Extracorporeal Membrane Oxygenation and Reperfusion Strategies in High-Risk Pulmonary Embolism Hospitalizations

Ioannis T Farmakis,Ingo Sagoschen,Stefano Barco,Karsten Keller,Luca Valerio,Johannes Wild,George Giannakoulas,Gregory Piazza,Stavros V Konstantinides,Lukas Hobohm
DOI: https://doi.org/10.1097/CCM.0000000000006361
2024-10-01
Abstract:Objectives: To investigate the contemporary use of extracorporeal membrane oxygenation (ECMO) in conjunction with reperfusion strategies in high-risk pulmonary embolism (PE). Design: Observational epidemiological analysis. Setting: The U.S. Nationwide Inpatient Sample (NIS) (years 2016-2020). Patients: High-risk PE hospitalizations. Measurements and main results: Use of ECMO in conjunction with thrombolysis-based reperfusion (systemic thrombolysis or catheter-directed thrombolysis) or mechanical reperfusion (surgical embolectomy or catheter-based thrombectomy) with regards to in-hospital mortality and major bleeding. We identified high-risk PE hospitalizations in the NIS (years 2016-2020) and investigated the use of ECMO in conjunction with thrombolysis-based (systemic thrombolysis or catheter-directed thrombolysis) and mechanical (surgical embolectomy or catheter-based thrombectomy) reperfusion strategies with regards to in-hospital mortality and major bleeding. Among 122,735 hospitalizations for high-risk PE, ECMO was used in 2,805 (2.3%); stand-alone in 1.4%, thrombolysis-based reperfusion in 0.4%, and mechanical reperfusion in 0.5%. Compared with neither reperfusion nor ECMO, ECMO plus thrombolysis-based reperfusion was associated with reduced in-hospital mortality (adjusted odds ratio [aOR] 0.61; 95% CI, 0.38-0.98), whereas no difference was found with ECMO plus mechanical reperfusion (aOR 1.03; 95% CI, 0.67-1.60), and ECMO stand-alone was associated with increased in-hospital mortality (aOR 1.60; 95% CI, 1.22-2.10). In the cardiac arrest subgroup, ECMO was associated with reduced in-hospital mortality (aOR 0.71; 95% CI, 0.53-0.93). Among all patients on ECMO, thrombolysis-based reperfusion was significantly associated (aOR 0.55; 95% CI, 0.33-0.91), and mechanical reperfusion showed a trend (aOR 0.75; 95% CI, 0.47-1.19) toward reduced in-hospital mortality compared with no reperfusion, without increases in major bleeding. Conclusions: In patients with high-risk PE and refractory hemodynamic instability, ECMO may be a valuable supportive treatment in conjunction with reperfusion treatment but not as a stand-alone treatment especially for patients suffering from cardiac arrest.
What problem does this paper attempt to address?