This chapter is relevant to Section G8(i) of the 2017 CICM Primary Syllabus, which asks the exam candidate to "understand the detailed pharmacology of inotropes and vasopressors". It is a summary of the pharmacological properties of vasopressin, with a focus on its use as a vasoactive infusion. It has come up in past papers, most notably in Question 20 from the first paper of 2017 and Question 22 from the first paper of 2013.
Now, when one normally calls something a "summary", one implies that the information is sensibly condensed into several small easily digestible snack-like pieces. Carrying on with the food metaphor, what follows is more like eating out of the garbage. Occasional interesting morsels are buried under layers of irrelevant or duplicated material. To save the readers from his worst excesses, the author has attempted to combine all the exam-essential elements into the grey box which follows.
Class Vasopressor Chemistry Pituitary hormone Routes of administration IV Absorption Basically zero oral availability due to destruction by intestinal peptidases, such as trypsin Solubility pKa = 10.26, good water solubility Distribution VOD = 0.14 to 0.2 L/kg; protein binding ~ 30% Target receptor Vasopressin binds to V1 receptors (vasoconstrictor effect) and V2 receptors (antidiuretic effect). Metabolism 35% is metabolised by endothelial peptidases in the liver Elimination 65% is excreted unchanged by the kidney; half-life 17-35 minutes Time course of action Rapid onset of effect Mechanism of action Vasopressor effects are exerted by V1 receptors, which are Gq-protein coupled receptors. Similarly to alpha-1 receptors, they increase intracellular calcium by means of increasing cAMP concentrations.
V2 receptors are Gs-coupled receptors and produce the insertion of aquaporins into the apical membrane of principle cells of the collecting tubule.
Unlike catecholamine receptors, vasopressin receptors do not lose their affinity for vasopressin with changing pH.
Clinical effects Vasoconstriction, redistribution of splanchnic blood flow, increased platelet aggregation, decreased urine output, increased circulating Factor VIII and von Willebrand factor Single best reference for further information TGA PI document
Chemical structure and chemical relatives of vasopressin are a fascinating topic, and
a characteristically unhinged discussion of the vasopressin receptor agonist class is carried on in the chapter on the pharmacology of vasopressin an its analogues, exiled into the lower hells of the endocrinology section because that is where the relevant CICM syllabus item was. On the other hand, it was felt that this vasopressor-focused entry was necessary and should belong in the cardiovascular section, as that is more logical for the intensivist who will be using this substance mainly to manipulate the cardiovascular system. That same intensivist will, more than likely, also have little interest in the details of disulfide bridges and cyclical peptide molecule structures. For that reason, here the reader will be spared those details, and instead redirected to works like Glavaš et al (2022).
Vasopressin really should be given via a central line. Like noradrenaline, is was once routinely administered via peripheral veins; and like noradrenaline gangrenous complications plagued its use.
Unlike noradrenaline, it did not seem to extravasate spontaneously, so its effect on the vasa venorum may not be as pronounced (given that the gangrenous complications were typically the effect of IVC dislodgement).
Half life = 17 – 35 minutes
65% of Vasopressin is excreted unchanged in the kidneys
35% of Vasopressin is metabolized by endothelial peptidases in the kidney and liver.
The key feature which renders vasopressin vulnerable to this form of elimination is the 8th amino acid, arginine. A substitution for dextro-arginine results in greatly increased half life (eg. desmopressin).
Vasopressin receptors are G-protein-coupled receptors and produce their effect very rapidly through IP3-mediated intracellular calcium release (in the vase of the Gq-coupled V1 receptors) and cAMP-mediated exocytosis of aquaporin-lined vesicles (in the case of the Gs-coupled V2 receptors). The "vasopressin and friends" chapter has a more content-rich explanation of their function, and what follows is a brief point-form reminder:
V1 vasopressin receptors are ubiquitous; the densest distribution is seen in the vascular smooth muscle. The distribution among vascular beds is also ubiquitous – with notable exceptions. For instance, vasopressin receptors are expressed in the efferent, but not the afferent arterioles of the kidney; the result is a tendency of vasopressin to increase the rate of glomerular filtration. This explains why shocked critically ill patients experience an increase (rather than a decrease) of urine output when the vasopressin infusion is commenced. Additionally, vasopressin receptors are found in the brain (where we don't really know what they do), and the platelets (where it increases intracellular calcium and creates conditions favourable for platelet activation).
At low doses, vasopressin acts as an antidiuretic hormone. The normal levels of vasopressin secretion in response to hyperosmolar states range up to a certain maximal effect.
Even though vasopressin concentration continues to rise, at a serum concentration of 5pM (picomoles) the maximal urinary concentration is achieved, and no greater water retention is possible. After this point, defence of osmolality must be accomplished by increasing water intake, rather than preventing water loss.
However, vasopressin levels will continue to increase in response to decreased blood pressure.
V1 receptor effects dominate this dose range. Via V1 receptors, vasopressin is a pure vasopressor, causing vasoconstriction without any chronotropic or inotropic effect. The levels increase in an exponential fashion; in animal models of haemorrhagic and septic shock the levels can increase 200-fold. The resulting systemic vasoconstriction preserves the perfusion of vital organs while sacrificing the more expendable tissues.
However, in sustained severe shock, this effect loses its oomph.
In both haemorrhagic and septic shock, the circulating levels of vasopressin seem to decrease, which probably contributes to the severity of shock. This is thought to be the consequence of pituitary vasopressin depletion. Obviously, this can be addressed with supplemental vasopressin. And yes, studies of septic pigs and haemorrhaging dogs have suggested that the vasopressin deficiency in shock states is readily answered by exogenous vasopressin infusion.
Vasopressin is not anything like the catecholamines. In contrast to them, vasopressin has variable effects depending on the expression of different receptors in different vascular beds. Not only that, but the effect can vary depending on dose.
And to top it off, in low (non-pressor) doses vasopressin has been show to actually act as a vasodilator in pulmonary arteries (of the rat), renal arteries (also of the rat), cerebral arteries (of the dog) and coronary arteries (of the monkeys), which seems to be related to L-arginine being synthesized into nitric oxide under the influence of V1 receptors.
Unlike noradrenaline, which causes a predictable vasopressor response in healthy people, vasopressin in the healthy human is a relatively weak pressor agent. The reason behind this is the increased sensitivity of the baroreceptor reflex mediated by V1 receptors in the brain. The administration of vasopressin causes more bradycardia than an equivalently vasoactive dose of noradrenaline – the heart rate decrease is greater.
The above graph is an extrapolation of the famous article by Allwood et al (1963), which investigated the effects of catecholamines on healthy humans. Vasopressin was not tested - the graph is my own confabulation. However, it is supported by the cardiovascular data presented by the VASST investigators, who found that the most significant hemodynamic difference between their noradrenaline and vasopressin group was the slower heart rate among the vasopressin-treated patients.
In the denervated heart, vasopressin has a positive inotropic effect, presumably by affecting V1 receptors in the myocardium, making more calcium ions available to support contraction. The
Vasopressin causes vasoconstriction of vessels in the skeletal muscle, fat, the pancreas, and the thyroid gland. The vasoconstricting effect on the mesentery, coronary and cerebral circulation is not as pronounced with vasopressin as it is with the catecholamines. However, the higher the dose, the more coronary vasoconstriction there will be.
Unlike the linear, predictable dose-response curve of noradrenaline, vasopressin has a more traditional S-shaped dose-response curve, which demonstrates the diminishing effect it has at high doses. It has a considerably (order of magnitude) greater potency than noradrenaline, and escalating doses yield a disproportionately greater response at the middle dose range.
So, Pang and Tabrizchi specified that their dose range for noradrenaline was from 3.0 x 10-10 to 8.0 x 10-9 mol/kg/min, and the dose range for vasopressin was 4.5 x 10-11 to 1.4x10-9 mol/kg/min. One can recognise that the molar mass of vasopressin (C46H65N13O12S2) is 1056.22g. Thus, the tested dose range is 0.0475 to 1.48 mcg/kg/min. If we believe that 1mg of vasopressin equates to about 600 units, we can say that 1 unit weighs 1.667mcg. Thus, the dose range in the rat was from 0.03 to 0.89 units per kg per min.
That to me seems like a completely insane dose rate, unless my maths are horribly wrong. Compare with the recommended human dose: the infusion rate for vasopressin in the treatment of shock in adults is 0.6-2.4 units per hr, i.e. 0.01– 0.04 units/min, or 0.00014 to 0.0066 units/min/kg for a 70kg human.
However, we must recall that in healthy mammals with a preserved autonomic nervous system, vasopressin is a relatively weak vasoconstrictor. Perhaps the rats were less responsive for this reason. In contrast, septic humans seem to have a greatly increased sensitivity to vasopressin, and the doses they require are substantially lower.
It seems the consenting participation of healthy humans was more difficult to secure for vasopressin research than for noradrenaline; we must make do with data collected from patients in severe septic shock. True, this does not reflect the "real" physiological dose-response relationship - given their known hypersensitivity to vasopressin - but it is more applicable to bedside use in the ICU.
Don't pay attention to the extrapolated graph. The abovementioned study did not find a dose-response curve for vasopressin. The degree of improvement in MAP was not related to dose. They did, however, establish that at doses over 0.03 units per minute (1.8 units/hr) the cardiac index decreased significantly.
Furthermore, they mentioned that beyond 0.04 units/min there was no further improvement in hemodynamic parameters, suggesting that this is the point where the dose-response curve begins to plateau. This may have given rise to the manufacturer's dosing recommendation.
In addition to its own direct effects, vasopressin seems to enhance the catecholamine sensitivity of mammalian blood vessels (in one specific instance, the mesenteric artery of the rat).
This means that the commencement of a vasopressin infusion together with noradrenaline will lead to a greater improvement in blood pressure than would be expected from a simple additive effect.
As previously mentioned, vasopressin can act as a vasodilator in some select groups of vessels, and certainly some monkey coronaries had dilated for some investigators at some point. However in a series of anaesthetised dogs vasopressin consistently decreased coronary blood flow and increased coronary vascular resistance.
What are we to make of this? The pragmatic intensivist would have two questions. In the patient with untreated ischaemic heart disease, will vasopressin decrease myocardial perfusion? And, in a normal patient with normal coronary arteries, will vasopressin precipitate a myocardial infarction?
The answer seems to be dose-dependent. It is known that at low doses, vasopressin actually dilates coronary arteries (at the same time constricting peripheral non-essential vascular beds, thereby distributing blood flow to vital organs).
At moderate doses there is no significant effect on coronary vascular resistance. This is supported by human data: the VASST investigators reported on the cardiac safety of vasopressin in septic shock, and concluded that there was no significant increase in adverse cardiac effects at their studied dose range (0.01 - 0.03 units/min, or 0.6-1.8 units per hr).
However, at high doses myocardial perfusion seems to suffer, and with it the cardiac output. So one would be advised to avoid high-dose vasopressin in patients with untreated coronary artery stenosis, as it could be counterproductive. With a high enough dose, even healthy coronaries will vasoconstrict enough to produce ischaemia. This correlates with human studies which report an increase in the frequency of cardiac arrest among septic patients receiving more than 0.05 units/min (3 units/hr).
Again, there is a strange relationship of dose and response here. At levels resembling normal circulating levels, activation of V1 vasopressin receptors in the heart increases cardiac output. So, it acts as an inotrope, but the doses required for this effect are very small, and one may never see these effects in clinical practice.
As the dose is increased, coronary perfusion decreases because of coronary vasoconstriction. Combine this decrease in myocardial perfusion with the savage increase in afterload, and you can see why the cardiac output is observed to decrease with high dose vasopressin infusion.
Yes, perhaps in low (non-pressor) doses vasopressin has some vasodilatory effects on cerebral arteries. However, we also know that at higher doses, in the anaesthetised rat cerebral blood vessels are constricted by vasopressin. One might be forgiven for thinking that this has negative implications for the human brain, particularly in the setting of cerebral vasospasm (i.e. following a subarachnoid haemorrhage). Though some authors have reported its (relatively safe) use as a supplemental vasopressor in SAH, they caution their readers to carefully watch for vasospasm.
Vasopressin infusion decreases pulmonary arterial pressure in the rat. The authors hypothesise that this is a nitric oxide related effect. Among the work performed on humans, we may view studies in septic shock patients and those recovering from cardiopulmonary bypass. Each group has failed to demonstrate much of an increase in pulmonary arterial pressure at therapeutic doses. This is a contrast with catecholamines, which (owing to the alpha-1 receptors in the pulmonary circulation) can cause significant pulmonary vasoconstriction.
Investigators who marinaded some rat tail arteries in acidic brine had concluded that in conditions of metabolic acidosis, alpha-1 receptor responses are blunted. However, no such blunting occurred for vasopressin receptors; the arteries constricted as briskly as ever. This suggests that in severe metabolic acidosis vasopressin receptor sensitivity is well preserved.
This is perhaps the most popular application for vasopressin.
There was a period in the history of vasopressin use, during which it was thought to be somehow superior to noradrenaline as a single agent for severe septic shock. Subsequent studies have demonstrated that together, the vasopressors are more effective than when they are used alone, and that there is no real difference between them in terms of survival.
So, even though it is an apparently safe way to augment haemodynamic support for septic shock patients, the use of vasopressin does not seem to confer a survival benefit, which is sad news for the vasopressin enthusiast.
Given the relatively benign effect of moderate-dose vasopressin on cardiac output, pulmonary vessels and coronary vascular resistance, it is surprising how little literature there is on the subject of its use in cardiogenic shock. One retrospective analysis of 36 patients who developed cardiogenic shock after MI has confirmed that it increases MAP without any additional fatal cardiac embarrassment. Similarly, the Germans have published a small case series.
Vasopressin may also be an attractive alternative to traditional catecholamines in the treatment of certain catecholamine-intolerant cardiogenic shock states as HOCM and Takotsubo cardiomyopathy (where you cannot afford to stimulate the beta-1 receptors).
The profoundly vasodilated state of the recently cardiotomised post-bypass patient calls for harsh vasopressors, and arginine vasopressin seems a logical choice. Indeed, this has been experimented with.
In a randomised controlled trial (where vasopressin was compared to placebo and standard care) a peri-operative infusion of 2 units/hr seemed to decrease post-CABG vasoplegia. A review article of vasopressor selection in vasoplegic shock has recommended the use of vasopressin for this purpose.
Additionally, vasopressin seems to have an advantage over noradrenaline when it comes to managing the hypotension which resulted from the use of milrinone. The authors of a 52-patient study of these drugs concluded that the beneficial effect of this vasopressin-on-milrinone interaction lies in the ability to increase systemic vascular resistance while decreasing right ventricular afterload.
There has been some interesting data arising from some dog models of irreversible haemorrhagic shock, resistant to volume replacement and catecholamines. In these dogs, MAP was effectively restored with vasopressin, when all else had failed. Paramilitary-sounding articles call attention to the fact that its use in the field may prevent progression to cardiac arrest in severe trauma. This contrasts with the empirical finding that vasopressin use is associated with increased 72-hour mortality in trauma patients (from 41% to 55%). A recent review article summarised the available animal studies and case series, and was unable to recommend anything except more studies. One such study is the VITRIS trial, carried out in the European prehospital setting. According to their website, it is still going.
As previously discussed, there is really only one study of vasopressin as a supplementary vasopressor in subarachnoid haemorrhage, and it was far from conclusive- though i this small series (22 patients) vasopressin had no adverse effects on cerebral perfusion, the authors were less than enthusiastic about its broad applicability, recommending that people make their own decisions.
Drugs which increase the effect of vasopressin
Drugs which inhibit the effect of vasopressin
There is no convenient antagonist to this drug. A vasodilator infusion of some variety may be useful.
Vasopressin receptor antagonists (such as the "vaptan" group of drugs) may play some role...