Extracorporeal membrane oxygenation induced cardiac dysfunction in newborn lambs.

Extracorporeal membrane oxygenation (ECMO) is routinely used to support cardiopulmonary failure in infants and children. Suboptimal outcomes for primary cardiac support suggest a need for investigation of the impact of ECMO on the heart. Twenty-four newborn lambs received a brief period of ECMO support to investigate the hypothesis that ECMO produces cardiac dysfunction in newborn lamb. Dorset newborn lambs, 4-7 days of age, were exposed to ECMO for 5 min at a 100 ml/kg flow rate and quickly weaned off. Measurements included echocardiographic mean left ventricular (LV) velocity of circumferential fiber shortening corrected for heart rate (mVCFc), LV shortening fraction, and peak systolic wall stress plus hemodynamic measurement of LV maximum rate of pressure change with time (LV dp/dt max), maximum rate of pressure change divided by developed pressure (LV dp/dtP), right atrial pressure, pulmonary capillary wedge pressure, mean pulmonary artery pressure, LV peak and end-diastolic pressure, and aortic pressure. These measures were also obtained after an exposure to 5 min of ECMO and immediate disconnect for 5 min, followed by ECMO administration for 1 h again, followed by discontinuation of ECMO. LV mVCFc is decreased after exposure to 5 min of ECMO support despite a decrease in LV peak systolic wall stress that provides afterload reduction. LV mVCFc is inversely related to peak systolic wall stress at a significance level of p < 0.0001. The time period after initiation of ECMO is a significant factor in the model (p = 0.0097). Time [baseline] was different from the other time points with p = 0.0010. Average mVCFc at baseline is 1.27 ± 0.35 and decreases to 1.01 ± 0.42 after 5 min of ECMO that is then withdrawn. Peak systolic wall stress decreases from 36.0 ± 13.1 at baseline to 29.8 ± 12.1 after 5 min of ECMO. LV dp/dt max decreases from 1,769 ± 453 mmHg/s at baseline to 1,311 ± 513 mmHg/s after exposure to 5 min of ECMO (p = 0.0005). Baseline LV dp/dt max is different from each point after start of ECMO. Diastolic LVdp/dt min increased from -1,340 ± 477 mmHg/s to -908 ± 393 mmHg/s at 5 min. Echocardiographic mVCFc, when considered in isolation or as a function of LV peak systolic wall stress, shows diminished LV function after ECMO. Hemodynamic measurement of LV dp/dt max and LV dp/dt min confirms the observation. Separation of the humoral from mechanical effect of ECMO with the short exposure to the extracorporeal circuit shows that an immediate decrement of LV function occurs at initiation of ECMO, a finding that has not been stressed with previous studies of extracorporeal support. This implies a potentially outcome-limiting deleterious effect for the patient who requires ECMO support for the heart rather than the lungs. We should continue to strive to understand and ameliorate this deleterious effect of the extracorporeal circulation circuit.