Extracorporeal Membrane Oxygenation for Pulmonary Embolism: A Systematic Review and Meta-Analysis

BACKGROUND: The use of extracorporeal membrane oxygenation (ECMO) for high-risk pulmonary embolism (HRPE) with haemodynamic instability or profound cardiogenic shock has been reported. Guidelines currently support the use of ECMO only in patients with cardiac arrest or circulatory collapse and in conjunction with other curative therapies. We aimed to characterise the mortality of adults with HRPE treated with ECMO, identify factors associated with mortality, and compare different adjunct curative therapies. METHODS: We conducted a systematic review and meta-analysis, searching four international databases from their inception until 25 June 2023 for studies reporting on more than five patients receiving ECMO for HRPE. Random-effects meta-analyses were conducted. The primary outcome was in-hospital mortality. A subgroup analysis investigating the outcomes with curative treatment for HRPE was also performed. The intra-study risk of bias and the certainty of evidence were also assessed. This study was registered with PROSPERO (CRD42022297518). RESULTS: A total of 39 observational studies involving 6409 patients receiving ECMO for HRPE were included in the meta-analysis. The pooled mortality was 42.8% (95% confidence interval [CI]: 37.2% to 48.7%, moderate certainty). Patients treated with ECMO and catheter-directed therapy (28.6%) had significantly lower mortality ( < 0.0001) compared to those treated with ECMO and systemic thrombolysis (57.0%). Cardiac arrest prior to ECMO initiation (regression coefficient [B]: 1.77, 95%-CI: 0.29 to 3.25, = 0.018) and pre-ECMO heart rate (B: -0.076, 95%-CI: -0.12 to 0.035, = 0.0003) were significantly associated with mortality. The pooled risk ratio when comparing mortality between patients on ECMO and those not on ECMO was 1.51 (95%-CI: 1.07 to 2.14, < 0.01) in favour of ECMO. The pooled mortality was 55.2% (95%-CI: 47.7% to 62.6%), using trim-and-fill analysis to account for the significant publication bias. CONCLUSIONS: More than 50% of patients receiving ECMO for HRPE survive. While outcomes may vary based on the curative therapy used, early ECMO should be considered as a stabilising measure when treating patients with HRPE. Patients treated concurrently with systemic thrombolysis have higher mortality than those receiving ECMO alone or with other curative therapies, particularly catheter-directed therapies. Further studies are required to explore ECMO vs. non-ECMO therapies in view of currently heterogenous datasets.