Venovenous extracorporeal membrane oxygenation for patients with single-ventricle anatomy: A registry report.

OBJECTIVE: Support with extracorporeal membrane oxygenation for cardiopulmonary failure is done so with venoarterial cannulation in the majority of children with single-ventricle anatomy. However, there is a growing experience for patients with pure oxygenation/ventilation impairment supported with venovenous extracorporeal membrane oxygenation. We describe that experience.

METHODS: Data were collected from the Extracorporeal Life Support Organization registry for patients with single-ventricle anatomy supported with venovenous extracorporeal membrane oxygenation from 1990 to 2012. Descriptive statistics and multivariate analyses for associations with mortality were conducted.

RESULTS: A total of 89 patients with single-ventricle anatomy had venovenous extracorporeal membrane oxygenation performed at a median age of 66 days (8-221). Survival to discharge was 48%. Fifty-four patients (61%) had shunt physiology, 22 patients (25%) had cavopulmonary connections, and 13 patients (14%) had single-ventricle anatomy but with no previous cardiac surgery. Indication for extracorporeal membrane oxygenation was respiratory failure in 59 patients (63%) and cardiac failure in 30 patients (32%). Double-lumen cannulas were used in 62 patients (70%). Bivariate analysis demonstrated that the duration of intubation before extracorporeal membrane oxygenation, mean airway pressure before cannulation, partial pressure carbon dioxide before cannulation, peak inspiratory pressure before cannulation, pump flow at 24 hours, extracorporeal membrane oxygenation run duration, and presence of renal injury were associated with mortality. Multivariate logistic analysis demonstrated that the duration of intubation (adjusted odds ratio, 1.01; 95% confidence interval, 1.003-1.016; P = .003), partial pressure carbon dioxide (adjusted odds ratio, 1.04; 95% confidence interval, 1.01-1.068; P = .007), mean airway pressure (adjusted odds ratio, 1.16; 95% confidence interval, 1.0-1.342; P = .05), and renal injury (adjusted odds ratio, 6.6; 95% confidence interval, 1.879-23.2; P = .003) were associated with mortality.

CONCLUSIONS: Patients with single-ventricle anatomy in respiratory failure may be treated successfully with venovenous extracorporeal membrane oxygenation, with survival comparable to those treated with venoarterial extracorporeal membrane oxygenation for cardiac failure. Future research on indications for venovenous extracorporeal membrane oxygenation may aid clinicians in deciding the optimal approach for this challenging cohort.