Early temporary mechanical circulatory support for cardiogenic shock: an emulated target trial analysis based on a prospective cohort

Abstract

Background: Encouraging results in several series of patients on refractory cardiogenic shock (CS) led to the consideration of including temporary mechanical circulatory support (TMCS) in the guidelines for CS management. Benefit on survival might be improved when TMCS devices are initiated early. This study assessed early TMCS for CS in the real-world condition of a cardiac assistance regional network. Methods: We examined the clinical features and outcomes of patients treated with TMCS, using data from a prospective cohort of patients admitted for CS in a cardiac assistance regional network. Patients were eligible when CS included systolic blood pressure (SBP) <90 mmHg despite adequate filling pressure or need of catecholamine to maintain SPB>90 mmHg, signs of congestion or tissue hypo-perfusion or blood lactate >2mmol/L. TMCS indication, veno-arterial extracorporeal membrane oxygenation (VA-ECMO) or percutaneous left ventricle assistance (Impella) was based on the initial severity of CS or if shock was refractory to medical treatment. To estimate the effect of early TMCS on mortality, we emulated a target trial of TMCS implantation versus no TMCS within 24 hours of ICU admission. Patients were followed until hospital discharge, death, or 180 days. Inverse probability of treatment weighting with propensity score was used to adjust for confounders. Results: Among 372 patients, 246 were included in the study: 121 treated by early TMCS (TMCS group) and 125 not treated by early TMCS (controls). Median age was 59.5 years, 71.9% were male, and the main etiologies were acute myocardial infraction (MI) (46.8%) and decompensated heart failure (27.2%). Crude in-hospital mortality was 38% (46/121) in the TMCS group and 22.4% (28/125) in the control group. After emulation, in-hospital mortality was not different between the TMCS group and the control group (HR=0.91, 95%CI:0.65-1.26). In sub-group analyses, in-hospital mortality was not different among patients with MI (HR=0.82, 95%CI:0.52-1.32), with other etiology than MI (HR=1.19, 95%CI:0.70-2.04), and with VA-ECMO only (HR=0.87, 95%CI:0.62-1.22). TMCS patients had two-fold longer hospital stays, and a higher rate of complications than controls. Conclusion: In this emulated target trial from real-world data, patients with CS treated with early TMCS had no improvement in survival compared to controls. Trial registration number: NCT03528291, registered trials May 18, 2017