Use of extracorporeal membrane oxygenation in acute pulmonary embolism: a pooled analysis

Background High-risk acute pulmonary embolism (PE) is burdened by a mortality as high as 65%. VenoArterial ExtraCorporeal Membrane Oxygenation (VA-ECMO) may offer a cardiopulmonary support and a precious time window to ensure pulmonary reperfusion therapies. No definite consensus exists on the use of VA-ECMO in high-risk PE patients as only sparse observational studies are available yielding conflicting outcomes. Purpose To provide insights on the use of ECMO in acute high-risk PE pooling together all available published experiences to date. Methods Two authors (LB, AB) searched PubMed, Embase, BioMedCentral and Google Scholar, from inception to 18/09/2019. All published clinical studies investigating ECMO support in patients with high-risk acute PE were evaluated for inclusion. Results Literature search identified 384 observational studies: a total of 66 were included for 584 acute high-risk PE patients receiving ECMO support. Mean age was 46.8±16.8 years (44% female). Most patients presented with cardiac arrest (56%) or obstructive shock (42%). Diagnosis of PE was confirmed by computed tomography (CT) in 72%, performed before ECMO cannulation in 65%. Echocardiography was obtained in 89%. Right ventricle dilatation or dysfunction was found in 90% and 87%. ECMO was primarily employed as upfront treatment (63%), in the VA-ECMO configuration (94%). ECMO was equally employed in conjunction with interventional/surgical pulmonary reperfusion treatments (38%), with thrombolysis (35%) and without adjunctive procedures (40%). Mean ECMO support duration was 100.3±12.9 hours. Notably, 92% received thrombolysis before ECMO cannulation. ECMO bailout implant was mostly adopted in patients receiving thrombolysis (81% vs 19%; p=0.010), as a rescue therapy. Most common reperfusion procedures were surgical embolectomy (28%), catheter-directed thrombolysis (12%) and transcatheter embolectomy (12%). The majority of these procedures (81%) took place after ECMO cannulation. Mean total hospital stay was 17.8±11.6 days. Hospital survival rate was 54% and did not differ in upfront vs bailout ECMO (p=0.184) and between thrombolysis, interventional procedure and ECMO alone recipients (p=0.423). Neurologic death and non-fatal neurologic injury occurred both in 10%. Most patients (70%) were successfully weaned off ECMO, while 30% died on support and 7% died after ECMO removal. Acute kidney injury was the most common complication (47%). Major bleeding occurred in 19% and was fatal for 5%. Patients undergoing thrombolysis had a tendency towards higher rates of major bleeding (48% vs 23%; p=0.05). At a mean follow-up of 365.0 (IQR 202.5–365.9) days, overall survival rate was 85% in those surviving hospitalization. Conclusions In this pooled population consisting mostly of cardiac arrest/obstructive shock PE patients, ECMO strategy was associated with acceptable in-hospital survival and was frequently used in conjunction with other reperfusion treatments. Central Illustration Funding Acknowledgement Type of funding source: None