Providing anesthesia for fetal surgery is a relatively new endeavor, but as the number of centers performing fetal surgery increases, so will the number of cases. Some fetuses are not at risk of death from their disease until they are born. A classic example is the fetus with a tumor obstructing the airway. In utero, the placenta serves as the organ of respiration, but at birth, airway obstruction will be fatal. An ex utero intrapartum therapy (EXIT) procedure is performed late in gestation with the aim of treating the fetus on “placental bypass.” After maternal hysterotomy, the fetus undergoes a procedure (eg, tracheostomy) that will allow it to transition to extrauterine life. The placenta continues to serve as the organ of respiration during the fetal procedure. One of the goals of the maternal anesthetic is to optimize fetal perfusion by ensuring uterine relaxation and adequate maternal blood pressure. The umbilical cord is not divided until after the fetal procedure is completed.
Much of the literature on the management of anesthesia for open fetal surgery is limited to case reports or descriptions of general principles and techniques.1–5 Detailed clinical data are only now beginning to be published.6 We present data focused on the anesthetic management of 65 EXIT procedures performed over 13 years at our institution. The details of the surgical management of many of these cases have been described in other articles.7–10
The IRB at the Children’s Hospital of Philadelphia approved this study and waived the requirement for informed consent. Demographic information from fetal surgical cases at our institution is recorded in a Microsoft Access database. From this database, patients who underwent EXIT procedures from 1998 to 2011 were identified and included in this study. Clinical data were collected from various sources, including electronic anesthesia records, operative reports, and progress notes. Fetal bradycardia was recorded for this analysis if bradycardia was noted in either the operative report or the anesthesia record. The clinical teams typically define fetal bradycardia as heart rate <100 beats/min. Neonatal mortality is defined as death in the first month of life. Atelectasis was present if it was reported in a progress note.
Statistical analysis (JMP 11, SAS Institute, Cary, NC) is largely limited to summary and description of the data. The sample size was chosen for convenience. Consistent information was not available from the anesthesia records before 1998, and 2011 marked the end of another project with EXIT procedure patients. Wilcoxon rank sum test and Fisher exact test were used to compare groups for exploratory analyses (eg, fetal procedure categories, maternal parity). Simple linear regression was used to examine the trends of Fio2 over time. Logistic regression was used to explore the relationship between continuous variables, such as length of fetal procedure and fetal bradycardia or need for fetal chest compressions. Normally distributed data are presented as mean (±SD), otherwise medians (range) are reported.
Sixty-seven EXIT procedures were performed between July 1998 and February 2011. Two cases were excluded from this analysis because of incomplete data. The preoperative demographics are presented in Table 1. The fetal diagnoses and procedure type are presented in Table 2. The airway category included cases of successful intubation using less invasive methods, such as direct laryngoscopy and flexible or rigid bronchoscopy. Complex airway procedures included tracheostomy, tumor resection, cyst drainage, neck dissection, or retrograde passage of wire via a tracheotomy to guide oral intubation. The complex procedures were performed only if less invasive approaches to secure the airway did not succeed. Thoracotomies were performed for resection of large lung lesions. Miscellaneous cases included cannulation for extracorporeal membrane oxygenation, debulking of sacrococcygeal teratoma, or median sternotomy for resection of mediastinal teratoma. Table 3 includes summary details of the operations based on procedure category.
Thirty-two percent of procedures occurred before their planned date and time. The reasons for earlier than desired timing included preterm premature rupture of membranes, onset of labor, development of hydrops fetalis, or maternal mirror syndrome. Twelve cases were designated as emergencies in the anesthesia record.
After aspiration prophylaxis, left uterine displacement, and rapid sequence induction, general endotracheal anesthesia with maternal paralysis was used in all cases. The majority of patients had one 18-g and one 16-g peripheral IV catheter. At our institution, anesthesia management evolved over time during the study period. Preoperative epidural placement for postoperative analgesia became routine in 2002. Two of the patients had epidural catheters placed while the mother had general anesthesia (after the fetal procedure but before tracheal extubation) because of the emergency nature of their procedures. Invasive monitoring of arterial blood pressure through a radial arterial line became routine in 2003. Phenylephrine infusions have been commonly used to maintain arterial blood pressure since 2006.
Maternal Fio2 during the fetal procedure (which is defined as the time between hysterotomy and clamping of the umbilical cord) has been decreasing over the years (P = 0.0001). Patients in the first half of the cohort had a Fio2 range of 75% to 100%, and Fio2 in the second half of the cohort ranged from 21% to 80%. Nitrous oxide was administered to 41 patients after the umbilical cord was clamped; nitrous oxide was administered for the entire operation to 13 patients. One patient received no nitrous oxide, 1 was not documented, and 1 had nitrous oxide before, but not after umbilical cord clamping. No trends in nitrous oxide use over the years were noted. Mothers who had delivered 1 or more children before the EXIT procedure were less likely to require nitroglycerin boluses for uterine relaxation than those who had not delivered any other children (3% vs 31%, P = 0.003). Use of a nitroglycerin infusion was 0% and 7% in these groups, respectively (P = 0.2). During the study period, nitroglycerin administration was typically reserved for high uterine tone refractory to increases of volatile anesthetic concentration or for patients who did not tolerate high doses of volatile anesthetic.
Volatile anesthetic concentration was recorded by the electronic anesthesia record keeping system (CompuRecord, Philips Healthcare, Amsterdam, The Netherlands). Desflurane was used in 61 of 65 patients. Unfortunately, when the electronic record keeping system detected desflurane levels >12%, the measure for a given time period was automatically invalidated. After the inspired and expired desflurane concentrations increased to 12%, periods with no recorded values were present in the record. Thus, the maximum doses of desflurane cannot be reported, but 47 of the 61 patients received inspired desflurane concentrations >12%. Oxytocin was typically used to establish uterine tone after delivery and umbilical cord clamping; methylergonovine and prostaglandins were added in some cases. There was no relationship between parity and estimated blood loss (960 vs 930 mL, nulliparous vs parous, P = 0.4). However, there may have been a relationship between parity and use of any uterotonic in addition to oxytocin (34% vs 8%, nulliparous vs parous, P = 0.01).
Fetuses were monitored using different techniques at various times during the surgical procedure. Ultrasound assessment of fetal heart rate was used before access to a fetal limb was obtained. Pulse oximetry was used after the fetal limb was exposed. For selected cases, such as fetal pulmonary lobectomy, a cardiologist conducted real-time echocardiography to monitor fetal heart rate, cardiac filling, and function.
Fetal bradycardia occurred in each of the operative categories (15 cases). Details of the resuscitation of the bradycardic fetuses are described in Table 4. Umbilical cord complications (ie, kinking, compression, stretch) were the most common reason for bradycardia. Resuscitation most commonly included repositioning of the fetus to ensure umbilical cord patency. In certain cases, the fetus required medications, chest compressions, crystalloid, or red blood cell transfusion. In cases involving the airway, expeditious control of the airway was often the definitive therapy.
There were 4 “rushed deliveries” where the umbilical cord was clamped and cut sooner than initially desired. Two of these were in cases of overt placental abruption, and the other 2 were in cases in which the surgical team noted placental abruption developing, but the delivery was completed before abruption occurred. In one of the fetuses undergoing thoracotomy, red blood cells and crystalloid were administered to the fetus with the guidance of the cardiologist, who noted underfilling of the fetal heart as the lung tumor was being mobilized and removed. There was no relationship between length of fetal procedure (hysterotomy to cord clamp interval) and incidence of fetal bradycardia (P = 0.88) or need for fetal chest compressions (P = 0.79). Maternal age, parity, Fio2, and gestational age of the fetus at the time of surgery had no relationship to the incidence of bradycardia (all P > 0.39).
Most fetuses received an IM injection of fentanyl (20 μg/kg), vecuronium or pancuronium (0.2 mg/kg), and atropine (20 μg/kg) before the start of the fetal procedure. Four fetuses in the airway category did not receive this injection. After fetal exposure, the airway management of these 4 fetuses was deemed easier than anticipated, and the fetal IM injection was deferred. A few fetuses in each group did not have either atropine or neuromuscular blockade documented in the anesthesia record, but did have IM fentanyl administration documented. This discrepancy was most likely a documentation error from the clinical team, because the drugs for IM injection are typically drawn into 1 syringe. Documentation of fentanyl can be expected to have much higher compliance than that of neuromuscular blocking drugs, because opioid administration must be reconciled with pharmacy records.
Two mothers required transfusion of 1 unit of packed red blood cells (PRBCs), 2 mothers received 2 units PRBCs, and 2 mothers were transfused with 3 units PRBCs. One mother received 6 units PRBCs and 750 mL albumin. Details of the intraoperative course of the mother/fetus dyads are presented in Tables 3 to 6. Reasons for blood transfusion included placental abruption (n = 3), bleeding from uterine venous lakes after hysterotomy (n = 1), uterine atony after delivery (n = 1), preoperative anemia (n = 1), and postpartum anemia (hematocrit of 23%; n = 1). There was no difference in the need for additional uterotonic agents among the 4 procedure categories (complex airway, airway, thoracotomy, or other; P = 0.5). Maternal postoperative complications are noted in Table 7.
The overall fetal/neonatal mortality rate was 15%. Intraoperative fetal demise occurred in 1 case in this series of 65 EXIT procedures. The diagnosis was congenital high airway obstruction syndrome. The fetus had a rudimentary trachea superior to the carina, and a stable airway could not be established. Neonatal mortality was 23% (5/21) in the complex airway group, including the fetus with intraoperative demise. Four neonates who required complex airway management for huge cervical teratomas died of pulmonary hypoplasia in the neonatal intensive care unit.
Mortality in the airway group was 5% (1/20). Eighteen EXIT procedures were performed, but 2 of the mothers were carrying twins, so the total number of infants in this group was 20. The twin who died was a parasitic thoracoomphalopagus twin with a rudimentary heart who was completely dependent on the twin with a normal heart. The EXIT was performed for airway and IV access before emergency conjoined twin separation in an adjacent operating room with planned sacrifice of the parasitic twin to permit salvage of the twin with the normal heart.
Two of 21 EXIT-to-thoracotomy infants died in the neonatal period (birth to 1 month of life), 1 from sequelae of severe pulmonary hypoplasia. The other infant was a 27-week gestation infant who succumbed to bleeding complications requiring emergency exploration of the chest. The 2 neonates who died in the “other” category were treated with an EXIT-to-extracorporeal membrane oxygenation strategy and died of severe congenital diaphragmatic hernia and complex congenital heart disease.
We present data from 65 EXIT procedures performed over 13 years at our institution. Providing anesthesia for fetal surgery has unique considerations in addition to the fact that there are 2 patients undergoing a procedure simultaneously. The mother is the first priority, and she must be kept safe in the face of major surgery. Challenges to this goal come from optimizing conditions for the fetal procedure. The uterus must be relaxed to optimize fetal perfusion; however, medications to relax the uterus cause hypotension, which may be detrimental to both mother and fetus. Fetal perfusion must be maintained by keeping maternal blood pressure and cardiac output high enough and ensuring adequate flow of blood from the placenta to the fetus. The primary operating room team must be ready to deal with fetal distress, and additional neonatology and operating room teams must be ready to resuscitate a fetus in extremis if the EXIT procedure does not go as planned. After the fetus is delivered, the uterus must be taken from a state of profound atony to adequate contraction to avoid bleeding complications in the mother.
As the field has developed, the anesthetic performed at our institution has evolved. In the earlier years, after aspiration prophylaxis, left uterine displacement, and rapid sequence induction of general anesthesia, isoflurane was administered through an endotracheal tube, and a second large bore peripheral IV line was placed. In addition to endotracheal intubation and 2 large bore IV catheters, our standard practice now includes invasive arterial blood pressure monitoring, postoperative epidural analgesia, and the use of desflurane instead of isoflurane.4 These changes in practice were the result of different factors. Desflurane became the volatile agent of choice in 1999 because it could be rapidly titrated to effect. Although studies do not clearly support choosing either low or high Fio2 in the perioperative period, Fio2 administered for EXIT procedures at our institution has been decreasing over time because our group has been examining literature calling into the question the need for high Fio2 during adult surgical procedures.11–15 Clinically, a bolus of phenylephrine or ephedrine given to the mother has been noted to improve poor fetal oxygen saturation (20%) to a more acceptable range (eg, 50%). In addition, temporal improvements in myocardial contractility have been noted with administration of maternal vasopressors. Management of nitrous oxide is variable and has not shown any trends over the years. It is still commonly administered after the umbilical cord is clamped to allow a rapid decrease of volatile anesthetic, thus speeding the recovery of uterine tone.
Invasive arterial blood pressure monitoring was introduced in 2002 in response to a mother who had intraoperative placental abruption. She required multiple units of PRBCs and aggressive resuscitation. Arterial catheters also allow for precise blood pressure manipulation that is critical in maintaining fetal perfusion. Because the dose of volatile anesthetic administered is high, maternal hypotension ensues and must be treated with vasopressors and fluid administration. As data from obstetric anesthesia literature began to show improved umbilical arterial pH with the use of phenylephrine to treat hypotension in healthy women undergoing spinal anesthesia for cesarean delivery,16 and as obstetric anesthesia practice filtered to our institution, phenylephrine infusions were introduced in 2006. There are no studies to date to support vasopressor choice in EXIT procedures.
Epidural analgesia was introduced in 2002 after continuing discussion and collaboration between the pediatric and obstetric anesthesia teams involved in the care of these patients. No pain scores are available for direct comparison, but the general perception of the team members is that postoperative analgesia improved dramatically after introduction of this technique.
Volatile anesthetic allows for rapid and easy titration to achieve uterine relaxation and also affords the fetus some degree of anesthesia to blunt the stress response to surgery. However, high-dose volatile anesthetic is associated with fetal myocardial depression and possibly an increased risk of fetal bradycardia.17,18 An alternative to this technique is epidural anesthesia and analgesia.19 Uterine relaxation is achieved with a nitroglycerin infusion. An advantage of this technique is decreased risk of fetal myocardial depression, but longer, technically complicated fetal procedures, such as pulmonary lobectomy, may not be amenable to such a technique. The operating room environment for these cases is highly charged, and a mother who is awake or lightly sedated may contribute to suboptimal operating conditions. A more recently described technique may mitigate the myocardial depression associated with high-dose volatile anesthesia, while still providing good operating conditions. General anesthesia is induced, and low-to-moderate dose volatile anesthetics are administered for uterine relaxation, but infusions of propofol and remifentanil are used to minimize exposure of the fetus to volatile anesthetic until uterine relaxation is required.18
Fetal bradycardia may occur during EXIT procedures. The anesthesia team must ensure that this is not a result of maternal cardiovascular depression. The surgical team must ensure an intact uteroplacental interface and a patent umbilical cord. Both teams must communicate and coordinate care to make decisions in cases of fetal bradycardia or hypoxemia. In some cases, fetal chest compression and medications are necessary, in others, expedited control of the airway is needed, and in some cases, intravascular volume expansion will suffice. Several rounds of emergency medications are drawn up sterilely in unit doses at the beginning of each case. The importance of fetal echocardiography cannot be overstated.20 At our institution, it is routinely used in cases of open fetal surgery for myelomeningocele. An appreciation for the myocardial depressant effects of volatile anesthetics has led to a recent trend (starting in late 2014) in our team for early administration of nitroglycerin instead of maximizing the doses of volatile anesthetic. There are no outcome studies for this new strategy. The hope is that fetal myocardial performance will be better, but the added nitroglycerin may place the mother at risk for uterine atony after the umbilical cord is clamped.
For centers performing EXIT procedures, several points should be emphasized. Undertaking these procedures is no small task. Although most of these cases occur at the scheduled date and time, many do not, and the centers performing these procedures must have the infrastructure to quickly assemble the teams needed during nights and weekends. Additional staff must be ready, including a neonatal team and a second operating room team. Equipment needed for all EXIT procedures (sterile pulse oximeter probes and cables, sterile Mapleson D circuits, oxygen tubing, uterine staplers, etc) stays in mobile storage units that are maintained at all times by fetal surgical specialty nurses. In addition, when a mother who must deliver through EXIT procedure is identified, instruments and trays specific to each case are chosen by the surgeons. These instruments and trays are reserved for use at any time, day or night, until the mother delivers. Fetal bradycardia may occur in any case, even in an uncomplicated EXIT for airway control. Just as in any crisis, roles must be clear, and the teams must communicate well to ensure the safety of the patients.
Fetal surgery as a field will continue to grow as prenatal diagnostic technology improves and surgical training and experience increase. The number of centers in the United States and internationally continues to increase. To optimize anesthetic management of these cases, the proper questions must be asked. The data from our study will aid in generating testable hypotheses. One question that comes to the forefront is the optimal anesthetic technique (high-dose volatile versus supplemental IV anesthesia versus neuraxial anesthetics with nitroglycerin) and the incidence of fetal myocardial dysfunction and bradycardia. Development of a method to accurately and precisely assess uterine tone would greatly help anesthesiologists titrate tocolytic medications. Simple descriptive studies of anesthetic pharmacology also need to be performed but may be technically challenging. For example, the pharmacodynamics and pharmacokinetics of desflurane in the human fetus are not well understood. Questions of neurodevelopmental outcomes after exposure to anesthesia in these fetuses are intriguing, but will be hard to answer given the confounding effect of their congenital anomalies and other comorbidities.
Name: Elaina E. Lin, MD.
Contribution: This author helped analyze the data and write the manuscript.
Name: Julie S. Moldenhauer, MD.
Contribution: This author helped design the study, conduct the study, and write the manuscript.
Name: Kha M. Tran, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Name: David E. Cohen, MD.
Contribution: This author helped write the manuscript.
Name: N. Scott Adzick, MD, MMM.
Contribution: This author helped write the manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
1. Gaiser RR, Kurth CDAnesthetic considerations for fetal surgery.Semin Perinatol19992350714
2. Abraham RJ, Sau A, Maxwell DA review of the EXIT (Ex utero Intrapartum Treatment) procedure.J Obstet Gynaecol20103015
3. Van de Velde M, De Buck FFetal and maternal analgesia/anesthesia for fetal procedures.Fetal Diagn Ther2012312019
4. Lin EE, Tran KMAnesthesia for fetal surgery.Semin Pediatr Surg201322505
5. Ferschl M, Ball R, Lee H, Rollins MDAnesthesia for in utero repair of myelomeningocele.Anesthesiology2013118121123
6. Tran KM, Maxwell LG, Cohen DE, Adamson PC, Moll V, Kurth CD, Galinkin JLQuantification of serum fentanyl concentrations from umbilical cord blood during ex utero intrapartum therapy.Anesth Analg201211412657
7. Hedrick HLEx utero intrapartum therapy.Semin Pediatr Surg2003121905
8. Hedrick HL, Flake AW, Crombleholme TM, Howell LJ, Johnson MP, Wilson RD, Adzick NSThe ex utero intrapartum therapy procedure for high-risk fetal lung lesions.J Pediatr Surg200540103843discussion 1044
9. Laje P, Johnson MP, Howell LJ, Bebbington MW, Hedrick HL, Flake AW, Adzick NSEx utero intrapartum treatment in the management of giant cervical teratomas.J Pediatr Surg201247120816
10. Mackenzie TC, Crombleholme TM, Johnson MP, Schnaufer L, Flake AW, Hedrick HL, Howell LJ, Adzick NSThe natural history of prenatally diagnosed conjoined twins.J Pediatr Surg2002373039
11. Akça O, Podolsky A, Eisenhuber E, Panzer O, Hetz H, Lampl K, Lackner FX, Wittmann K, Grabenwoeger F, Kurz A, Schultz AM, Negishi C, Sessler DIComparable postoperative pulmonary atelectasis in patients given 30% or 80% oxygen during and 2 hours after colon resection.Anesthesiology1999919918
12. Rothen HU, Sporre B, Engberg G, Wegenius G, Hedenstierna GAirway closure, atelectasis and gas exchange during general anaesthesia.Br J Anaesth1998816816
13. Meyhoff CS, Wetterslev J, Jorgensen LN, Henneberg SW, Høgdall C, Lundvall L, Svendsen PE, Mollerup H, Lunn TH, Simonsen I, Martinsen KR, Pulawska T, Bundgaard L, Bugge L, Hansen EG, Riber C, Gocht-Jensen P, Walker LR, Bendtsen A, Johansson G, Skovgaard N, Heltø K, Poukinski A, Korshin A, Walli A, Bulut M, Carlsson PS, Rodt SA, Lundbech LB, Rask H, Buch N, Perdawid SK, Reza J, Jensen KV, Carlsen CG, Jensen FS, Rasmussen LSPROXI Trial GroupEffect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial.JAMA2009302154350
14. Zoremba M, Dette F, Hunecke T, Braunecker S, Wulf HThe influence of perioperative oxygen concentration on postoperative lung function in moderately obese adults.Eur J Anaesth200917
15. Wetterslev J, Meyhoff CS, Jørgensen LN, Gluud C, Lindschou J, Rasmussen LSThe effects of high perioperative inspiratory oxygen fraction for adult surgical patients.Cochrane Database Syst Rev20156CD008884
16. Cooper DW, Carpenter M, Mowbray P, Desira WR, Ryall DM, Kokri MSFetal and maternal effects of phenylephrine and ephedrine during spinal anesthesia for cesarean delivery.Anesthesiology200297158290
17. Rychik J, Tian Z, Cohen MS, Ewing SG, Cohen D, Howell LJ, Wilson RD, Johnson MP, Hedrick HL, Flake AW, Crombleholme TM, Adzick NSAcute cardiovascular effects of fetal surgery in the human.Circulation2004110154956
18. Boat A, Mahmoud M, Michelfelder EC, Lin E, Ngamprasertwong P, Schnell B, Kurth CD, Crombleholme TM, Sadhasivam SSupplementing desflurane with intravenous anesthesia reduces fetal cardiac dysfunction during open fetal surgery.Paediatr Anaesth20102074856
19. Clark KD, Viscomi CM, Lowell J, Chien EKNitroglycerin for relaxation to establish a fetal airway (EXIT procedure).Obstet Gynecol200410311135
20. Rychik J, Cohen D, Tran KM, Szwast A, Natarajan SS, Johnson MP, Moldenhauer JS, Khalek N, Martinez-Poyer J, Flake AW, Hedrick HL, Adzick NSThe role of echocardiography in the intraoperative management of the fetus undergoing myelomeningocele repair.Fetal Diagn Ther2015371728