Routine mechanical ventricular assist following the Norwood procedure - Improved neurologic outcome and excellent hospital survival

Background. Although excellent survival following the Norwood procedure for palliation of hypoplastic left heart syndrome (HLHS) is being achieved by some, most centers, especially the ones with small surgical volume and limited experience, continue to struggle with initial results. Survivors often showed evidence of significant neurologic injury. The early postoperative care is labor-intensive as attempts are made to balance the systemic and pulmonary circulation for these infants. We report our experience with routine use of mechanical circulatory assist to support the increased cardiac output requirements present following Norwood procedure. Methods. Eighteen consecutive infants undergoing Norwood operation for HLHS (Oregon Health & Science University [OHSU] 13; University of Louisville [UL] 5) were placed on a ventricular assist device (VAD) immediately following modified ultrafiltration in the operating room using the cardiopulmonary bypass (CPB) cannulas that were in the right atrium and the neoaorta. VAD flows were maintained at approximately 200 mL . kg(-1) . min(-1) and the patients were transported to the intensive care unit (ICU). Patients operated at OHSU also received neurodevelopmental testing before their Glenn procedure, approximately 4 to 6 months following their Norwood operation. Results. All patients were stable on VAD support and no attempt was made to balance the systemic and pulmonary circulation. The ventilator was manipulated to achieve systemic PaO2 between 30 and 45 mm Hg and PaCO2 between 35 and 45 mm Hg. Evidence of hypoperfusion (increasing lactates) was managed by increasing the VAD flow. Lactates normalized [< 2 mmol/L]) by 1.8 +/- 1.1 days following surgery. Average time of VAD support was 3.1 +/- 1.0 (range, 2 to 5 days) and average time until chest closure was 3.4 +/- 1.5 (range, 2 to 8 days). There were two cases of postoperative bleeding (11.1%) requiring reexploration and one case of mediastinitis (5.5%) in a patient who has now gone on to successful Glenn. Sixteen of the eighteen patients survived (hospital survival mean 89% with a 95% confidence interval of 63.9% to 98.1%; 12/13 OHSU [92.3%]; 4/5 UL [80%]). Neurodevelopmental testing using the Mullen Scales of Early Learning and the Vineland Adaptive Behavior Scale were normal for all infants tested. Conclusions. Routine postoperative use of VAD can support the increased cardiac output demands of infants following Norwood operation and results in a stable postoperative convalescence that does not require aggressive ventilator or inotrope manipulation. Although not a panacea, this strategy can simplify postoperative management, lead to excellent hospital survival, and possibly augment cerebral oxygen delivery, resulting in improved neurologic outcomes for this challenging group of patients.