Subclavian or axillary artery cannulation for extracorporeal membrane oxygenation: A systematic review

Subclavian or axillary artery cannulation for venoarterial extracorporeal membrane oxygenation (VA-ECMO) provides a valuable alternative to femoral access; however, a comprehensive overview of such an approach in this setting is lacking. This review examined types of access, clinical complications, and outcomes of subclavian/axillary cannulation, emphasizing the pros and cons of this VA-ECMO approach as well as areas for further investigation.

A systematic search was conducted in PubMed, Scopus, and Web of Science for articles published between January 2000 and December 2024. Studies were included that (1) reported on axillary or subclavian artery cannulation in VA-ECMO, (2) involved adult patients, and (3) provided in-hospital outcomes. Exclusion criteria were case reports with fewer than 5 patients and non-English language articles. Data on patient demographics, indications, cannulation techniques, complications, and in-hospital survival were collected.

Seventeen studies with a total of 2030 patients were selected and analyzed. Subclavian or axillary cannulation was predominantly right-sided (99%) and involved graft interposition in 81% of cases. Indications for VA-ECMO with this arterial access included acute myocardial infarction shock (11%-42%), acute decompensation with chronic heart failure (4.9%-52%), postcardiotomy shock (7%-75.2%), acute myocarditis (4%- 7.7%) and refractory respiratory distress (5%-50%). Complications included limb ischemia (1.2%-9.6%), stroke (5.2%-18.8%) and bleeding events (4%-37.5%, often at the graft site). Left ventricular (LV) unloading strategies were used in 5.2% to 15.0% of cases but were not documented in almost 60% of the studies. In-hospital and 1-year survival rates ranged from 82.7% to 55% and from 50.6% to 25.7%, respectively.

VA-ECMO with axillary or subclavian artery access can be a safe and effective configuration, particularly for patients with contraindications or need of change of femoral cannulation. However, conflicting data on stroke risk, hemodynamic effects on LV ejection, the need for LV unloading (particularly with the right-sided approach), and both short- and long-term survival rates warrant further investigation.