Résumé
Introduction: Mixed cardiogenic-vasodilatory shock carries high mortality. Vasopressin and veno-arterial extracorporeal membrane oxygenation (VA-ECMO) are used in refractory cases, but evidence on the optimal timing of vasopressin initiation remains limited.Methods: This bicentric retrospective cohort study included 480 patients with mixed shock treated with vasopressin (n = 191) and VA-ECMO. Patients were categorized into no vasopressin (n = 289), early (<6 hours after VA-ECMO initiation; n = 98), and late (≥6 hours; n = 93) vasopressin groups. The primary outcome was new-onset or worsening AKI within 30 days. Analyses included multivariate logistic regression and propensity score-based inverse probability of treatment weighting (IPTW).Results: Early vasopressin recipients had higher baseline norepinephrine-equivalent requirements than late vasopressin group and no vasopressin group (1.12 µg kg-1 min-1 [0.41;2.29] vs. 0.55 µg kg-1 min-1 [0.17;1.20], vs. 0.33 µg kg-1 min-1 [0.12;0.90], p < 0.001). In the overall cohort, early vasopressin use was associated with lower AKI incidence (49 patients (50.0%) vs. 216 patients (74.7%) %), and 49 patients (50.0%) vs. 73 patients (78.5%), p = 0.001), compared to no vasopressin use. The association was confirmed with unadjusted (OR: 0.35 [CI₉₅%: 0.21-0.59], p = 0.001 vs. no vasopressin use), and adjusted multivariate logistic regression (OR: 0.42 [CI₉₅%: 0.231-0.771], p = 0.005 vs. no vasopressin use). The rate of digestive ischemia and mortality did not differ between the three groups of patients.Conclusions: Early vasopressin administration in VA-ECMO-supported mixed shock may reduce the incidence of acute kidney injury compared with no vasopressin use, without increasing adverse events. Prospective studies are warranted to confirm these findings and to determine the optimal timing of vasopressin therapy in this high-risk population.