Hemodynamic effects of adjunct arginine vasopressin to norepinephrine in septic shock: insights from a prospective multicenter registry study

Abstract

Background: The Surviving Sepsis Campaign guidelines suggest adding arginine vasopressin (AVP) when norepinephrine (NE) doses reach 0.25-0.50 µg/kg/min in septic shock patients. However, relying solely on a NE threshold has limitations, as other factors may be valuable in guiding AVP therapy during septic shock. Therefore, we aimed to identify additional patient characteristics associated with AVP hemodynamic responsiveness.

Methods: A multicenter, prospective, observational study was conducted among adult ICU patients who met the predefined criteria for septic shock (not reaching the individual target mean arterial pressure despite adequate fluid resuscitation and NE base dose > 0.25 µg/kg/min) and received AVP therapy. AVP hemodynamic responsiveness was the primary study outcome, defined as stabilization or decrease of NE infusion rate two hours after initiating AVP. Secondary outcomes included shock duration and rebound hypotension following termination of AVP infusion. Univariate and multivariable regression analyses were performed to detect associations between characteristics and outcomes.

Results: Between May 2020 and October 2023, 200 septic shock patients originating from 11 different ICUs were included. Of these, 153 (79%) met the definition for AVP hemodynamic responsiveness. Obesity and hyperlactatemia was negatively associated with AVP-response (adjusted Odds Ratio [aOR] 0.30, 95%CI 0.14-0.65 and aOR 0.86, 95%CI 0.75-0.99, respectively), while a NE infusion rate ≥ 0.30 µg/kg/min showed positive odds of AVP response (aOR 2.33, 95%CI 1.06-5.14). Incidence of new-onset atrial fibrillation was lower in AVP responders than non-responders (4% vs. 14%, p = 0.013). Higher body mass index (BMI) , NE infusion rate and duration prior to AVP initiation was associated with longer shock duration (aOR 1.06, 95%CI 1.02-1.11, aOR 1.12, 95%CI 1.01-1.25, and 1.01 95% CI 1.00-1.03, respectively), while higher pH associated with lower likelihood of prolonged shock (aOR 0.80, 95%CI 0.64-0.99). Rebound hypotension occurred in 9% when AVP was terminated, and AVP duration > 24 h was negatively associated with rebound hypotension (OR 0.22, 95%CI 0.05-0.85).

Conclusions: Arterial lactate, pH, BMI, and NE duration and dose were associated with AVP responsiveness and shock duration during septic shock, and rebound hypotension occurred in 9% during recovery. Our findings suggest that beyond NE thresholds, specific factors could be considered to optimize adjunctive AVP therapy in septic shock patients.

Keywords: Arginine vasopressin; Body mass index; Intensive care unit; Norepinephrine; Sepsis; Shock.

Fig. 1

Flowchart of the study population. * No NE infusion rate registered at AVP initiation (n = 4), NE infusion rate only registered once daily (n = 2) or NE infusion rate was only registered at AVP initiation but not during follow-up (n = 1). ICU Intensive Care Unit, eCRF electronic case report form, DNR Do-Not-Resuscitate, AVP Arginine vasopressin, NE Norepinephrine

Fig. 2

Comparison of dynamics between AVP responders and non-responders. A. NEE dynamics before and after start of AVP. A LMM was conducted applying robust standard errors to evaluate the difference in NEE dynamics between responders and non-responders during the five hours prior to AVP initiation. The change in NEE was not significantly different between responders and non-responders (p = 0.716). B. Arterial lactate levels at baseline and after start of AVP. A LMM showed responders had lower baseline lactate (p = 0.018), but lactate levels did not change significantly over time (p = 0.517), nor indicated different trajectories between responders and non-responders over time (p = 0.980). C. Mean Arterial Pressure (MAP) at baseline and after start of AVP. Results of LMM showed a significantly higher baseline MAP in responders compared to non-responders (p = 0.005) and significant increase in MAP over time overall (p < 0.001), but similar trajectory among responders and non-responders (p = 0.075). After applying robust standard errors due to heteroscedasticity the baseline difference remained significant (p = 0.029) with an overall significant increase over time (p = 0.015), but no significant interaction effect with response (p = 0.170), indicating similar MAP trajectories between groups. D. Net fluid balance at baseline and after start of AVP. A LMM showed no significant difference at baseline net fluid balance between responders and non-responders (p = 0.993). Non-responders accumulated fluid at an average rate of 1.5 ml/kg IBW per hour more than responders (p < 0.001), which remained significant after applying robust standard errors (p = 0.021) due to heteroscedasticity. In all figures medians alongside interquartile ranges are presented. NEE Norepinephrine Equivalent, AVP Arginine vasopressin, MAP Mean Arterial Pressure, IBW Ideal Body Weight, LMM Linear Mixed Model