Extracorporeal membrane oxygenation use in poisoning: a narrative review with clinical recommendations

Cameron Upchurch, Adam Blumenberg, Daniel Brodie, Graeme MacLaren, Bishoy Zakhary, Robert G. Hendrickson

Research output: Contribution to journalReview articlepeer-review

15 Scopus citations

Abstract

Context: Poisoning may lead to respiratory failure, shock, cardiac arrest, or death. Extracorporeal membrane oxygenation (ECMO) may be used to provide circulatory support, termed venoarterial (VA) ECMO; or respiratory support termed venovenous (VV) ECMO. The clinical utility of ECMO in poisoned patients remains unclear and guidelines on its use in this setting are lacking. Objectives: To perform a literature search and narrative review on the use of ECMO in poisonings. Additionally, to provide recommendations on the use of ECMO in poisonings from physicians with expertise in ECMO, medical toxicology, critical care, and emergency medicine. Methods: A literature search in Ovid MEDLINE from 1946 to October 14, 2020, was performed to identify relevant articles with a strategy utilizing both MeSH terms and adjacency searching that encompassed both extracorporeal life support/ECMO/Membrane Oxygenation concepts and chemically-induced disorders/toxicity/poisoning concepts, which identified 318 unique records. Twelve additional manuscripts were identified by the authors for a total of 330 articles for screening, of which 156 were included for this report. Narrative literature review: The use of ECMO in poisoned patients is significantly increasing over time. Available retrospective data suggest that patients receiving VA ECMO for refractory shock or cardiac arrest due to poisoning have lower mortality as compared to those who receive VA ECMO for non-poisoning-related indications. Poisoned patients treated with ECMO have reduced mortality as compared to those treated without ECMO with similar severity of illness and after adjusted analyses, regardless of the type of ingestion. This is especially evident for poisoned patients with refractory cardiac arrest placed on VA ECMO (termed extracorporeal cardiopulmonary resuscitation [ECPR]). Indications: We suggest VA ECMO be considered for poisoned patients with refractory cardiogenic shock (continued shock with myocardial dysfunction despite fluid resuscitation, vasoactive support, and indicated toxicologic therapies such as glucagon, intravenous lipid emulsion, hyperinsulinemia euglycemia therapy, or others), and strongly considered for patients with cardiac arrest in institutions which are structured to deliver effective ECPR. VV ECMO should be considered in poisoned patients with ARDS or severe respiratory failure according to traditional indications for ECMO in this setting. Contraindications: Patients with pre-existing comorbidities with low expected survival or recovery. Relative contraindications vary based on each center’s experience but often include: severe brain injury; advanced age; unrepaired aortic dissection or severe aortic regurgitation in VA ECMO; irreversible organ injury; contraindication to systemic anticoagulation, such as severe hemorrhage. Conclusions: ECMO may provide hemodynamic or respiratory support to poisoned patients while they recover from the toxic exposure and metabolize or eliminate the toxic agent. Available literature suggests a potential benefit for ECMO use in selected poisoned patients with refractory shock, cardiac arrest, or respiratory failure. Future studies may help to further our understanding of the use and complications of ECMO in poisoned patients.

Original languageEnglish (US)
Pages (from-to)877-887
Number of pages11
JournalClinical Toxicology
Volume59
Issue number10
DOIs
StatePublished - 2021

Keywords

  • ECLS
  • ECMO
  • ECPR
  • cardiac arrest
  • overdose
  • poisoning
  • respiratory failure
  • shock
  • toxin

ASJC Scopus subject areas

  • Toxicology

Fingerprint

Dive into the research topics of 'Extracorporeal membrane oxygenation use in poisoning: a narrative review with clinical recommendations'. Together they form a unique fingerprint.

Cite this