ADVANCES in prenatal ultrasonography have improved the ability to detect fetal anomalies of the airway and respiratory system that would be incompatible with life outside the uterus. Extracorporeal membrane oxygenation (ECMO) is increasingly used to care for the neonate with severe pulmonary pathology and respiratory distress. 1Some disorders of the neonatal airway are so severe, however, that techniques have been developed to provide oxygenated blood to the neonate by maintaining fetal–placental circulation until the infant's airway is secured. The ex utero intrapartum technique (EXIT) procedure was initially developed to allow uteroplacental circulation to continue while an artificial airway could be established in cases in which congenital abnormalities of the fetal head and neck could impair spontaneous respiration after delivery. 2This procedure has been predominately used for managing giant fetal neck masses, such as teratomas and lymphangiomas, as well as for neonates with in utero –applied tracheal clips to induce lung growth stunted by congenital diaphragmatic hernia. 3–6However, some neonates with a significantly compromised airway may also require ECMO due to concomitant severe lung pathology. The time required to initiate ECMO may lead to unacceptable hypoxia if an airway cannot be immediately established. We present a case report describing the anesthetic management involved when an EXIT procedure was used to facilitate initiation of ECMO in a neonate with a large pulmonary arteriovenous malformation.
A 26-yr-old gravida 2 para 1 woman with an uncomplicated pregnancy and no significant medical history was referred to our institution after an ultrasonographic examination at 21 weeks gestation revealed a fetus with a large pulmonary arteriovenous malformation, severe cardiomegaly, and a large-diameter pulmonary artery that may have been obstructing the fetal airway. The fetus had no other malformations or obvious pathology. Because further diagnostic studies would be required to evaluate the cardiopulmonary anomaly before attempting definitive surgical correction, a multidisciplinary team decided that ECMO should be instituted immediately upon delivery. Cannulation could take several minutes, however, and it would be necessary to provide some method of oxygenation. In anticipation of severe respiratory distress and the potential difficulty securing an airway immediately after birth, the decision was made to perform an EXIT procedure before establishing ECMO. General anesthesia was planned to permit control of uterine tone.
The patient was admitted at 38 1/7 weeks for elective cesarean section. A preoperative physical exam revealed a 64-in, 60-kg healthy parturient with a Mallampati I airway. Preoperatively, the patient received routine aspiration prophylaxis and was prehydrated with a liter of normal saline. Following rapid sequence induction, anesthesia was maintained with a 50/50 mix of nitrous oxide and oxygen with 1% end-tidal isoflurane. Fentanyl, 300 μg, was given within the first 10 min. Muscle relaxation was maintained with atracurium.
At the point of uterine incision, Fio2was increased to 100%, and isoflurane was increased from 1.8% to 2.6% end-tidal to provide uterine relaxation for delivery of the fetal head and neck. Ephedrine boluses, totaling 95 mg over the course of the case, were necessary to maintain maternal systolic blood pressure greater than 100 mmHg.
Fetal membrane rupture occurred 5 min after uterine incision; the neonate's head, shoulders, and right arm were delivered a minute later. A 24-gauge peripheral intravenous catheter was placed in the right hand of the infant while he was orally intubated with a 3.5 endotracheal tube, using a Miller O laryngoscope blade. Because of fetal lung hypoplasia, we did not plan to mechanically ventilate the child. The goal was to protect the airway and to have the means to ventilate if ECMO cannulation was prolonged or fetal distress occurred. The infant was monitored with continuous pulse oximetry on the right hand to measure preductal oxygen saturation and echocardiography. Fentanyl (10 μg) and pancuronium (1 mg) were administered intravenously to the infant just prior to ECMO cannulation. The neonatal Spo2ranged from 26% to 65%, and 0.2 mg intravenous atropine was given for a heart rate of 90 beats/min. After successful cannulation of the internal jugular vein, the child was delivered, and ECMO was initiated 6 min later. The total elapsed time from delivery of the head to ECMO was 44 min. Umbilical blood gas measurements immediately after initiation of ECMO revealed an arterial pH of 7.26, a Paco2of 52, a Pao2of 36, an HCO3of 23, and an Sao2of 76%; a venous pH of 7.33, a Pvco2of 42, a Pvo2of 51, an HCO3of 21.9, and an Spo2of 91%.
After ECMO was established, isoflurane was decreased to 1.1% end-tidal, and nitrous oxide was restarted at 70% with 30% O2. The placenta was delivered spontaneously 5 min after the child was delivered. Though uterine tone was firm, 30 U oxytocin in 1,000 ml lactated Ringer's solution were administered to the mother intravenously. Total estimated blood loss was 600 ml. The mother's postpartum course was uncomplicated, and she was discharged to home on the second postoperative day.
The approximately 3,000-g baby was taken to the Neonatal Intensive Care Unit, where he remained on ECMO. On the second day of life, repair of the pulmonary arteriovenous fistula, patent ductus arteriosus, and atrial septal defect were performed. The child's postoperative course was complicated by bleeding and hemodynamic instability. Ultimately, the severity of pulmonary hypoplasia was deemed nonsurvivable. On the fourth postpartum day, ECMO was withdrawn, and the child expired secondary to respiratory failure.
Increasingly sophisticated techniques used in perinatal medicine, such as the EXIT procedure, may change how an anesthesiologist balances the risks and benefits of interventions and therapeutics that affect the mother and those that may affect the neonate. For example, frequently, medications with the potential to blunt neonatal ventilation, such as opioids, are used sparingly or not at all. In this case, anticipation of the use of ECMO and the subsequent need for neonatal sedation allowed use of narcotic prior to delivery. Fentanyl was selected for its rapid onset, ability to cross the placental membrane, and relative lack of effect on uterine tone, uterine blood flow, and maternal and fetal hemodynamics. 7
Optimal uterine relaxation was essential to prevent premature placental separation prior to the establishment of ECMO and to allow for fetal oxygenation. We used 2 to 3 minimum alveolar concentration (MAC) isoflurane to achieve adequate uterine relaxation while not compromising both maternal systolic blood pressure and fetal perfusion. Although higher MACs of 3 to 4 have been shown to produce significant tocolysis, 8isoflurane administered at an end-tidal concentration of greater than 2.0% for greater than 30 min has been shown in animal studies to result in fetal acidosis due to decreased fetal heart rate, oxygen saturation, cardiac output, and blood pressure. 9We were prepared to provide additional uterine relaxation if indicated with bolus doses of 50–100 μg intravenous nitroglycerine.
The rapid establishment of neonatal intravenous access permitted administration of additional opioid and neuromuscular blocker prior to ECMO cannulation and provided a route for emergency drugs. The presence of several anesthesiologists allowed for multiple interventions to be accomplished at the same time while assuring proper care to the anesthetized mother.
Following the successful tracheal intubation, the neonate demonstrated relatively low oxygen saturations of 26–65%. Previous investigations of fetal arterial oxygen saturation during labor have identified Spao2values of 30% as the threshold value for fetal compromise. Initially, we were concerned that these values reflected inadequate uteroplacental perfusion. 10One explanation is that with uterine incision and decompression, partial placental separation may have occurred, leading to inadequate uteroplacental blood flow. 11Also, the large pulmonary arteriovenous malformation may have been a cause of this apparent hypoxia and underscored the urgency for establishing ECMO. The low Spo2values may also have been attributed to unique aspects of the EXIT to ECMO period. We have observed in all previous EXIT procedures that initial fetal Spo2values are in the low 30% range. Finally, movement artifact 12from the difficulty maintaining good oximetry probe contact may have contributed to the observed oxygen saturation readings. Obtaining a cord blood sample was entertained but rejected because it would not alter our management and could delay ECMO. However, the near-normal cord gases soon after ECMO was initiated suggest that fetal perfusion was adequate during the time of the EXIT procedure.
In summary, in the setting of neonatal pulmonary and airway abnormalities, the combination of two very technically challenging procedures, such as EXIT and ECMO, may be attempted. Familiarity with intrauterine neonatal physiology and maternal uterine response to anesthetics is critical to successfully manage an EXIT to ECMO procedure.