Special Feature

Vasopressin in Septic Shock

By Uday B. Nanavaty, MD

Septic shock is one of the common causes of hypotension in the ICU. For decades, adrenergic agents with variable alpha- and beta-adrenergic activities have been the mainstay among vasopressor agents. Very few randomized controlled trials have been done that identify the efficacy or superiority of any one particular agent. Amongst the vasopressors in common use, norepinephrine (NE) and phenylephrine remain widely used, with dopamine losing favor over the last decade. Epinephrine is rarely used as the first vasopressor of choice.

Reviewing the available literature, it seems that NE is most commonly used as the vasopressor of choice in septic shock. However, it remains a common observation that in some patients NE is ineffective as a vasopressor. There are multiple potential pathophysiologic reasons for the failure of adrenergic vasopressors. Due to the lack of effectiveness of NE and other pressors in some patients with septic shock, as well as concern about toxicity at higher doses, alternatives have been sought. Vasopressin has become one alternative that is gaining increasing acceptance, although both the safety and efficacy of vasopressin in septic shock remain unproven. Here, I review the physiologic basis of using vasopressin in septic shock, as well as the available limited data about its safety and efficacy.1,2

Vasopressin in Health

Vasopressin, also known as antidiuretic hormone (ADH), is a well-known peptide hormone that is formed primarily in the supraoptic and paraventricular nuclei of hypothalamus and secreted in the posterior pituitary gland. ADH is secreted in response to osmotic stimuli and plays a vital role in the reabsorption of water in the collecting tubules and ducts in the kidney. Absence of ADH leads to central diabetes insipidus, a condition characterized by large water losses. Resistance to actions of ADH on kidney tubules results in a similar condition called nephrogenic diabetes insipidus. The effects of ADH on water regulation are observed in response to osmotic stimuli and are observed at minute physiologic concentrations.

Aside from these effects on kidney function, higher concentrations of ADH are associated with arteriolar vasoconstriction—hence the name vasopressin. Decreased blood volume and or a decrease in blood pressure exert their effects via dis-inhibition of atrial stretch receptors as well as decreased stretch of baroreceptors in the carotid, aortic and pulmonary vasculature.

The primary regulator of ADH secretion seems to be the osmolarity of extracellular fluids, with a 1% increase in osmolarity resulting in 7-fold increase in ADH secretion. Comparatively, more than a 10% decrease in blood volume is required before ADH secretion is affected. Besides osmolarity, blood pressure and blood volume, such diverse stimuli as hypoxia and nausea as well as wide variety of chemicals such as morphine, nicotine, and alcohol, as well as drugs such as cyclophosphamide and clonidine, affect ADH secretion. The vasoconstrictor effects are mediated through V1 vascular receptors, whereas the anti-diuretic effects are mediated through V2 renal receptors.

Vasopressin Levels in Septic Shock1,3,4

Animal and human studies suggest that early on in septic shock, ADH levels are markedly elevated and after several hours of septic shock, levels of circulating ADH drop. ADH levels remain relatively low in patients who remain hypotensive after approximately 24 hours of therapy for septic shock. In patients with hemorrhagic shock, the levels of ADH are elevated initially, as would be expected based on physiological stimuli.

The precise mechanisms of low ADH levels in persistent septic shock are unclear. It is possible that the stores of vasopressin are exhausted in patients with persistent vasodilatory shock. It is also described that there may be autonomic dysfunction in baroreflexes. High levels of NE (endogenous or exogenously administered) are known to inhibit release of vasopressin, and nitric oxide overproduction in the posterior pituitary may have an inhibitory effect.

Effects of Vasopressin Infusion in Patients with Septic Shock

Landry et al3 first described their findings of using vasopressin infusion in vasodilatory shock and suggested a concept of "vasopressin hypersensitivity" as very low doses of vasopressin generated "pressor" responses. Since then, several investigators have evaluated the effects of vasopressin in small studies5,6 or more often in form of case series and at times in retrospective fashion. The published literature is so scant that a meta-analysis would not be of significance but several similarities exist to draw the following conclusions.

Is Vasopressin Effective as a Vasopressor in Septic Shock?1,5,6

It is clear that vasopressin, as an infusion, is associated with improvement in blood pressure and reduction in the doses of pressors required to maintain reasonable blood pressure in patients suffering from septic shock. It has been observed in several small studies that abrupt withdrawal of vasopressin will result in hypotension and re-institution of vasopressin results in prompt response with elevation of blood pressure. Similarly, it has been consistently noted that patients who receive vasopressin infusion with septic shock often demonstrate increased urine output. It is believed that this response may be simply due to increased renal perfusion pressure. As with other agents, failure to respond to pressor doses of vasopressin has been reported as well.

What Dose of Vasopressin Is Safe in Treating Hypotension Associated with Septic Shock

The smallest possible dose of vasopressin should be used to achieve the mean arterial pressure goal. Although a wide variety of doses have been used, several different reports suggest that to avoid possible cardiac and other side effects, vasopressin should only be used in doses between 0.01 to 0.04 U/min. Effectiveness of vasopressin as a pressor agent has been demonstrated for the lower doses and at these small doses, very few if any side effects have been reported. Most of the studies that report few side effects have used vasopressin for 4 to 24 hours only. Two small studies have reported that even in small doses, vasopressin infusion can result in impaired gastric mucosal blood flow. This phenomenon of redistribution of blood flow may have potential harmful effects. Some of the studies have also reported a drop in cardiac index (CI). Most studies however have not reported any significant changes in lactate levels or other indices to suggested impaired systemic perfusion. Peripheral extravasation of vasopressin solution used for these low doses has been reported to induce skin necrosis.

When Should One Use Vasopressin?

When vasopressin infusion should be initiated remains unclear. Unfortunately, various authors have used different doses of NE prior to starting the study of vasopressin. There is no study that compares vasopressin against NE as the first line treatment of hypotension during septic shock. Based on the limited literature, vasopressin infusion should be used after volume resuscitation and failure of pharmacologic doses of NE or other vasoactive agents. If escalating doses of pressor agent fail, it would be reasonable to try vasopressin.

Can Vasopressin Be Used To Replace Norepinephrine?

At least one study tried to replace NE with vasopressin. Klinzing et al7 studied substitution of vasopressin for NE. They found that much higher dose of vasopressin (0.06-1.8 U/min) was needed to maintain blood pressure. They also observed that at these doses of vasopressin, CI tended to decrease, heart rate tended to decrease and although splanchnic blood flow was maintained, there was concern that mucosal blood flow may be reduced. Thus it seems reasonable to use vasopressin in low doses (0.01-0.04 U/min) enough to achieve the blood pressure goals when NE seems to be failing. Currently available studies suggest that replacing NE with vasopressin may have harmful effects.

Does Vasopressin Infusion Improve Outcome in Septic Shock?

There are no studies that have looked at the effect of vasopressin infusion on the outcome of septic shock. It would be unwise to conclude that improvement in blood pressure will automatically result in improved outcome. We have seen the effects of nitric oxide inhibition with dramatic improvement in blood pressure in septic shock patients. Unfortunately, inhibiting nitric oxide resulted in higher mortality in a large randomized controlled study.

No significant change in creatinine clearance or dialysis requirement has been noted, and it is not possible at this point to draw any conclusions about improved renal function.

When Should We Avoid Vasopressin?

Most authors have excluded any element of cardiogenic shock either by using invasive monitoring and excluding patients with low CI or by using echocardiographic assessment. Patients with unstable angina or recent myocardial infarction should be excluded as well.

What Do We Know About Terlipressin in Septic Shock?

Terlipressin is a long acting analogue of vasopressin. There are some case reports that suggest that small doses of terlipressin administered every six hours result in improved blood pressure in patients with septic shock. Again, no randomized controlled studies have been performed to assess the safety or efficacy of intermittent dosing of terlipressin in septic shock.


It seems safe to conclude that vasopressin acts as a pressor agent in the dose range of 0.01 to 0.04 U/min in septic shock-associated hypotension, especially if the patient remains hypotensive in spite of adequate fluid resuscitation and conventional pressor therapies. One should be careful in making sure that the patient does not have an underlying cardiac problem that can be exacerbated by vasopressin before starting this unproven therapy. Care should be taken to avoid complications such as skin necrosis. In view of lack of evidence to the benefit of using vasopressin in septic shock, extreme care should be taken in its use, at least until it is proven that the therapy does not harm the patient. Terlipressin, a vasopressin analog, may become valuable, if it can be shown to be effective and safe in treating septic shock-associated hypotension.


1. Holmes CL, et al. Chest. 2001;120:989-1002.

2. Russell JA. Crit Care Med. 2003;31:2707-2709.

3. Landry DW, et al. Circulation. 1997;95:1122-1125.

4. Sharshar T, et al. Crit Care Med. 2003;31:1752-1758.

5. Patel BM, et al. Anesthesiology. 2002. 96:576-582.

6. Tsuneyoshi I, et al. Crit Care Med. 2001;29 :487-493.

7. Klinzing S, et al. Crit Care Med. 2003;31:2646-2650

Uday B. Nanavaty, MD, Pulmonary and Critical Care Medicine, Rockville, Maryland, is Associate Editor for Critical Care Alert.