Venoarterial extracorporeal membrane oxygenation to rescue sepsis-induced cardiogenic shock: a retrospective, multicentre, international cohort study

Abstract

Background: Patients with sepsis-induced cardiomyopathy with cardiogenic shock have a high mortality. This study assessed venoarterial extracorporeal membrane oxygenation (VA-ECMO) support for sepsis-induced cardiogenic shock refractory to conventional treatments.

Methods: In this retrospective, multicentre, international cohort study, we compared outcomes of 82 patients (aged ≥18 years) with septic shock who received VA-ECMO at five academic ECMO centres, with 130 controls (not receiving ECMO) obtained from three large databases of septic shock. All patients had severe myocardial dysfunction (cardiac index 3 L/min per m² or less or left ventricular ejection fraction [LVEF] 35% or less) and severe haemodynamic compromise (inotrope score at least 75 μg/kg per min or lactic acidaemia at least 4 mmol/L) at time of inclusion. The primary endpoint was survival at 90 days. A propensity score-weighted analysis was done to control for confounders.

Findings: At baseline, patients treated with VA-ECMO had more severe myocardial dysfunction (mean cardiac index 1·5 L/min per m²vs 2·2 L/min per m², LVEF 17% vs 27%), more severe haemodynamic impairment (inotrope score 279 μg/kg per min vs 145 μg/kg per min, lactataemia 8·9 mmol/L vs 6·5 mmol/L), and more severe organ failure (Sequential Organ Failure Assessment score 17 vs 13) than did controls, with p<0·0001 for each comparison. Survival at 90 days for patients treated with VA-ECMO was significantly higher than for controls (60% vs 25%, risk ratio [RR] for mortality 0·54, 95% CI [0·40-0·70]; p<0·0001). After propensity score weighting, ECMO remained associated with improved survival (51% vs 14%, adjusted RR for mortality 0·57, 95% CI [0·35-0·93]; p=0·0029). Lactate and catecholamine clearance were also significantly enhanced in patients treated with ECMO. Among the 49 survivors treated with ECMO, 32 who had been treated at the largest centre reported satisfactory Short Form-36 evaluated health-related quality of life at 1-year follow-up.

Interpretation: Patients with severe sepsis-induced cardiogenic shock treated with VA-ECMO had a large and significant improvement in survival compared with controls not receiving ECMO. However, despite the careful propensity-weighted analysis, we cannot rule out unmeasured confounders.

Funding: None.