Feasibility of airway pressure release ventilation in infants with near end stage- neonatal bronchopulmonary dysplasia

Background: Ventilated premature infants with severe, advanced bronchopulmonary dysplasia (BPD) often require high mean airway pressure (MAP) and FiO $_{2}$ to maintain acceptable gas exchange using conventional mechanical ventilation techniques such as synchronized intermittent mandatory ventilation (SIMV). In addition, they are at risk to develop episodes of severe respiratory failure ("BPD spells"). We report our experience with airway pressure release ventilation (APRV), a ventilation modality which allows spontaneous respiration during a relatively long high pressure phase followed by a shorter release period allowing for CO $_{2}$ removal from the conducting airways. Despite the higher MAP, the unique APRV flow pattern is thought to decrease volutrauma.

Objective: To determine the feasibility and safety of APRV in neonates with severe BPD.

Design/methods: Approval for this retrospective review was obtained from Childrens Hospital Los Angeles (CHLA) Institutional Review Board. Patients with advanced BPD treated in the Newborn and Infant Critical Care Unit at CHLA that received APRV at the discretion of the attending neonatologist for developing severe respiratory failure between May 2004 and June 2005 were identified and their medical records reviewed.

Results: Nine patients were placed on APRV using the AVEA ventilator (CareFusion, San Diego, CA). Eight infants were former extremely premature infants with severe BPD and 1 infant was born at term infant with suspected surfactant protein-B deficiency. Mean gestational and postmenstrual age was 26.7 ± 4.9 and 50.5 ± 9.8 weeks, respectively at the time of initiation of APRV. The duration of APRV was 17.5 ± 11.9 days. On APRV, patients received higher MAP (29 vs. 15 cm H $_{2}$ O; p< 0.001) and lower FiO $_{2}$ (0.5 vs. 0.3; p = 0.38) compared to SIMV with pressure support, high-frequency oscillatory ventilation or high-frequency jet ventilation. Seven of the 9 patients had clinical improvement on APRV within 3 days defined by fewer episodes of severe desaturation, variability in FiO $_{2}$ , and vasopressor/inotrope support. In addition, none of the infants received pharmacological neuromuscular blockade after the initiation of APRV. Despite the higher MAP, patients did not develop air-leaks and remained hemodynamically stable with normal blood pressure and urine output.

Conclusions: These preliminary data indicate that infants with severe BPD tolerate the use of APRV without apparent adverse effects. However, it remains unclear whether APRV is indeed safe and beneficial in former very preterm neonates with severe BDP and episodes of respiratory failure. Randomized controlled studies are required to determine the safety, efficacy and long-term morbidities of this mode of ventilation in neonates.