(EXIT) Procedure: Anesthetic Implications

Four cases of the ex utero intrapartum treatment (EXIT) procedure: anesthetic implications

G. Dahlgren, D. C. Tornberg,€ K. Pregner, L. Irestedt Department of Anesthesia and Intensive Care, Karolinska Hospital and Institute, SE 171 76 Stockholm, Sweden

SUMMARY. The ex utero intrapartum treatment (EXIT) procedure is a method of maintaining utero-placental circulation during cesarean section to gain time to secure a potentially obstructed fetal airway. Four cases of the EXIT procedure are described with special reference to the maternal anesthetic technique. Deep volatile anesthesia (ap- proximately 2 MAC) with isoflurane or sevoflurane for a prolonged period of time, in three cases in combination with an intravenous nitroglycerin infusion, was used to ensure a fully relaxed uterus during the procedure. All mothers were maintained hemodynamically stable with preserved utero-placentary perfusion. It was possible to intubate the tracheas of two fetuses, whereas in the other two tracheostomies had to be performed. Fetal gas exchange was not negatively affected during the EXIT procedure as evidenced by normal blood gas values in the umbilical artery at the time of delivery. After reducing the concentration of volatile anesthetic, delivery of the neonate and administration of oxytocin, uterine contractility was promptly re-established and there were no signs of uterine atony in the postoperative period. All four neonates survived the procedure without complications. Ó 2004 Elsevier Ltd. All rights reserved.

INTRODUCTION The EXIT procedure is performed in conjunction with an elective cesarean section. The fetus is only The ex utero intrapartum tracheoplasty procedure was partially delivered from the uterus with maintenance of originally developed for the fetus with severe congenital the utero-placental circulation allowing for diagnostic diaphragmatic hernia treated in utero with temporary and/or therapeutic procedures to the fetal airway. tracheal occlusion to enhance lung growth. With this Commonly, the fetus has been anesthetized and immo- technique utero-placental circulation could be main- bilized by the intramuscular route before the start of tained during removal of the tracheal clips immediately airway manipulations. When the airway is secured the followed by cesarean delivery.1 The technique has since fetus is delivered and the umbilical cord clamped. Main been used also for fetuses with giant neck masses and prerequisites during the EXIT procedure include com- other diseases with an increased risk for airway man- plete uterine relaxation and maintained uterine volume agement problems directly after delivery. It has for this and utero-placental circulation. Uterine relaxation has in reason been renamed ex utero intrapartum treatment most published cases been accomplished by deep vola- (EXIT) or operation on placental support (OOPS).2–4 tile anesthesia but other means of treatment are also After the first EXIT procedures were reported in the available.3;6 Uterine volume is preserved by the fact that mid-nineties many case reports and case reviews have the main part of the torso and the lower part of the fetus been published of which the review of 31 cases from the are retained in the uterus during the procedure. In many Children’s Hospital of Philadelphia is the largest.5 centers warmed Ringer’s lactate solution is also infused into the uterine cavity. In order to maintain the utero- placental blood flow, maternal cardiac output and blood ––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Accepted January 2004 pressure have to be kept at acceptable levels. Between 1998 and 2002 four EXIT procedures G. Dahlgren MD, PhD, Consultant Anesthetist; D.C. Tornberg€ MD, Consultant Anesthetist; K. Pregner, Nurse Anesthetist; have been performed at the Karolinska Hospital, L. Irestedt MD, PhD, Associate Professor, Chairman. Stockholm. Case number one has previously been published in detail.7 The aim of this case review is to Correspondence to: Dr. Gunnar Dahlgren, Department of Anesthesia and Intensive Care, Karolinska Hospital, SE 171 76 Stockholm, highlight the maternal anesthetic technique and also to Sweden; E-mail: [email protected] discuss possible implications for the use of volatile

178 Four cases of the ex utero intrapartum treatment (EXIT) procedure 179 anesthetics during general anesthesia for standard ce- direct laryngoscopy confirmed the diagnosis of lar- sarean section. yngeal atresia and a tracheostomy was performed. Manual ventilation was started and surfactant instilled into the tracheostomy tube after which the fetus was CASE PRESENTATIONS delivered and the cord clamped. ET isoflurane was re- duced to 0.9% and oxytocin 5 units was administered Relevant maternal and neonatal data are summarized in i.v. followed by slow infusion of 50 units. The recovery Tables 1 and 2. of uterine contractility was indeed very quick (within Case one was a healthy 28-year-old primigravida. In minutes) and there were no problems with uterine atony week 27 the fetus was found to have a laryngeal ste- postoperatively. Maternal hemodynamics remained sta- nosis/atresia. An EXIT procedure was performed at 35 ble throughout the procedure and blood loss was mod- weeks’ gestation. The patient was placed supine with erate. At four years of age the child is doing fine with a left lateral tilt to avoid aortocaval compression. General normal clinical and neurological status, although still anesthesia was then induced with thiopentone 400 mg, breathing via a tracheostomy. fentanyl 200 lg and succinylcholine 100 mg. After en- Case two was a healthy 29-year-old multiparous dotracheal intubation anesthesia was maintained with woman. In week 28 the fetus was found to have a large isoflurane and nitrous oxide in oxygen (50-50) com- oral cyst. An EXIT procedure was performed at 37 bined with additional doses of fentanyl and vecuronium weeks’ gestation. The patient was placed supine with for muscle relaxation. The nitrous oxide was discontin- left lateral tilt to avoid aortocaval compression. Gen- ued before uterine incision and the FiO2 was increased eral anesthesia was then induced with thiopentone 500 to 1.0. End-tidal (ET) isoflurane concentration was in- mg, fentanyl 100 lg and succinylcholine 100 mg. After creased to 2.2% during 10–15 min before uterine inci- intubation, anesthesia was maintained with sevoflurane sion. At the time of uterine incision the obstetrician in oxygen-air with FiO2 0.5. Vecuronium was admin- confirmed that the uterus was fully relaxed. During the istered for muscular relaxation together with additional EXIT procedure ET isoflurane concentration was kept small doses of fentanyl. The sevoflurane concentration between 1.8 and 2.2%. Bleeding was controlled by a was increased to 4.5% ET for 10 min to achieve full continuous suture around the uterotomy. The head, right uterine relaxation. Nitroglycerin 1 lg Á kgÀ1 minÀ1 was arm and part of the torso of the fetus were then deliv- infused i.v. beginning before uterine incision and run- ered. Based on the estimated fetal weight an intramus- ning until clamping of the cord. During the EXIT cular injection of fentanyl 10 lg/kg, vecuronium 0.2 procedure ET sevoflurane concentration was kept be- mg/kg and atropine 10 lg/kg was given to the fetus. A tween 2.8 and 4.5%. Bleeding was controlled by a sterile pulse oximetry probe was positioned on the right continuous suture around the uterotomy. The fetus was hand but a reliable reading could not be obtained. The anesthetized intramuscularly and monitored as in case fetal heart rate was monitored with ultrasonography. A 1. The fetus could be intubated nasally after aspiration

Table 1. Maternal data

Case 1 Case 2 Case 3 Case 4

Anesthetic agent Isoflurane Sevoflurane Sevoflurane Sevoflurane Additional tocolytic No Nitroglycerin Nitroglycerin Nitroglycerin Skin incision to uterotomy (min) 11 10 8 10 Blood loss (mL) 700 200 1700 2000 Post partum uterine atony No No No No

Table 2. Neonatal data

Case 1 Case 2 Case 3 Case 4

Diagnosis Laryngeal atresia Dermoid cyst Lymphangioma Lymphangioma Airway procedure Tracheostomy Intubation Tracheostomy Intubation Uterotomy to cord clamp time (min) 24 14 62a 29 Umbilical artery pH 7.30 7.30 7.16 7.22 Umbilical artery PO2 (kPa) 5.5 8.6 3.9 3.3 Umbilical artery PCO2 (kPa) 6.9 7.0 9.1 8.8 Umbilical artery base deficit (meq/L) 2.8 2.1 6.1 2.8 Umbilical artery lactate (mmol/L) Not analyzed 2.3 6.0 4.1 a See text Case three for comments. 180 International Journal of Obstetric Anesthesia of liquid contents from the sublingual cyst. After was a lymphangioma which was to be treated with in- clamping of the cord oxytocin 10 units was given i.v. jections of the sclerosing agent OK-432 (Picibanil). At followed by a slow infusion of 100 units. Prostinfenem one year of age the child was normally developed al- (15-metyl prostaglandin F2a) 0.25 mg was given into though still breathing through the tracheostomy. the uterine wall. Within a minute after clamping the Case four was a 24-year-old healthy nulliparous cord and reducing the ET sevoflurane concentration to woman with a history of three missed abortions. She was 1.4%, the uterus started to contract. Maternal hemo- pregnant in week 29 when the fetus was found to have a dynamics remained stable throughout the procedure tumor on the left side of the face and neck. At 31 weeks and bleeding was minimal. The sublingual cyst which the tumor measured 7 Â 8 cm. An EXIT procedure was turned out to be a dermoid cyst increased in size performed at 35 weeks’ gestation. The patient was postoperatively; it was marsupialised the day after de- placed supine with left lateral tilt to avoid aortocaval livery and resected surgically two months later. The compression. General anesthesia was then induced with neonate developed normally and had no airway thiopentone 425 mg, fentanyl 100 lg and succinylcho- symptoms at one year of age. line 100 mg. Following intubation anesthesia was Case three was a 37-year-old multiparous woman who maintained with sevoflurane in oxygen-air with FiO2 had two years earlier been successfully treated for a ma- 0.7. Vecuronium was used for muscular relaxation. Ni- lignant mediastinal lymphoma. At 17 weeks’ gestation the troglycerin 1 lg Á kgÀ1 minÀ1 was infused i.v. beginning fetus was found to have a tumor anteriorly in the neck. In before uterine incision and running until clamping of the the following weeks the tumor showed substantial growth cord. The sevoflurane concentration was increased to and at 33 weeks had a diameter of 10.5 cm. An EXIT 3.5% ET for more than 10 min before incision of a procedure was performed at 36 weeks’ gestation. The completely relaxed uterus. During the EXIT procedure patient was placed supine with left lateral tilt to avoid sevoflurane concentration was kept between 1.8 and aortocaval compression. General anesthesia was then in- 3.5%. The fetus was anesthetized intramuscularly as in duced with thiopentone 400 mg, alfentanil 1 mg and case 1. Fetal heart rate was monitored by ultrasonogra- succinylcholine 100 mg. After intubation, anesthesia was phy and oxygen saturation was successfully monitored maintained with sevoflurane in oxygen-air with FiO2 0.7. via the right hand. Fetal oxygenation was normal Vecuronium was used for muscle relaxation. Nitroglyc- throughout the procedure. The fetal airway could be erin 1 lg Á kgÀ1 minÀ1 was infused i.v. beginning before secured with a nasal endotracheal tube. After delivery of uterine incision and running until clamping of the cord. the neonate and clamping of the cord the sevoflurane Sevoflurane concentration was increased to 4% ET for 8 concentration was decreased and oxytocin 5 units was min before the completely relaxed uterus was incised. given i.v. followed by a slow intravenous infusion of 50 During the EXIT procedure ET sevoflurane concentration units. Uterine contractility was quickly re-established. was kept between 1.9 and 4%. The fetus was anesthetized Blood loss was considerable despite the continuous su- intramuscularly and monitored as in case 1. The fetus was ture around the uterotomy and three units of packed red successfully intubated. After ventilation of the fetus was cells were transfused. Maternal hypotension was, how- established, the placental circulation seemed to cease as ever, well controlled with small ephedrine boluses and a manifested by disappearing pulsations in the umbilical low-dose phenylephrine infusion. There were no bleed- cord. Due to the huge size of the tumor and the risk for ing problems after the delivery of the neonate. The fetal dislocation of the endotracheal tube postoperatively the tumor turned out to be a lymphangioma. The neonate situation was reconsidered and a tracheostomy per- could be extubated after three days. At 7 months of age a formed. A total time of 62 min passed between uterotomy facial nerve palsy was present but the child was other- and clamping the cord. Out of these 62 min we roughly wise fine and did not require an artificial airway. estimate that the fetus depended on utero-placental perfusion only during the first 30 min. The blood sample from the umbilical artery at the time of cord clamping is DISCUSSION therefore difficult to interpret. After delivery and cord clamping, the sevoflurane concentration was decreased The dose of volatile anesthetics required for EXIT and oxytocin 5 units was given i.v. followed by a slow procedures is quite different from that for standard ce- intravenous infusion of 100 units. Blood loss was fairly sarean section. In the 1960s a thiopentone/nitrous oxide/ high despite the continuous suture around the uterotomy muscular relaxant technique of anesthesia was in general but blood transfusion was not considered necessary. The use for cesarean section. In 1970 it was shown that the maternal circulation was maintained by intermittent in- addition of a small amount of halothane (0.5% inspired jections of ephedrine. Uterine contractility was quickly concentration) to 2–3 mg/kg thiopentone and 50% ni- reestablished after delivery of the neonate and from that trous oxide in oxygen decreased the incidence of ma- point on there were no bleeding problems. The fetal tumor ternal awareness with postoperative recall from 4 to 0%. Four cases of the ex utero intrapartum treatment (EXIT) procedure 181

Blood loss was not increased and Apgar scores were with fetal blood acidosis.19 Generally, during anes- improved compared to the old combination of thiopen- thesia for fetal procedures lasting less than one hour tone and 70% nitrous oxide.8 Many textbooks still rec- in animal models, 1–1.5 MAC of a potent agent seem ommend the use of such a low dose of volatile agent to be well tolerated. This dose-range is also used for (corresponding to 0.5–0.7 MAC inspired concentration) human fetal surgical procedures.20 In our cases all to decrease the risk for maternal awareness without in- four neonates were anesthetized and given muscule ducing neonatal depression, uterine atony and increased relaxants by the intramuscular route during the EXIT maternal blood loss.9;10 It is, however, important to re- procedure, in addition to the volatile agent given alize that higher concentrations of volatile anesthetics through the mother. After delivery the neonates were could be necessary to eliminate the risk of awareness11 mechanically ventilated until the effect of the anes- and also to counteract the negative effects on utero- thetics wore off. Assessment of the neonatal effects of placental perfusion from the maternal autonomic stress the volatile anesthetic was therefore not possible at response.12 During an EXIT procedure the ET concen- the time of birth. We can nevertheless conclude that tration of the volatile agent is kept very high (approxi- neither prolonged deep volatile anesthesia (approxi- mately 2 MAC) in order to keep the uterus fully relaxed, mately 2 MAC) nor uterine incision and manipulating provided that maternal hemodynamics and hence utero- of the fetus caused any serious disturbances in fetal placental flow are not jeopardized. In standard cesarean gas exchange as evidenced by the blood gas values. section, the time from uterine incision to clamping the This finding is in line with others.5 umbilical cord is usually intended to be well below three Maternal intraoperative blood loss during the EXIT minutes, in order to avoid the assumed deterioration of procedure is exacerbated by the relaxed state of the fetal acid-base status with time.13 In contrast, during the uterus and surgical bleeding from the uterine incision in EXIT procedure the uterine incision to clamping the combination with the duration of the procedure. In the cord time is as long as is needed for the fetal diagnostic/ present cases, bleeding from the edges of the uterine therapeutic interventions to take place. incision was minimized by application of a continuous There are many studies comparing the effects on the suture. A uterine incision stapling device developed and neonate of general versus regional anesthesia for stan- used for this purpose may, however, be more effective in dard cesarean section in terms of Apgar scoring and fetal this respect.5 Except for the first case, our patients re- blood gases and acid-base status. It seems obvious that ceived a nitroglycerin infusion in addition to the volatile general anesthesia with a combination of thiopentone anesthetic. In two of these patients blood loss was and volatile anesthetic, compared to regional anesthesia, greater than with a standard cesarean section, although may induce transient sedation associated with decreased maternal hemodynamics remained well preserved. In one-minute Apgar values. The exact role of the volatile retrospect the nitroglycerin treatment in our patients was agent in this respect is not known. Unfortunately ran- probably unnecessary; it may be more useful for rescue domized studies on neonatal outcome comparing dif- if uterine relaxation is insufficient, or as an alternative to ferent ET concentrations of modern volatile anesthetics high-dose volatile anesthesia.3;6 during general anesthesia for cesarean section are lack- In-vitro measurements using pregnant human uterine ing. In contrast to the possible effect of general anes- muscle strips show a dose-dependant decrease in con- thesia on one-minute Apgar scores, two large tractility with increased MAC values (0.5–1.5). This is epidemiological studies suggest that fetal acidosis is true for the older agents halothane and enflurane as well more common after regional than general anesthesia in as for the modern agents isoflurane, sevoflurane and elective cesarean section.14;15 It should not be forgotten, desflurane.21 The uterine responsiveness to oxytocin, as however, that surgical management also has an impact investigated in patients after vaginal delivery, is how- on the status of the neonate. The time from uterine in- ever maintained up to 0.8 and 0.9 MAC for halothane cision to delivery may even be a factor of greater im- and enflurane respectively.22 In this study it was also portance for neonatal well-being than the type of shown that “with both agents there was a rapid recovery anesthetic.13 of spontaneous uterine activity when deep planes of In healthy sheep, utero-placental blood flow, fetal anesthesia was lightened.” The risk of increased blood blood acid-base status and fetal oxygen saturation loss from prolonged use of high doses of volatile anes- remain well preserved with 1.0 and 1.5 MAC halo- thetics is obvious, but as our four patients clearly show, thane and isoflurane, although 2 MAC caused a de- the depressed state of uterine contractility could be terioration in these parameters due to maternal quickly reversed at the time of delivery without any circulatory depression.16 In the asphyxiated lamb fetus signs of intra- or postoperative recurrence. These find- a 15-minute exposure to 2 MAC halothane did not ings are in keeping with the results from 31 consecutive cause further deterioration in fetal status.17;18 In longer EXIT procedures where blood loss not was increased fetal operations halothane may, however, be associated compared to standard cesarean sections.5 182 International Journal of Obstetric Anesthesia

In conclusion, four cases of the EXIT procedure are 8. Moir D D. Anaesthesia for caesarean section. An evaluation of a described with special reference to maternal anesthetic method using low concentrations of halothane and 50 per cent of oxygen. Br J Anaesth 1970; 42: 136–142. technique. Based on this experience and the literature, it 9. Glosten B. Anesthesia for obstetrics. In: Miller R D (ed). seems that the fear of sustained uterine atony intra- or Anesthesia. Philadelphia: Churchill Livingstone, 2000: 2024–2068. postoperatively from high doses of modern short-acting 10. Biribo M A. General anesthesia for cesarean section. In: Birnbach D J, Gatt S P, Datta S (eds). Textbook of Obstetric Anesthesia. volatile anesthetics used for a short period during a Philadelphia: Churchill Livingstone, 2000: 239–243. standard cesarean section under general anesthesia, may 11. King H, Ashley S, Brathwaite D, Decayette J, Wooten D J. be exaggerated. From a fetal point of view the EXIT Adequacy of general anesthesia for cesarean section. Anesth Analg 1993; 77: 84–88. procedures were well tolerated in terms of acid-base 12. Jouppila P, Kuikka J, Jouppila R, Hollmen A. Effect of induction of status and blood gas values indicating a well preserved general anesthesia for cesarean section on intervillous blood flow. utero-placental perfusion despite the prolonged use of Acta Obstet Gynecol Scand 1979; 58: 249–253. 13. Datta S, Ostheimer G W, Weiss J B, Brown Jr W U, Alper M H. high ET concentration of volatile agents. Neonatal effects of prolonged anesthetic induction for cesarean section. Obstet Gynecol 1981; 58: 331–335. 14. Roberts S W, Leveno K J, Sidawi J E, Lucas M J, Kelly M A. Fetal REFERENCES acidemia associated with regional anesthesia for elective cesarean delivery. Obstet Gynecol 1995; 85: 79–83. 1. Harrison M R, Adzick N S, Flake A W et al. Correction of 15. Mueller M D, Bruhwiler H, Schupfer G K, Luscher K P. Higher congenital diaphragmatic hernia in utero VIII: Response of the rate of fetal acidemia after regional anesthesia for elective cesarean hypoplastic lung to tracheal occlusion. J Pediatr Surg 1996; 31: delivery. Obstet Gynecol 1997; 90: 131–134. 1339–1348. 16. Palahniuk R J, Shnider S M. Maternal and fetal cardiovascular and 2. Liechty K W, Crombleholme T M, Flake A W et al. Intrapartum acid-base changes during halothane and isoflurane anesthesia in the airway management for giant fetal neck masses: the EXIT (ex utero pregnant ewe. Anesthesiology 1974; 41: 462–472. intrapartum treatment) procedure. Am J Obstet Gynecol 1997; 177: 17. Yarnell R, Biehl D R, Tweed W A, Gregory G A, Sitar D. The 870–874. effect of halothane anaesthesia on the asphyxiated foetal lamb in 3. Mychaliska G B, Bealer J F, Graf J L, Rosen M A, Adzick N S, utero. Can Anaesth Soc J 1983; 30: 474–479. Harrison M R. Operating on placental support: the ex utero 18. Cheek D B, Hughes S C, Dailey P A et al. Effect of halothane on intrapartum treatment procedure. J Pediatr Surg 1997; 32: 227–231. regional cerebral blood flow and cerebral metabolic oxygen 4. Skarsgard E D, Chitkara U, Krane E J, Riley E T, Halamek L P, consumption in the fetal lamb in utero. Anesthesiology 1987; 67: Dedo H H. The OOPS procedure (operation on placental support): 361–366. in utero management of the fetus with prenatally diagnosed 19. Sabik J F, Assad R S, Hanley F L. Halothane as an anesthetic for tracheal obstruction. J Pediatr Surg 1996; 31: 826–828. fetal surgery. J Pediatr Surg 1993; 28: 542–546. 5. Bouchard S, Johnson M P, Flake A W et al. The EXIT procedure: 20. Fogel S T, Langer J C. Anesthesia for fetal surgery. In: Norris M C experience and outcome in 31 cases. J Pediatr Surg 2002; 37: (ed). Obstetric Anesthesia, 2nd Ed. Philadelphia: Lippincott 418–426. Williams & Wilkins, 1999: 187–210. 6. Gagnon A L, Bebbington M W, Kamani A, Solimano A. Prenatally 21. Turner R J, Lambros M, Kenway L, Gatt S P. The in-vitro effects diagnosed fetal neck teratoma. Fetal Diagn Ther 1998; 13: of sevoflurane and desflurane on the contractility of pregnant 266–270. human uterine muscle. International Journal of Obstetric 7. Bui T H, Grunewald C, Frenckner B et al. Successful EXIT (ex Anesthesia 2002; 11: 246–251. utero intrapartum treatment) procedure in a fetus diagnosed with 22. Marx G F, Kim Y.-I., Lin C.-C., Halevy S, Schulman H. congenital high airway obstruction syndrome due to laryngeal Postpartum uterine pressures under halothane or enflurane atresia. Eur J Pediatr Surg 2000; 10: 328–333. anesthesia. Obstet Gynecol 1978; 51: 695–698.