Timing and Strategy for Weaning From Venoarterial ECMO are Complex Issues

Abstract

Objective: Weaning from venoarterial extracorporeal membrane oxygenation (VA ECMO) usually is performed without clear guidelines; yet, patients still die after removal of extracorporeal circulation because of inadequate heart or end-organ recovery. The aim of the study was to address the weaning procedure, analyzing the hemodynamic and echocardiographic picture of patients weaned and to identify predictors of poor outcome among this population.

Design: Observational study.

Setting: University hospital.

Participants: One hundred twenty-nine VA ECMO cases.

Interventions: None.

Measurements and main results: Forty-nine patients (38%) were weaned, 7 (5.4%) were bridged to a ventricular assist device, and 6 (5.2%) were listed for heart transplantation. Weaned patients showed a significant increase of pulse pressure (35 [0-50] mmHg before ECMO, 59 [53-67] mmHg at weaning, 61 [51-76] mmHg after ECMO (p<0.001]) and reduction of dose of inotropes (inotropic score [as defined in the text] 20 [14-40] before ECMO, 10 [3-15] at weaning, and 10 [5-15] after ECMO, p<0.001). Left ventricular ejection fraction (LVEF) increased from 19 (0-22.5)% before ECMO to 35 (22-55)% after ECMO (p<0.001). A significant improvement of right ventricular (RV) function was observed in weaned patients (RV dysfunction from 52% to 21%, p<0.001). Among weaned patients, 15 (31%) died. Patients who died after weaning had longer ECMO duration compared to discharged patients (8 [5-11] v 4 [2-6] days, p = 0.01) and more transfusions (22 [10-37] v 7 [0.5-15] units, p = 0.02); survival was lower in patients with central ECMO (postcardiotomy) compared to peripheral ECMO (p = 0.045). Mortality was higher in those with persistence of RV failure, continuous venovenous hemofiltration, higher inotropic score, lower systolic pressure, or higher leucocyte count at weaning.

Conclusions: Successful weaning from ECMO is a multifaceted process, which encompasses consistent recovery of myocardial and end-organ function; LVEF, though improved, is still low at weaning. Hospital survival is correlated significantly to the duration of ECMO support and to bleeding complications.

Keywords: echocardiography; hemodynamics; venoarterial ECMO; weaning.