Virtual Medical Library Home | Table of Contents
David Johnson, MD
Extracorporeal membrane oxygenation (ECMO) is a highly invasive mechanism of cardiopulmonary support. ECMO is only initiated when all other therapies have failed to reverse severe neonatal respiratory failure. ECMO, a modification of the cardiopulmonary bypass techniques initially developed for cardiac surgery, was first reported in neonatal respiratory failure in 1976. [5] Initially, only moribund patients were treated and some survived. As experience with the technique increased, better selection and exclusion criteria were developed. Though ECMO seemed to be improving survival in these selected neonatal patients, mortality risks of 80% that were commonly used to select ECMO patients were based on historical controls, [4] [61] and the real benefit from ECMO was unclear. Two prospective, randomized trials were performed [4] [81] that suggested improved survival with ECMO, but both used unusual statistical analyses and adaptive designs that limited the number of conventionally treated patients. Because of the ethical dilemma of withholding potentially life-saving therapy that had become widely available, it became almost impossible to conduct a prospective randomized trial in the United States. In the United Kingdom, ECMO was not as widely available or enthusiastically embraced initially. Between 1993 and 1995, a randomized, prospective trial of ECMO in neonatal respiratory failure was conducted. This study was terminated early when a clear survival advantage was demonstrated in the ECMO treated group. [105] The mortality rate in the conventional group was 59%, compared with a 30% mortality in the ECMO group. This survival advantage was seen in all diagnostic categories.
ECMO is a support technique that allows the underlying disease process to be corrected. Selecting patients who are at high risk for mortality, and therefore are likely to benefit from this potentially risky support, is often difficult. The most common criteria to select neonatal ECMO patients is an oxygenation index (defined as: [Mean Airway Pressure × Fi O2 × 100 / Pa O2 ] greater than 40. [82] Alveolar-arterial gradient greater than 600 [61] and acute deterioration (pH < 7.15 and Pa O2 < 40 mm Hg) [92] have also been used to predict mortality and qualify for ECMO. Contraindications to ECMO include weight less than 2 kg, gestational age less than 35 weeks, intracranial hemorrhage more severe than grade 1, age more than 10 days, irreversible lung disease, and significant brain injury or congenital anomaly that precludes future meaningful life experiences. Cranial ultrasound to identify intracranial bleeding and echocardiogram to exclude structural cardiac anomalies are generally performed before starting ECMO.
Cannulation for ECMO is via the right internal jugular vein for veno-venous bypass or the vein and the right common carotid artery are cannulated for veno-arterial bypass. In patients who have maintained adequate cardiac function, veno-venous bypass is preferred to preserve the carotid artery which may be ligated when veno-arterial ECMO support is terminated.
The most worrisome complication of ECMO support is bleeding. Systemic heparinization is required to prevent clotting in the ECMO circuit. The most serious bleeding complication, intracranial hemorrhage, is becoming less common with decreased amounts of heparinization. [31] Major mechanical problems occur in less than 10% of patients and are becoming less common as experience with the technique is increasing and devices designed for longer-term use are developed. [31]
Overall survival from the 1997 ELSO registry is 80% but differs according to diagnosis: meconium aspiration, 94%; RDS, 84%; primary pulmonary hypertension, 82%; sepsis, 77%; and diaphragmatic hernia, 58%. [31]
Survival is one measure of success, but sequelae affecting quality of life are important as well. Generally the outcome of a patient relates to the underlying disease. Occasionally, however, because of the lack of a timely referral, the infant may experience increasing episodes of hypoxemia and acidosis with potential long-term neurologic morbidity. Patients who survive a prolonged course of mechanical ventilation and develop severe BPD have a similar neurodevelopmental outcome as those requiring ECMO. [1] [67] It is difficult to separate the pre-ECMO treatment from ECMO itself in determining the cause of any neurodevelopmental impairment. Close follow-up of these patients by a developmental pediatrician or neurologist is required.
In summary, ECMO is used to support a population of patients with a historically high risk for mortality and has been shown to improve survival. Determination of disease severity and timely referral will help ensure an optimal outcome. Though certainly not free of complications, ECMO offers a chance for survival to a significant number of newborns who have failed all other therapy.