Arginine Vasopressin for the Treatment of Septic Shock in Adults

Seth R. Bauer, Pharm.D.; Simon W. Lam, Pharm.D.


Pharmacotherapy. 2010;30(10):1057-1071. 

In This Article

Dosing of Arginine Vasopressin

Physiologic versus Pharmacologic Dosing

The Surviving Sepsis Campaign recommends arginine vasopressin 0.03 unit/minute as an adjunct to first-line catecholamines for hemodynamic management of septic shock;[3] however, this dosing strategy has been questioned by others.[26] Numerous studies have evaluated different dosing regimens for arginine vasopressin in patients with vasodilatory shock, ranging from 0.01–1.8 units/minute, with doses of up to 0.04 unit/minute being termed "physiologic" and doses above that termed "pharmacologic."[8,15,21–23,27–31] A major challenge in the attempt to determine which dose of arginine vasopressin should be used is that the true physiologic arginine vasopressin response during septic shock has not been fully elucidated.

One group of authors was the first to demonstrate that patients in advanced, late-stage, catecholamine-resistant vasodilatory shock had inappropriately low endogenous plasma levels of vasopressin.[8] In this study, all patients with septic shock had hypotension, which persisted after adequate fluid administration (determined by pulmonary capillary wedge pressure ≥ 12 mm Hg), and required catecholamines to maintain systolic blood pressure greater than 90 mm Hg for 1–2 days. In this septic population, the mean vasopressin concentration was 3.1 pg/ml. In contrast, a group of patients with hypotension of similar duration and severity due to cardiogenic shock had an appropriately increased vasopressin concentration of 22.7 pg/ml. Based on this observation, the authors concluded that plasma vasopressin concentrations are inappropriately low in patients with vasodilatory shock.

After this initial pioneering work, many investigators recommended the use of arginine vasopressin as a low-dose, 0.01–0.04 unit/minute, physiologic replacement in patients with septic vasodilatory shock.[6,28,32] The investigators also recommended that the physiologic replacement dose should be targeted to a "normal" arginine vasopressin response to hypotension (i.e., 20–30 pg/ml) as demonstrated by the patients in cardiogenic shock from the initial investigation.[8] However, in the literature, the range of endogenous plasma vasopressin levels during acute vasodilatory events is vast, ranging from 3.1–53.8 pg/ml.[8,15,22,27,33–38] Furthermore, during cardiopulmonary arrest, arguably the most severe form of shock, the initial plasma vasopressin concentration can be as high as 193 pg/ml.[39] In addition, the levels of initial vasopressin surge during endotoxemic shock in animal models ranged from 180–1800 pg/ml.[6] Thus, clinicians are left with the following questions: which "normal" vasopressin level should be targeted, and does it make a difference?

To further complicate matters, the exogenous administration of low-dose arginine vasopressin 0.01–0.04 unit/minute for the treatment of vasodilatory shock during clinical investigations has yielded vast differences in plasma vasopressin concentrations, ranging from 30–289 pg/ml.[8,15,40] Similarly, the administration of pharmacologic doses of arginine vasopressin (> 0.04 unit/min) have also yielded a considerable range in vasopressin concentrations (17.1–745 pg/ml).[21,27,36] One unifying point about all these clinical investigations is that both physiologic and pharmacologic doses of arginine vasopressin routinely yield plasma levels that exceed what is traditionally seen as physiologic replacement levels (30 pg/ml). In the initial investigation,[8] the use of an infusion of 0.04 unit/minute of arginine vasopressin caused the plasma concentration to increase to nearly 100 pg/ml, which corresponded with an increase in systolic blood pressure from 92 to 146 mm Hg. The use of a 0.01–unit/minute infusion in two patients led to an increase in plasma concentrations to about 30 pg/ml and also corresponded with an increase in systolic blood pressure.[8] Similarly, in VASST, a maximum dose of arginine vasopressin of 0.03 unit/minute corresponded with a plasma concentration of about 108 pg/ml after 24 hours of arginine vasopressin infusion.[15]

Rather than focusing on whether a certain dose of exogenous arginine vasopressin will adequately replenish the body's normal physiologic response to hypotension, perhaps a better question would be as follows: do higher doses of arginine vasopressin provide clinical improvements? Studies in this area are underpowered to evaluate hard-outcome differences and primarily focus on hemodynamic measurements.

Titration versus Fixed Dosing

One group of authors found that higher doses of arginine vasopressin, sufficient for the replacement of norepinephrine, had mixed results in patients with septic shock.[29] In this study, 12 patients with septic shock who were receiving a mean norepinephrine dose of 0.56 μg/kg/minute were initiated on increasing doses of arginine vasopressin until norepinephrine could be weaned. Arginine vasopressin was started at 0.06 unit/minute and increased to as high as 1.8 units/minute. The investigators found that arginine vasopressin, at higher doses, could sufficiently support the blood pressure without the use of norepinephrine. However, the switch to arginine vasopressin from norepinephrine was associated with a decrease in cardiac index and increase in gastric regional partial pressure of carbon dioxide, suggestive of reduced gastrointestinal mucosal blood flow. Despite worsening adverse effects, the authors sufficiently demonstrated that increasing doses of arginine vasopressin could adequately replace norepinephrine requirements.

Another group subsequently evaluated, in a retrospective historical cohort design, two different dosing regimens for arginine vasopressin, 0.033 versus 0.067 unit/minute, for the treatment of vasodilatory shock.[31] In this study, 46% of the patients had septic shock, whereas the remainder had vasodilatory shock from either previous cardiotomy or systemic inflammation after major surgery. Patients were matched based on MAP and catecholamine requirements before starting arginine vasopressin, age, and severity of organ dysfunction. Overall, 87% of patients in both dose groups had an MAP increase within the first 30 minutes of starting arginine vasopressin therapy. Also, no significant difference was noted in the magnitude of MAP change during that time period. However, in the first 24 hours after arginine vasopressin initiation, norepinephrine requirements were reduced more frequently in patients receiving the 0.067-unit/minute dose of arginine vasopressin. No adverse effects were seen with the higher dose of arginine vasopressin. Conversely, more patients in the arginine vasopressin 0.033-unit/minute group had increases in hepatic transaminase levels (72% vs 28%, p<0.001). In addition, at the 12-hour time point, arterial lactate levels were significantly higher among patients who received 0.033 unit/minute (76 ± 67 vs 46 ± 38 mg/dl, p<0.001). The authors concluded that the 0.067-unit/minute dose of arginine vasopressin, compared with 0.033 unit/minute, appeared to be more effective in reversing cardiovascular collapse in patients with vasodilatory shock. Of note, at baseline, the median dose of norepinephrine was 1.07 μg/kg/minute. The authors speculated that the higher dose of arginine vasopressin may have been more beneficial because such doses are needed to stabilize hemodynamic function when norepinephrine requirements are high.

The same group of investigators studied the two different dosing regimens for arginine vasopressin in a prospective randomized controlled study for the treatment of vasodilatory shock. The study included 50 patients, with 25 patients randomly assigned to each group (Table 1).[24] All patients had vasodilatory shock and required norepinephrine doses greater than 0.6 μg/kg/minute to maintain MAP above 60 mm Hg. The primary objective was to compare the hemodynamic response of the two different arginine vasopressin doses at various time points up to 48 hours after randomization. The study defined hemodynamic response as either an MAP increase or norepinephrine decrease. The mean ± SD baseline norepinephrine doses were 0.98 ± 0.6 and 0.86 ± 0.34 μg/kg/minute in the arginine vasopressin 0.033- and 0.067-unit/minute groups, respectively.

Overall, MAP increased, and heart rate and norepinephrine dose decreased during the study period in both groups. However, patients who received arginine vasopressin 0.067 unit/minute required less norepinephrine during the study. Patients in both dose groups also had improvements in arterial pH, lactate level, and base deficit, suggesting an improvement in tissue perfusion. Plasma arginine vasopressin levels were increased in both groups but were higher in the 0.067-unit/minute group (p<0.001). The authors also noted that plasma arginine vasopressin levels and norepinephrine requirements were correlated inversely (R=−0.49, p<0.01), which suggests a dose-dependent arginine vasopressin effect. No increase in the rate of adverse effects was observed in the 0.067-unit/minute group. Since all patients in this study were receiving relatively high doses of norepinephrine and would be considered in "more severe" vasodilatory shock according to the VASST criteria, the authors hypothesized that the 0.01–0.03-unit/minute dose of arginine vasopressin used during the VASST study may have been insufficient for patients with more severe shock.

Taken together, these three studies demonstrate that higher doses of arginine vasopressin (> 0.04 unit/min) may lead to further blood pressure response. Thus, higher doses of arginine vasopressin may be considered in patients with refractory shock. However, we found no studies that suggest that the higher MAP achieved with the use of higher doses of arginine vasopressin leads to improved outcomes. Since arginine vasopressin doses as low as 0.01 unit/minute have yielded both physiologic replacement levels of vasopressin and blood pressure increase, it may be a reasonable starting dose for patients. Starting at a low dose and titrating to the lowest arginine vasopressin dose that restores cardiovascular function would be a similar approach to that used in the VASST study[15] (dose titrated to a maximum of 0.03 unit/min) and other studies.[21,23,29]


Concerns for severe adverse effects from arginine vasopressin doses higher than 0.04 unit/minute have largely been based on retrospective, noncomparative, descriptive investigations.[28,30,41] One group retrospectively evaluated 50 patients who received arginine vasopressin for the treatment of severe septic shock.[28] Hospital mortality was 85% in these patients, with six occurrences of cardiac arrest. Five of the six cardiac arrests occurred when the patient was receiving arginine vasopressin doses of 0.05 unit/minute or more. The authors concluded that arginine vasopressin should be used in patients with severe septic shock at a fixed low dose of 0.01–0.04 unit/minute because beneficial effects are achieved at doses of 0.04 unit/minute or lower and adverse effects are more likely at higher doses. These statements by the authors may not be well founded based on the data presented in the study. The authors clearly state that arginine vasopressin was used in their ICU as rescue therapy in patients who appeared to be dying of refractory hypotension and that no conclusions can be drawn from this study since it was not from a matched cohort.

Similarly, another group retrospectively reported the frequency of and risk factors for development of ischemic skin lesions in critically ill patients with catecholamine-resistant vasodilatory shock who were treated with a continuous arginine vasopressin infusion.[41] In this investigation, 63 patients were included, of whom 19 (30%) developed ischemic skin lesions. The median doses of norepinephrine at the time arginine vasopressin was started in patients who developed ischemic skin lesions and those who did not were 0.67 and 0.65 μg/kg/minute, respectively (p=0.708). Presence of septic shock and preexistent peripheral arterial occlusive disease were significant independent risk factors for the development of ischemic skin lesions during arginine vasopressin infusion. The median dose of arginine vasopressin used in this study was 0.0009 unit/kg/minute (range 0.0001–0.0022 unit/kg/min). For a 70-kg patient, this equates to 0.063 unit/minute (0.007–0.154 unit/min). The authors did not find any significant relationship between arginine vasopressin dosages or length of infusion and the development of ischemic skin lesions.

Contrary to the above retrospective analyses, prospective randomized studies have not found any significant differences in the frequency of adverse effects when comparing norepinephrine with arginine vasopressin[15,23] or pharmacologic doses with physiologic doses of arginine vasopressin. [24] These data suggest that the occurrence of adverse effects of arginine vasopressin in general and with arginine vasopressin doses above 0.04 unit/minute may be lower than previously hypothesized.

Hemodynamically, arginine vasopressin may decrease cardiac output and heart rate through increased vagal tone and decreased sympathetic tone and coronary blood flow.[6] Several observational studies have noted decreases in cardiac output during septic shock.[28,42,43] However, discrepant data exist among prospective randomized controlled studies.[19,20,23] As discussed above, patients with NYHA class III or IV heart failure were excluded from the VASST study.[15] A recent retrospective, observational study evaluated the use of arginine vasopressin in septic patients with heart failure.[44] Forty-eight patients were evaluated for mortality outcomes and hemodynamic changes 4 hours after the initiation of arginine vasopressin. The authors found a statistically significant increase in MAP and decrease in heart rate after the initiation of arginine vasopressin. Despite significant decreases in heart rate, no significant changes were seen in cardiac output. The mortality rate in this study was 83%, which was drastically higher than that reported in other septic shock studies. In addition, 15 patients (31%) experienced cardiac arrest during the study. Given the noncomparative, observational nature of the study, it is difficult to determine whether the poor patient outcomes were from baseline comorbidities or associated with the use of arginine vasopressin. However, clinicians should continue to be cautious when using arginine vasopressin in patients with actual, or the potential for, cardiac dysfunction.

Factors that may Affect Dosing

Patient Weight Certain patient and disease-state factors may provide a further level of complexity to the dosing of arginine vasopressin in patients with septic shock. Arginine vasopressin, unlike catecholamines, is not routinely dosed based on body weight. It is largely unknown whether a patient's weight may affect the vasopressin levels produced by exogenously administered arginine vasopressin, and clinical data in this area are sparse and conflicting. However, given the relative low volume of distribution of arginine vasopressin (0.14 L/kg), it is conceivable that the extremes of body weight may have considerable effect on the serum concentration generated from fixed-dosed arginine vasopressin infusion.[45]

We retrospectively evaluated the effect of body mass index (BMI) on hemodynamic response in 66 patients who received arginine vasopressin monotherapy for septic shock.[46] Hemodynamic response was defined as an MAP of at least 65 mm Hg at 1 hour after arginine vasopressin initiation. Patients were divided into four groups based on predefined BMI ranges. The mean BMI values in the four groups were 22.4, 27.8, 32.2, and 45.0 kg/m2, respectively. Overall, 91% of the patients achieved a hemodynamic response to arginine vasopressin. There was no significant difference in response rate or change in MAP among the four BMI groups. All patients in this study received arginine vasopressin 0.04 unit/minute with no titration.

Although this study did not find a correlation between body mass and hemodynamic response to arginine vasopressin, caution should be exercised in its interpretation because of a number of limitations. As previously mentioned, even low-dose arginine vasopressin could yield a wide range of vasopressin levels. Therefore, it is conceivable that an arginine vasopressin dose of 0.04 unit/minute, even for the patients in the highest BMI quartile, may have reached vasopressin levels capable of producing a hemodynamic effect. Furthermore, all patients in this study were receiving arginine vasopressin alone for the hemodynamic management of septic shock. Based on this study, it is unknown whether body mass affects hemodynamic response to arginine vasopressin in patients who have significant catecholamine requirements.

A similar retrospective study evaluated 55 patients who were receiving concomitant catecholamine and arginine vasopressin therapy.[47] The primary outcome measure was a Pearson correlation between change in norepinephrine requirements and arginine vasopressin dosing normalized by body weight. The investigators found a significant correlation (R=−0.63, p<0.001); however, a linear regression model revealed no predictive utility of weight to hemodynamic response to arginine vasopressin.

Continuous Renal Replacement Therapy Patients with septic shock frequently develop acute renal failure, and continuous renal replacement therapy (CRRT) is increasingly used in this population.[48] Arginine vasopressin has many properties that makes its removal from continuous dialysis plausible. It is a relatively small molecule (molecular weight 1084 daltons) that is not protein bound, with a plasma half-life of 24 minutes.[45] The active compound is 65% renally eliminated, with the remainder of drug metabolized by tissue peptidase.[45] Studies have demonstrated that arginine vasopressin may be completely dialyzable and completely removed by ultrafiltration. In fact, it appears that the mean ± SD sieving coefficient for arginine vasopressin is 0.85 ± 0.11,[49] and arginine vasopressin concentrations found in ultrafiltrate and dialysate are almost identical with those of control plasma.[50]

In a recent study, the correlation between arginine vasopressin and copeptin response was PHARMACOTHERAPY Volume 30, Number 10, 2010 evaluated in patients with septic shock.[36] Copeptin is a stable precursor to arginine vasopressin, which is more readily accessible in clinical settings. In a linear regression model of 44 patients with vasodilatory shock, plasma concentrations of arginine vasopressin and copeptin significantly correlated with each other (r=0.749, p<0.001). This correlation was not influenced by creatinine clearance, but by the use of CRRT (p=0.001). In fact, patients not receiving CRRT had a correlation coefficient of 0.803, whereas the correlation coefficient was 0.385 for patients receiving CRRT. Furthermore, in exploratory regression analysis, arginine vasopressin levels inversely correlated with the use of CRRT. The authors suggested that this may indicate that more arginine vasopressin is removed during CRRT than the endogenous arginine vasopressin production rate.

With regard to the decreased correlation between copeptin and arginine vasopressin in patients receiving CRRT, it is conceivable that CRRT removes differing amounts of these two substance from the plasma. Of note, the molecular weight of copeptin is 4021 daltons,[51] which suggests it may be less readily removed by CRRT than arginine vasopressin. This removal of arginine vasopressin may lead to hemodynamic instability in patients with septic shock, unmasking relative vasopressin deficiency.

In summary, further research is needed to define the optimal dosing of arginine vasopressin. Patient and disease factors that might affect arginine vasopressin levels during infusion should be investigated. These issues are particularly important for those who adopt a "one size fits all" arginine vasopressin dosing strategy.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.