Extracorporeal membrane oxygenation in infants with meconium aspiration syndrome: a decade of experience with venovenous ECMO

Abstract

Background: Despite the emergence of new therapies for respiratory failure of the newborn with meconium aspiration syndrome (MAS), extracorporeal membrane oxygenation (ECMO) has a significant role as a rescue modality in these infants. Our objective was to compare the use of venovenous (VV) vs venoarterial (VA) ECMO in newborns with MAS who need ECMO and to ascertain the impact of new therapies in these infants during the last decade. We also evaluated how disease severity or time of ECMO initiation affected mortality and morbidity.

Methods: A report of 12 years experience (1990-2002) of a single center, comparing VV and VA ECMO, is given. Venovenous ECMO was the preferred rescue modality for respiratory failure unresponsive to maximal medical therapy. Venoarterial ECMO was used only when the placement of a VV ECMO 14-F catheter was not possible; 128 patients met ECMO criteria, 114 were treated with VV ECMO, and 12 with VA ECMO. Two patients were converted from VV to VA ECMO.

Results: Venovenous and VA ECMO patients had comparable birth weight (mean +/- SEM, 3.48 +/- 0.05 vs 3.35 +/- 0.15 kg) and gestational age (40.3 +/- 0.1 vs 40.7 +/- 0.3 weeks). Before ECMO, there was no difference between VV and VA ECMO patients in oxygenation index (60 +/- 3 vs 63 +/- 8), mean airway pressure (19.5 +/- 0.4 vs 20.8 +/- 1.5 cm H2O), alveolar-arterial O2 gradient (630 +/- 2 vs 632 +/- 4 torr), ECMO cannulation age (median [25th-75th percentiles], 23 [14-47] vs 26 [14-123] hours), or in the % of patients who needed vasopressors/inotropes (98% vs 100%). From November 1994, inhaled nitric oxide (NO) was available. Before VV ECMO, 67% of the patients received NO, 24% received surfactant, and 48% were treated with high-frequency ventilation (HFV). There was no significant difference between VV and VA ECMO patients in survival rate (94% vs 92%), ECMO duration (88 [64-116] vs 94 [55-130] hours), time of extubation (9 [7-11] vs 14 [9-15] days), age at discharge (23 [18-30] vs 27 [15-41] days), or incidence of short-term intracranial complications (5.3% vs 16.7%). For the total cohort of 126 infants, indices of disease severity (oxygenation index, alveolar-arterial O 2 gradient, mean airway pressure) did not correlate with outcome measures. Delay in ECMO initiation (> 96 hours) was associated with prolonged mechanical ventilation and hospitalization (P < .01). New therapies (NO, HFV, surfactant) in the second part of the decade were associated with a longer ECMO duration (98 [80-131] vs 87 [60-116] hours; P < .05), no delay in ECMO initiation time (23 [10-40] vs 24 [14-52] hours), and no significant change in survival (97% vs 92.5%). No patient was treated with VA ECMO after 1994.

Conclusions: Venovenous ECMO is as reliable as VA ECMO in newborns with MAS in severe respiratory failure who need ECMO. Delay in ECMO initiation may result in prolonged mechanical ventilation and increased length of hospital stay. The emergence of new conventional therapies (NO, HFV, surfactant) and particularly increased experience enable sole use of VV ECMO with no significant change in survival in infants with MAS.