Vasopressin: the preferred vasopressor in sepsis? ... Not today, not yet.

Crit Care Resuscitation ISSN: 1441-2772 4 September 2006 8 3 239-240 © C r i t C a r e R es u s c i t a t i on 20 06 www.jficm.anzca.edu.au/aaccm/journal/publications.htm Points of view retic hormone, and is deficient in patients w shock. It is therefore being increasingly consid alternative or additional vasopressor therapy patients. Small cases series and a small pha suggest it as clinical benefits, causing some c rapidly adopt vasopressin as a routine clinic Although the concept of replacing a deficie Vasopressin acts as both a vasopressor and an antidiuith septic ered as an for these se II trial1 linicians to al therapy. ncy of an endogenous vasopressor with a low therapeutic dose is appealing, we caution that all published clinical studies have been very small, that the risk–benefit ratio is unclear, and that routine clinical use in patients with septic shock is premature. The outcomes from a large international randomised trial are to be reported later in 2006. Vasopressin is an endogenous nonapeptide, which is synthesised in the hypothalamus, stored in the posterior pituitary, and released in response to hypotension, decreased intravascular volume and hyperosmolality.2 Vasopressin mediates vasoconstriction by activating V1 receptors on vascular smooth muscle, and has an antidiuretic effect by activating V2 receptors on renal collecting ducts. Vasopressin also increases plasma cortisol concentrations by nitric oxide-mediated release of corticotropin (ACTH) via central V3 receptors.3,4 Interestingly, low plasma concentrations of vasopressin also mediate vasodilatation of the coronary, cerebral and pulmonary arterial circulations.5-7 Vasodilatation of these key vascular beds, together with other favourable effects on the kidney (increased urine output and improved renal function despite systemic vasoconstriction) are the key reasons why vasopressin may become a useful adjunct therapy in patients with septic shock, in comparison with conventional non-selective vasoconstrictors used alone. In septic shock, there is an endogenous deficiency of vasopressin. Vasopressin is stored in the posterior pituitary, but only 10%–20% of the total hormonal pool can be rapidly released.2 In keeping with this limited capacity for acute release, vasopressin levels are very high in early haemorrhagic and septic shock, but low in established septic and vasodilated shock. Landry and colleagues first reported inappropriately low mean vasopressin levels of 3.1 pg/mL in patients with vasodilated septic shock, compared with 4 pg/mL in normal patients, and an increase to 22.7 pg/mL in those with cardiogenic shock.8 This led to the concept of endogenous vasopressin deficiency in patients with septic shock, and to the hypothesis that vasopressin deficiency contributes to their mortality. Potential mechanisms for vasopressin deficiency include depletion of pituitary stores, autonomic dysfunction during septic shock, and increased vascular release of nitric oxide, which may down-regulate vasopressin production in the posterior pituitary.2 However, low-dose vasopressin infusions are sufficient to restore the endogenous deficiency — with intravenous infusions of 0.01 IU/min resulting in a plasma level of 30 pg/mL, and 0.04 IU/min resulting in 100 pg/mL.8 Furthermore, and in keeping with clinical observations in individual cases, higher-dose infusions of vasopressin may be counterproductive. In patients with septic shock, vasopressin infusions in doses of 0.06–1.8 IU/min were associated with decreased cardiac output, decreased oxygen delivery, and gastric mucosal ischaemia suggested by gastric tonometry.9 Not only are low infused doses sufficient to restore more “appropriate” blood vasopressin concentrations for patients with shock, but the vascular effects of infused vasopressin are also dose-dependent. In high doses, vasopressin causes non-selective vasoconstriction. In low doses, vasopressin potentiates the alpha effects of noradrenaline and adrenaline, but also causes vasodilatation of the critical vascular beds (coronary, renal, cardiac and pulmonary). Case reports and small series have reported that vasopressin infusion allows dose-sparing of other infused vasopressors (usually noradrenaline) in various groups of patients with shock, including those with septic shock.10-13 For example, a retrospective series of 50 patients with septic shock in Vancouver, Canada, reported decreased noradrenaline requirements and increased urine output after varying doses of infused vasopressin.2 In this series, mortality was very high (85%), perhaps because vasopressin infusions were often commenced in an attempt at salvage when death was imminent. However, simply replacing one vasopressor with an equivalent alternative is unlikely to improve patient outcomes. The next level of support for vasopressin therapy is found in a small, prospective, phase II randomised trial in septic shock patients in Vancouver.1 This