Vasopressin Plasma Concentrations in Responders and Non-responders to Exogenous Vasopressin Infusion in Patients With Septic Shock

Vasopressin is an endogenous hormone that decreases serum osmolarity and increases blood pressure. As a part of the stress response to hypotension, vasopressin is released from the posterior pituitary and leads to vasoconstriction through agonism of the vascular vasopressin V1 receptor. In patients with septic shock, endogenous vasopressin levels are initially elevated but quickly fall to levels at or below those of normal physiology (1.4-3.6pg/mL) because of the depletion of endogenous store. Sharshar et al. evaluated two sets of patients with septic shock, one of which was evaluated earlier in the septic shock course (3.6 ± 2.3 hours, n=18) and one evaluated at a later time from shock onset (mean 38.7 ± 28.4 hours, n=44). The group of patients evaluated earlier in their septic shock course were more likely than patients evaluated later to have elevated (>3.6 pg/mL) plasma vasopressin levels (88.9% vs. 38.6%, respectively). Similarly, a case series evaluated single vasopressin levels in three patients with septic shock, one of whom was in the first day of shock onset and two of whom were in the fifth and sixth day of shock onset. The patient in the earlier stages of septic shock had a plasma vasopressin level that was increased (16pg/mL), while the two patients in the later stages of septic shock had decreased plasma vasopressin levels (1.6 and 1.8pg/mL). The exact timing of when patients transition from having elevated endogenous vasopressin levels to having normal levels of vasopressin is currently unclear. In a clinical trial enrolling patients within the first 12 hours of shock onset, median endogenous vasopressin levels were 3.5 pg/mL (interquartile range 1.8, 5.3 pg/mL; n=54). Some have even hypothesized that vasopressin levels rise before clinical hypotension is apparent and the decline in vasopressin levels is associated with the onset of apparent hypotension. Further complicating this issue, endogenous vasopressin levels have been shown to be lower in patients with septic shock compared to other shock etiologies such cardiogenic shock (3.1 ± 1pg/mL in patients with septic shock vs. 22.7 ± 2.2pg/mL in patients with cardiogenic shock, p<0.001). The etiology of this discrepancy in endogenous vasopressin response by shock type is unclear, but a "relative deficiency" of vasopressin is theorized to exist in patients with septic shock.

In light of these findings, exogenous arginine vasopressin (AVP) has been added to exogenous catecholamines to increase mean arterial pressure (MAP) and to decrease catecholamine requirements in patients with vasodilatory shock. The use of AVP for these purposes in patients with septic shock is in keeping with the Surviving Sepsis Campaign Guidelines. In the Vasopressin and Septic Shock Trial (VASST), low-dose AVP was infused at a rate of 0.01-0.03 units/min in combination with norepinephrine to achieve a goal MAP of 65-75mmHg. Plasma vasopressin levels in patients receiving AVP were elevated at 6 (68.3pg/mL) and 24 hours (90.5pg/mL) in comparison to patients not receiving AVP (3.0pg/mL at baseline with no significant change at 6 or 24 hours). Association of plasma vasopressin levels with hemodynamic response to AVP, though, was not evaluated in VASST.

Concomitant corticosteroid use has been observed to decrease the total dose of administered AVP, to increase the proportion of patients alive and free of vasopressors at day 7, to increase plasma vasopressin concentrations by 33% at 6 hours and 67% at 24 hours, and to lead to lower 28- and 90-day mortality (35.9% vs. 44.7%, p=0.03 and 42.5% vs. 55.5%, p=0.01, respectively) than in those that received AVP alone. These findings generated the hypothesis that concomitant administration of AVP and corticosteroids results in increased plasma vasopressin levels versus AVP administration alone, leading to positive clinical outcomes in septic shock. Furthering the hypothesis that plasma vasopressin levels may influence outcomes in septic shock, genetic differences in leucyl/cystinyl aminopeptidase, the primary vasopressin metabolic enzyme, have been associated with more rapid vasopressin clearance, lower plasma vasopressin levels, and increased mortality in patients with septic shock. However, a study evaluating vasopressin plasma concentrations in patients with multiple shock types not administered exogenous AVP observed higher vasopressin concentrations in those with hemodynamic dysfunction than in those without (mean 14.1 ± 26 vs. 8.7 ± 10.8pg/mL, respectively) regardless of shock type. This suggests that plasma vasopressin concentration may not directly correlate with MAP.

The impact of body mass (which may influence vasopressin levels when fixed-dose AVP is administered) on hemodynamic response to AVP has been inconsistent. Studies have observed a negative correlation between BMI and change in MAP at 6 hours and a correlation between increasing weight-adjusted AVP dose and reduction in catecholamine requirements, suggesting that hemodynamic response to AVP is associated with body mass. In contrast, a third study observed no association between BMI and AVP dose required to meet goal MAP when AVP was administered as the sole vasopressor. Finally, a fourth found an inverse correlation between BMI and APACHE II-adjusted 28-day mortality, regardless of the fact that overweight and obese patients received less weight-adjusted vasopressin than underweight or normal weight patients. This suggests that while BMI may impact plasma vasopressin concentration, the change in vasopressin concentration may not have an impact on clinical outcomes.

Recently, a retrospective study was completed at the Cleveland Clinic to evaluate predictors of hemodynamic response to fixed-dose AVP in patients with septic shock. Patients were considered to be responders to AVP if a decrease in catecholamine dose was achieved with MAP≥65mmHg at 6 hours. The overall response rate to fixed-dose vasopressin was 45.4%. Within this study, only admission to surgical or neurosciences intensive care units (ICU) vs. medical ICU and lower lactate level were associated with increasing chance of response to AVP (OR 1.71, 95% CI 1.175-2.463, p=0.0049 and OR 0.925, 95% CI 0.887-0.965, p=0.0003, respectively) on logistic regression. Factors previously found to impact vasopressin levels (such as concomitant use of corticosteroids) were not associated with hemodynamic response. However, plasma vasopressin levels were not evaluated in this retrospective study.