrine in survival to hospital admission in out-of-hospital cardiopulmonary arrest or survival to hospital discharge after in-hospital cardiopulmonary arrest.^74,76,77 In the study by Wenzel et al.,⁷⁶ a significant increase in survival to hospital admission and discharge was observed in the subgroup of patients with asystole that were treated with AVP. In contrast, no increase in survival was observed in patients with ventricular fibrillation.⁷⁶ Subgroup examination in the larger meta-analysis by Aung and Htay⁷⁷ did not find any difference in initial cardiac rhythm and the success of using AVP or epinephrine in CPR. The advantage of AVP over epinephrine or the combination of AVP and epinephrine during CPR has yet to be clearly identified.

Recommendations

The 2004 Surviving Sepsis Campaign Guidelines for the Management of Septic Shock and the 2005 American Heart Association (AHA) Guidelines for Cardiopulmonary Resuscitation include recommendations for initial resuscitation and the use of AVP.^82,83 The recommendations (Tables 1 and 2) are supported by a grading and rating classification, respectively. Early goal-directed therapy for human patients presenting with septic shock have shown improved survival.⁸⁴ In addition to many targets for fluid resuscitation and cardiovascular performance goals outlined in the Surviving Sepsis Guidelines, the use of vasopressors is recommended if adequate fluid administration fails to restore blood pressure and organ perfusion. Either norepinephrine (0.05–1 μg/kg/min) or dopamine (2.5–10 μg/kg/min) are likely first choices in humans. The evidence, though, for either drug to be a first-choice vasopressor in humans is Grade D.⁸² Norepinephrine is more potent than dopamine and may be more effective in septic shock in humans.⁸⁵ If systemic hypotension continues to be unresponsive in humans, AVP (0.01–0.04 U/min) can be started. The evidence for this recommendation in humans is Grade E.⁸² Doses >0.04 U/min have been associated with myocardial ischemia and cardiopulmonary arrest in humans.⁸⁶

Table 1.   Grading system

Grading recommendations
A.	Supported by at least 2 level I investigations
B.	Supported by at least 1 level I investigation
C.	Supported by level II investigations only
D.	Supported by at least 1 level III investigation
E.	Supported by level IV or V evidence
Grading of evidence
I.	Large, randomized trials with clear-cut results; low risk of false-positive (alpha) error or false-negative (beta) error
II.	Small, randomized trials with uncertain results; moderate-to-high risk of false-positive (alpha) and/or false-negative (beta) error
III.	Nonrandomized, contemporaneous controls
IV.	Nonrandomized, historical controls, and expert opinion
V.	Case series, uncontrolled studies, and expert opinion

Table 2.   Classification of recommendations

Class I	Class IIa	Class IIb	Class III	Class indeterminate
Benefit>>>Risk	Benefit>>Risk	Benefit≥Risk	Benefit≥Risk
Adapted from ECC Committee. Subcommittees and Task Force of the American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2005; 112:IV–59. Reprinted with permission.
High-level prospective evidence	Weight of evidence supports the intervention	Lower level of evidence to support the intervention, or short-term benefits documented		Research is ongoing
Intervention should be performed	Intervention considered acceptable and useful	Intervention may be considered and may be useful	Intervention should not be performed and may be harmful	Cannot recommend for or against the intervention

The revised AHA Guidelines for Cardiopulmonary Resuscitation recommends the use of vasopressin for the treatment of ventricular fibrillation and pulseless ventricular tachycardia in humans in specific circumstances. If CPR and defibrillation do not cause ROSC, the guidelines suggest using epinephrine (1 mg, IV, every 3–5 min) (Class IIb evidence) or a single dose of AVP (40 U, IV) (Class Indeterminate evidence) to replace either the first or second dose of epinephrine.⁸³ PEA and asystole as presenting arrest rhythms in humans are not responsive to defibrillation but the recommendations for the use of drugs during CPR remain the same.^83,86

The application of vasopressin in small animal veterinary medicine should be considered in advanced vasodilatory states such as septic shock and as a component of cardiopulmonary-cerebral resuscitation (CPCR) in PEA and ventricular asystole. Septic shock has been defined as sepsis with refractory hypotension or an arterial systolic blood pressure <90 mm Hg despite aggressive fluid resuscitation.⁸⁷ Optimal management of the cardiovascular abnormalities present in vasodilatory shock requires assessment of CO and cardiac contractility. In clinical situations where such assessments are not available or not practical the following recommendations may be useful. Suggested parameters for optimal small animal fluid resuscitation are a central venous pressure of 5–10 mm Hg, systolic blood pressure of 100–120 mm Hg, urine production of 0.5–1.5 mL/kg/h, PCV ≥25%, lactate level <2.5 mmol/L, improved mentation, and normothermia.⁸⁸ Hypothermia may interfere with fluid resuscitation goals by down regulating catecholamine receptors. If aggressive fluid resuscitation results in an adequate intravascular volume as may be suggested by a normal central venous pressure accompanied by low systolic blood pressure, positive inotropes or vasopressors may be indicated (Figure 6). Dobutamine is a sympathomimetic β₁ adrenoreceptor agonist that increases CO via myocardial stimulation. It is delivered as a constant rate infusion (2.5–10 μg/kg/min). Side effects are unusual but include the development of ventricular arrhythmias. If this occurs the infusion can be temporarily discontinued and restarted at a lower infusion rate. If dobutamine administration is unsuccessful dopamine (2.5–10 μg/kg/min) or norepinephrine (0.01–1 μg/kg/min) can be added. Dopamine is a precursor of norepinephrine and exerts its effect in a dose-dependent manner. At low doses (2.5 μg/kg/min) dopamine stimulates canine dopaminergic receptors causing vasodilation in the renal and mesenteric vasculature. In contrast, low-dose dopamine infusion may not dilate the feline renal vasculature.⁸⁹ Higher doses (5–10 μg/kg/min) stimulate β₁ and α₁ adrenoreceptors resulting in increased myocardial contractility and vasoconstriction. Norepinephrine (0.01–1 μg/kg/min) causes vasoconstriction via potent α₁ adrenoreceptor effects. Patients with systemic hypotension unresponsive to sympathomimetic therapy (dobutamine, dopamine, and norepinephrine) and fluid resuscitation are candidates for AVP (0.01–0.04 U/min, IV Constant Rate Infusion [CRI]) administration. AVP stimulates vasoconstriction by stimulation of V₁ receptors in vascular smooth muscle. AVP diluted in normal saline⁹⁰ should be administered IV and titrated from a low starting dose (0.01 U/min, IV CRI). Doses in excess of 0.04 U/min should be avoided because of the risk of myocardial ischemia. AVP should not be used in patients with cardiogenic shock because increased systemic vascular resistance will worsen cardiac failure.

Figure 6.  AVP, arginine vasopressin; CRI, constant rate infusion.

AVP may be considered as a part of CPCR for PEA and ventricular asystole in the small animal veterinary patient (Figure 7). AVP improves CoPP, myocardial blood flow, and cerebral blood flow by peripheral vasoconstriction and shunting of blood toward the heart and CNS. Once the ‘ABCs’ (airway, breathing, circulation, or cardiac compressions) of CPCR have been established and PEA is identified, AVP (0.2–0.8 U/kg, IV) or epinephrine (0.01 mg/kg, IV) can be administered followed by CPCR for 3–5 minutes. The optimal dose of epinephrine in small animal patients is not known. The IV dose provided is currently the accepted starting dose.⁹¹ The decision whether to give AVP or epinephrine first is debatable. The 2005 AHA Guidelines for Cardiopulmonary Resuscitation states AVP can be used instead or with epinephrine in PEA because human studies have not shown a difference in survival outcome.⁸³ In the authors' opinion, other drug therapies that may be helpful in PEA are naloxone and atropine. Naloxone (0.02–0.04 mg/kg, IV) may be effective in augmenting CO by blocking endogenous endorphins or exogenous narcotics. Atropine (0.04 mg/kg), on the other hand, may be helpful in decreasing parasympathetic tone in PEA. If ventricular asystole is observed after the ‘ABCs’ of CPCR have been implemented AVP (0.2–0.8 U/kg, IV) or epinephrine (0.01 mg/kg, IV) can be administered followed by CPCR as above. The decision to give AVP or epinephrine first is again debatable. Some studies as well as expert opinion support using AVP first or without epinephrine.^61,76,91,92 The 2005 AHA Guidelines for Cardiopulmonary Resuscitation again allow for the use of AVP or epinephrine for the treatment of ventricular asystole.⁸³ Again, in the authors opinion atropine (0.4 mg/kg, IV) and naloxone (0.02–0.04 mg/kg, IV) may be useful in ventricular asystole for the above reasons.

Figure 7.  AVP, arginine vasopressin; ROSC, return of spontaneous circulation.

Beneficial effects of AVP have been observed in human clinical and animal experimental studies of septic shock, hemorrhagic shock, and CPR. Large, prospective, and randomized small animal clinical trials in advanced vasodilatory shock and CPCR evaluating important and clinically relevant outcomes are needed to elucidate whether AVP is superior to other currently accepted therapies for these conditions. Considering the available data, however, it may be reasonable to consider the use of AVP as a component of therapy in small animal veterinary patients with advanced vasodilatory shock or requiring CPCR.