When should we start vasopressin in septic shock? Revisiting an old question

Introduction

Optimal vasopressin timing in septic shock lacks guidance. Guidelines suggest “moderate” norepinephrine doses (0.25–0.50 µg/(kg·min)), based on consensus.^[1] High norepinephrine doses are associated with risks including arrhythmias and ischemia.^[2] However, adjunctive vasopressin is not without risks; dose-dependent digital ischemia, splanchnic hypoperfusion, and immunomodulatory changes have been reported, underscoring the need to balance benefits and harms when combining vasopressors.^[3,4] Vasopressin exerts its vasoconstrictive, antidiuretic, and endocrine effects through three distinct receptor subtypes: V1receptors on vascular smooth muscle cells mediate potent arteriolar constriction; V2 receptors in the renal collecting ducts promote water reabsorption independent of sodium; and V3 receptors in the hypothalamic–pituitary axis regulate adrenocorticotropic hormone (ACTH) release and modulate inflammatory signaling.^[3]

A Look Back: VASST, VANISH, and Beyond

The Vasopressin and Septic Shock Trial (VASST) did not show an overall mortality benefit of vasopressin compared with norepinephrine, although a less-severe subgroup suggested possible benefit.^[5] The Vasopressin vs Norepinephrine as Initial therapy in Septic Shock (VANISH) trial enrolled patients within 6 h of vasopressor initiation (median norepinephrine dose 0.08 µg/(kg·min)),^[5] selecting a cohort with relatively preserved renal perfusion.^[6] Although overall mortality did not differ between groups, vasopressin recipients experienced a non-significant reduction in stage ≥2 acute kidney injury (22% vs. 28%; P=0.07) and a lower requirement for renal replacement therapy among those with baseline renal impairment, highlighting a potentially meaningful, though statistically borderline, renal benefit in this specific population. VANISH did not confirm the VASST subgroup signal; in fact, outcomes trended toward a slight advantage with norepinephrine in this early, low-dose cohort.^[5,6]

Defining Vasopressin Timing

Defining vasopressin’s optimal window will require clearer markers – whether fixed thresholds, clinical criteria, or physiologic indicators. The optimal moment to introduce vasopressin in septic shock has been characterized along three axes rather than by clock time alone. First, the onset-to-adjunction interval measures hours from clinical diagnosis of refractory hypotension to vasopressin start, with trials like VANISH (<6 h) and analyses by Nagendran et al.^[7] (<12 h vs. >12 h), suggesting that very early initiation may confer renal and hemodynamic advantages without a clear mortality benefit. Second, dose-based thresholds define timing according to the intensity of norepinephrine support. Several studies have reported greater vasopressor-sparing effects when vasopressin is added at lower norepinephrine doses (≤0.1–0.15 µg/(kg·min)), as observed in the VASST subgroup analysis^[5] and supported by target trial emulation data from White et al.^[8] Third, duration-based thresholds consider hours of norepinephrine exposure before escalation; observational studies.^[9] indicate that adding vasopressin within 4–8 h of NE initiation can reduce subsequent catecholamine requirements and possibly renal support needs. By framing timing in these three dimensions, clinicians can tailor vasopressin deployment to individual shock trajectories rather than a single temporal cut-off. Observational data indicate that higher catecholamine dose, elevated lactate, and prolonged shock duration at vasopressin initiation are independently associated with increased mortality.^[10] In the PRESS Survey, the most common trigger for vasopressin initiation was norepinephrine dose alone (98% of respondents; most frequent 0.25–0.50 µg/(kg·min), followed by duration alone (53%; most frequent > 2–6 h) or a combination of both (78%). The most common initial vasopressin dose was 0.01–0.03 U/min (50%).^[11]

From Timing to Thresholds

“Moderate dose norepinephrine” varies across intensive care units (ICU). Early trials reported norepinephrine as tartrate salt (1 mg tartrate ≈ 0.5 mg base); modern studies use base. Historical doses have been converted to base-equivalent (tartrate × 0.5) for consistency, and norepinephrine-equivalent doses recalculated using the standard (vasopressin 0.03 U/min = 0.1 µg/(kg·min) norepinephrine), as described in Kotani et al.^[12]

Vasopressin Meta-analysis Reappraised

Nagendran et al.^[7] performed an individual patient data meta-analysis and observed a significant interaction between lower vasopressor doses at vasopressin initiation and improved 28-day survival. However, the 90-day outcomes – available for only half of the cohort – did not corroborate this finding, and the overall effect estimate in the full dataset did not demonstrate a consistent survival advantage. No interaction was found based on timing (<12 h vs. >12 h), and while subgroup analyses suggested a potential benefit in patients with normal lactate levels, these signals require cautious interpretation given the selective patient population at 90 days. In addition, there was a non-significant trend toward a higher incidence of adverse events (e.g., digital ischemia, hyponatremia) in patients who received vasopressin within 12 h of norepinephrine initiation. Moreover, there was no significant interaction between either timing or dose thresholds and subsequent renal replacement therapy requirements in the full cohort.^[7]

Emerging Approaches: Artificial Intelligence (AI) and Vasopressor Stewardship

AI models using ICU data suggest earlier vasopressin, personalizing strategies. The OVISS study (14,000 patients) recommended vasopressin at a dose of approximately 0.20 µg/(kg·min), reducing mortality.^[13] An emulation trial using real-world ICU data compared vasopressin initiation at norepinephrine doses ≤0.15 µg/(kg·min) vs. >0.15 µg /(kg·min).^[8] This study found that starting vasopressin at the lower threshold significantly reduced cumulative catecholamine requirements and was associated with improved markers of renal perfusion.

Clinical Caution and Future Directions

Despite growing interest, early initiation of vasopressin is not yet standard practice. The lack of randomized data comparing timing strategies remains a key barrier. In addition, concerns persist about vasopressin’s potential adverse effects – such as digital ischemia, splanchnic hypoperfusion, and rebound hypotension – though these are rare and appear dose-dependent.^[3,4,13] Still, the accumulating evidence supports a shift in perspective: from vasopressin as a “rescue drug” to an adjunctive therapy with a preventive role. Reframing the question from “Should we use vasopressin?” to “When should we start?” opens new possibilities for trial design, protocol development, and personalized care. Upcoming studies should aim to test fixed-dose thresholds prospectively, evaluate physiologic markers (e.g., lactate, vasopressor responsiveness), and consider composite outcomes beyond mortality – including kidney function, time on vasopressors, and ICU-free days.^[7,13]

Conclusions

The question of vasopressin timing is not new; however, the tools to answer it are more sophisticated than ever. Although definitive guidance is still pending, a growing body of evidence – including physiologic rationale, subgroup analyses, and data-driven modeling – suggests that earlier initiation may confer real benefit. For now, clinicians should remain alert not only to how much norepinephrine their patients are receiving, but also when the time might be right for vasopressin.

CRediT authorship contribution statement

Javier Muñoz: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Investigation, Formal analysis, Conceptualization. Jamil Antonio Cedeño: Writing – review & editing, Validation, Investigation.

Managing Editor: Jingling Bao/Zhiyu Wang