Neonatal cavopulmonary assist: pulsatile versus steady-flow pulmonary perfusion.

BACKGROUND: Morbidity and mortality associated with single-ventricle physiology decrease substantially once a systemic venous, rather than systemic arterial, source of pulmonary blood flow is established. Cavopulmonary assist has potential to eliminate critical dependence on the problematic systemic-to-pulmonary shunt as a source of pulmonary blood flow in neonates. We have previously demonstrated feasibility of neonatal cavopulmonary assist under steady-flow conditions. We hypothesized that pulsatile pulmonary perfusion would further improve pulmonary hemodynamics.

METHODS: Lambs (weight 7.2 +/- 1.1 kg, age 7.9 +/- 1.5 days) underwent total cavopulmonary diversion using bicaval venous-to-main pulmonary artery cannulation. A miniature centrifugal pump was used to augment cavopulmonary flow. Pulsatility was created with an intermittently compressed compliance chamber in the circuit. Hemodynamic and gas exchange data were measured for 8 hours. Pulsatile (n = 6), steady-flow (n = 13), and control (n = 6) groups were compared using two-way analysis of variance with repeated measures.

RESULTS: All animals remained physiologically stable with normal gas exchange function. Mean pulmonary arterial pressure was elevated in pulsatile and steady-flow groups compared with the control group and within-group baseline values. Pulmonary vascular resistance was elevated initially in both assist groups but decreased significantly over the last 4 hours of the study and normalized after hour 4 in the pulsatile perfusion group. Pulmonary vascular resistance also normalized to control in the steady-flow group after hour 7.

CONCLUSIONS: Both steady-flow and pulsatile pulmonary perfusion demonstrated normalization of pulmonary vascular resistance to control in a neonatal model of univentricular Fontan circulation. These results suggest that there is no benefit to pulsatile flow in this model.