Vasopressor therapy is one of the cornerstones in the management of septic shock when intravenous fluid resuscitation is insufficient to maintain a mean arterial pressure (MAP) above 65 mmHg [1]. Norepinephrine is the recommended first-choice vasopressor, but hyporesponsiveness represents a significant clinical problem and high doses are sometimes required to achieve the target MAP. Such high-dose catecholamine therapy may increase the risk of life-threatening arrhythmias, immunosuppression, and mortality [2,3,4].

In response to concerns about high-dose norepinephrine therapy in septic shock, adjunctive treatment with vasopressin has been suggested [1]. In the Vasopressin and Septic Shock Trial (VASST), adding low-dose vasopressin (0.01–0.03 U/min) to existing norepinephrine treatment did not decrease mortality in patients with septic shock compared with norepinephrine mono-therapy [5]. However, among patients with less severe shock and among those enrolled within 12 h there appeared to be a possible survival benefit with vasopressin, suggesting that early initiation may be beneficial.

Yet, the Vasopressin vs Norepinephrine as Initial Therapy in Septic Shock (VANISH) trial found no difference in kidney failure-free days or mortality with early (within 6 h) initiation of vasopressin therapy (up to 0.06 U/min) compared with norepinephrine alone among 409 septic shock patients [6].

In addition to vasoconstriction via vasopressin V1-receptor activation, vasopressin may cause unwanted side effects via activation of V2-receptors in the renal collecting ducts (antidiuretic effect) and on endothelial cells (prothrombotic effect via Von Willebrand factor release), V3-receptors in the pituitary gland (increased ACTH secretion) and oxytocin-receptors on vascular endothelial cells (increased nitric oxide synthase activity causing vasodilation). Terlipressin, a synthetic vasopressin analogue with greater selectivity for the V1-receptor, may therefore be an attractive alternative to vasopressin.

Experimental animal data suggest that terlipressin attenuates fluid accumulation and prolongs survival time compared with vasopressin [7]. In addition, data from small randomised controlled trials suggest that terlipressin improves short term renal function in patients with type 1 hepatorenal syndrome [8] and that it may even improve survival in patients with liver cirrhosis and septic shock compared with norepinephrine [9]. However, data on the safety and efficacy of terlipressin therapy in septic shock patients without liver failure are scant.

In a recent article in Intensive Care Medicine, Liu et al. report the results of a randomised, multicenter, double blind, controlled trial comparing norepinephrine alone with early terlipressin infusion (20–160 µg/h) plus norepinephrine in patients with septic shock [10]. The study was stopped due to futility after enrolment of 50% of scheduled patients. In the 526 randomised and analysed patients, they found no difference in 28-day mortality (primary outcome; 38 vs. 40%, P = 0.63), vasopressor-free days or change in SOFA score during the first week after randomisation. This informative study hence confirms the findings of previous pilot investigations showing no mortality benefit when terlipressin is used alone or is added to norepinephrine (Fig. 1) [11,12,13].

Fig. 1
figure 1

Meta-analysis of mortality in randomised trials comparing terlipressin (Treatment) with norepinephrine (Control) in adult patients with septic shock

Full size image

The striking difference in serious adverse events, particularly digital ischemia, reported by Liu et al., requires attention. Overall, 1.5% of patients treated with norepinephrine alone suffered from digital ischemia; an identical incidence to VANISH trial patients receiving norepinephrine only. However, in the study by Liu et al., digital ischemia occurred in 13% of patients receiving terlipressin (c.f. 5.4% with vasopressin in the VANISH trial). In addition, the authors found a greater prevalence of diarrhea (2.7% vs 0.3%), but importantly not acute mesenteric ischemia, in the terlipressin group.

The high incidence of digital ischemia may have several possible explanations. Firstly, the risk of terlipressin-induced ischemic events increases with hypovolemia. In the study by Liu et al., most cases of digital ischemia occurred in the first 24 h. Unfortunately, the amount of pre-randomisation fluid administration was not reported. It is therefore unclear whether patients were “adequately” fluid resuscitated before entering the trial.

Secondly, some studies demonstrate that cardiac index is significantly decreased with terlipressin but not with norepinephrine [11, 12]. An increase in systemic vascular resistance at the expense of cardiac index may indeed compromise organ blood flow and contribute to adverse ischemic events. But since cardiac index was not measured by Liu et al., we can only speculate about whether such hemodynamic alterations contributed to the high rate of adverse events.

Finally, terlipressin is known to have a longer effective half-life than both norepinephrine and (arginine) vasopressin, but the pharmacokinetics of lysin-vasopressin, the active metabolite of terlipressin, during continuous infusion in septic shock has not been established. Preliminary safety with low-dose infusion (approximately 110 µg/h) was reported in one small trial but adverse events other than atrial fibrillation were not assessed [13]. However, in patients with liver cirrhosis and septic shock who received up to 312 µg/h, the overall rate of adverse events was 41% with 29% experiencing “peripheral cyanosis” [9]. In view of the high rate of serious adverse events, the maximum terlipressin infusion rate should likely be significantly less than 160 µg/h in patients with septic shock.

In all, it appears that the terlipressin doses used to treat patients with hepatorenal syndrome may be too high in patients with septic shock. If there is a place for terlipressin infusion in such patients, a safe dose range needs to be established first. In the meantime, how should clinicians manage vasopressors in septic shock? It makes sense to continue to use norepinephrine as first line therapy. As doses rise, then early use of vasopressin appears to be the logical second line vasopressor. It has been tested in multiple randomised controlled trials and a recent meta-analysis found its use led to lower rates of atrial fibrillation and possibly lower rates of mortality and requirement for renal replacement therapy [14]. However, digital ischemic events were higher in that analysis too, reminding us that adequate fluid resuscitation, repeated assessment of cardiac output and targeting the lowest acceptable MAP for each individual patient to avoid high-dose vasopressors where possible, remain important components in the management of septic shock.