If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Peri-operative management of percutaneous fetoscopic spina-bifida repair: a descriptive review of five cases from the United Kingdom, with focus on anaesthetic implications
Percutaneous fetoscopic spina bifida repair under general anaesthesia is safe.
•
Fetal well-being can be assessed by umbilical artery pulsatility index.
•
Intravenous anaesthesia and inhaled sevoflurane provide reversible tocolysis.
•
Transplacental anaesthesia achieves adequate fetal anaesthesia and immobilisation.
•
Intra-uterine pressure during CO2 inflation best maintained near ‘opening’ pressure.
Abstract
We present a case-based review of the first five percutaneous fetoscopic in-utero spina bifida repair procedures undertaken in the UK. Our focus is on implications of anaesthesia and analgesia for the mother and fetus, provision of uterine relaxation and fetal immobilisation while providing conditions conducive to surgical access. Minimising risks for fetal acidosis, placental and fetal hypoperfusion, maternal and fetal sepsis and maternal fluid overload were the foremost priorities. We discuss optimisation strategies undertaken to ensure fetal and maternal well-being under anaesthesia, shortcomings in the current approach, and possible directions for improvement.
Current practice is surgical repair within 1–2 days of birth. Recently, several animal and human reports have suggested that prenatal repair could offer better postnatal neurological function.
In 2011, the Management of Myelomeningocele Study (MOMS) compared open (hysterotomy) prenatal repair with the standard postnatal technique. They found reduced rates of neurosurgical intervention for hydrocephalus in a significant proportion of those treated prenatally and an improved composite score for mental development and motor function at 30 months.
However, it was associated with increased rates of preterm birth (13% before 30 weeks’ gestation) and uterine dehiscence.
In the quest to reduce maternal and fetal complications, fetoscopic approaches have been adopted by many centres around the world with promising results.
There is no published guidance on optimal modes of anaesthesia. We report our first experiences of percutaneous fetoscopic spina bifida repair, including optimisation strategies undertaken to ensure fetal and maternal well-being under anaesthesia, shortcomings of the current approach and directions for improvement in the future.
Case series
Eligibility
The MOMS trial criteria were used to determine eligibility for antenatal repair. All cases had fetal spina bifida recognised on antenatal ultrasound and confirmed by magnetic resonance imaging (MRI) early in the second trimester. None of the fetuses had any other congenital abnormalities. The parents were counselled on the options available for repair and, following a multidisciplinary discussion, the percutaneous fetoscopic approach was offered. The parents gave consent and surgery was scheduled for the late second trimester.
Operative course
Extensive pre-operative planning was undertaken with advice from international experts. All team members participated in a simulation on the day of surgery.
Before surgery, the women were fasted, provided with standard antacid premedication and anti-thrombo-embolic stockings, and given rectal (PR) indomethacin 100 mg 12 h pre-operatively. Following pre-oxygenation, a rapid sequence induction was carried out with intravenous (IV) fentanyl 3 µg/kg, propofol 1.5 mg/kg and rocuronium 0.8 mg/kg. This generally achieved a bispectral index (BIS) of 25–30. Following intubation, positive-pressure ventilation was established with hyperventilation during intra-uterine carbon dioxide (CO2) insufflation to maintain end-tidal CO2 (ETCO2) below 3.5 kPa. Antibiotic prophylaxis was administered at induction (cefuroxime 1.5 g in case 1 and clindamycin 600 mg in cases 2–5), with three further doses postoperatively.
Anaesthesia was maintained with sevoflurane at a minimum alveolar concentration (MAC) of 0.4–1, intravenous propofol infusion 2–3 mg/kg/h and remifentanil infusion 10 µg/kg/h adjusted to maintain cardiovascular stability. The depth of maternal anaesthesia was monitored using BIS and maintained at 40.
An arterial line was inserted to assist regular sampling and cardiovascular monitoring.
A phenylephrine infusion was titrated to maintain a target mean maternal blood pressure (65–70 mmHg in cases 1 and 2 and 80 mmHg in cases 3–5) during the procedure. Atropine 600 µg (cases 3–5) was given if the maternal heart rate fell below 60/min. A glyceryl trinitrate (GTN) infusion was available (50 mg/50 mL) to provide additional uterine relaxation. This was required in case 4 during which intermittent uterine contractions were observed following an intraperitoneal CO2 leak during the procedure.
All women were placed in a semi-lithotomy/supine position, a urinary catheter inserted and PR indomethacin 100 mg given for uterine tocolysis. An 18-gauge needle was inserted into the uterus under ultrasound guidance and 500 mL of warm Ringer’s lactate solution infused. Four percutaneous ports (in case 1) and three ports (in cases 2–5) were respectively inserted into the uterus under ultrasound guidance. Approximately 1 L of amniotic fluid was removed and humidified warm CO2 insufflated. The uterine ‘opening pressure’ was measured and the maximum insufflation pressure set at 4 mmHg above it.
In cases 2–5, pressure was maintained at 2–3 mmHg above the opening pressure. We observed momentary drops in blood pressure during the insertion of trocars. Uterine insufflation with CO2 seemed most stimulating for the mother, with a rise in blood pressure and BIS that was mitigated by IV morphine. The spinal defect was dissected by the neurosurgeon and repaired with a bio-cellulose patch placed over the neural placode (exposed open spinal cord), covered by an additional layer of a skin substitute (Nevelia®, Symatese, Montpelier, France) as needed (Fig. 1).
Fig. 1Completed fetoscopic patch repair of the spina-bifida (case 1)
The uterine cavity was irrigated with warm (37 °C) Ringer’s lactate solution, antibiotic instilled (cefuroxime 1.5 g in case 1, clindamycin 500 mg in cases 2–5) and the cavity re-filled with 1L Ringer’s lactate solution or the drained amniotic fluid if not blood-stained.
Uterine tone was restored with termination of the GTN infusion and sevoflurane inhalation 5 min before removal of the trocars. Intravenous anaesthesia was maintained during uterine port-site haemostasis and closure of abdominal wounds. The duration of the procedure ranged from 4.0 to 5.5 h. Maternal temperature was monitored and maintained with a warm-air blanket. Women received IV paracetamol 1 g, IV metoclopramide 10 mg and PR indomethacin 100 mg. The maternal blood loss was approximately 150–200 mL and total fluids during the procedure limited to Hartmann’s solution 1000–1500 mL. Following recovery from general anaesthesia, the mothers and fetuses were observed in the obstetric high dependency unit for 24–48 h. Adequate pain control was achieved with regular paracetamol and doses of oral morphine (0.15 mg/kg) except in case 2, for whom morphine patient controlled analgesia followed by epidural analgesia was required from days 2–4 because of intermittent shooting postoperative pain triggered by fetal movements impinging upon the uterine wounds. None of the women required a blood transfusion. There were no post-procedure placental abruptions or uterine dehiscences.
Fetal monitoring
Intra-operative fetal monitoring was obstetrician led. For case 1 this included pre- and postoperative ultrasound assessment of umbilical and middle cerebral artery blood flow and cardiotocogram (CTG), with visual monitoring of the umbilical cord pulsation intra-operatively. The uterine and middle cerebral artery blood flows were reported as normal before and after surgery. The beat-to-beat variation of the fetal heart on CTG was reduced postoperatively in case 1, although the baseline rate remained normal. Normal fetal movements were felt by the mother 48 h post-surgery.
In cases 2–5, additional intra-operative fetal monitoring was instituted every 15 min using a Doppler probe placed on the maternal flank under sterile conditions to gain umbilical artery signals from the umbilical cord, below the amniotic fluid level so as to avoid the CO2 interface (Fig. 2). This fetal umbilical cord blood flow study was used to calculate the pulsatility index (PI), the systolic velocity-diastolic velocity/mean velocity, which is a measure of umbilical cord forward blood flow, and placental resistance.
A summary of fetal outcomes is shown in Table 1. Baby case 1 was born in a good condition, weighing 1.49 kg (<50th centile), but required intubation, surfactant and continuous positive-airway pressure for acute respiratory distress syndrome (ARDS) for 48 h. Haemoglobin concentration at birth was 101 g/L, platelet count 327 × 109/L, total white cell count 58.3 × 109/L, s. Na 134 mmol/L and s. K 5.0 mmol/L. The cranial ultrasound performed on day 1 showed subtle changes suspicious of periventricular leukomalacia and ventriculomegaly (Fig. 3). On neurological examination, the baby displayed good power and movement in all limbs and normal anal tone, and the bladder scan was within the normal range. At 48 h, there was renal impairment with the s. creatinine rising to a maximum of 289 µmol/L. This was managed without dialysis. There was no positive culture but sepsis was considered likely and treated. At two weeks of age, the baby developed a pulmonary haemorrhage and required mechanical ventilation and a course of antibiotics again. Ventricular index and occipito-frontal head circumference (OFC) increased in size, this being resolved by regular tapping initially and a ventriculo-peritoneal shunt (VP). His course continued to be complicated by respiratory issues. The baby died from complications of severe aspiration pneumonia at three months of age.
Table 1Demographic outline of cases
Case number
Gestation at fetoscopic repair (weeks + days)
Onset of labour
Mode of delivery
Gestation at birth (weeks + days) Apgar scores at 1, 5 min
Infant age at this review (months, days) Current status
1
27 + 6
Spontaneous
Forceps
30 + 3 8, 10
VP-shunt Died at 3 months from complications of prematurity
Babies 2–5 were born in good condition and their spinal wounds healed well. All were at home at the time of this review. The haemoglobin at birth in case 2 was 152 g/L. All these babies are regularly monitored for hydrocephalus and two have required VP shunt insertion (Table 1).
Discussion
Experience in providing maternal and fetal anaesthesia has been gained from ex-utero intrapartum treatment (EXIT) procedures.
The aims of feto-maternal anaesthesia include fetal immobility, uterine relaxation and tocolysis, reducing the risk of sepsis, and preventing placental-fetal hypoperfusion and fetal acidosis.
Choice of anaesthetic
Inhalational agents, propofol, fentanyl and remifentanil given to the mother will all cross the placenta to provide anaesthesia and analgesia to the fetus.
If regional anaesthesia is used, separate provision of fetal anaesthesia and analgesia by direct fetal intramuscular injection is necessary. Drugs given to the fetus also have the potential to cross the placenta into the maternal circulation. Although the doses are significantly smaller, drug specific implications for the mother and fetal oxygenation have been reported.
There is a further risk from absorption of amniotic and irrigating fluid through the myometrium and fluid retention due to tocolytic agents, especially beta-agonists, magnesium sulphate and atisoban, all of which contribute to maternal pulmonary oedema independent of anaesthetic technique. We opted for fluid-restrictive general anesthesia.
We used a rapid sequence induction with cricoid pressure and tracheal intubation to reduce the risk of aspiration of gastric contents. General anaesthesia with controlled ventilation allowed manipulation of maternal ETCO2 (and hence fetal arterial CO2) in response to uterine insufflation. Animal studies suggest that fetal acidosis may predispose to preterm delivery.
A study evaluating the consequences of a pneumoperitoneum on mid-term sheep fetuses reported that CO2 inflation to 15 mmHg for 90–120 min induced fetal respiratory acidosis and correction via maternal hyperventilation was late and incomplete.
but as umbilical venous blood is already equilibrated with maternal blood it may not give a reliable assessment of fetal acidosis. Umbilical arterial blood sampling is not feasible during fetoscopy because the sample would be contaminated by the pressurised CO2 environment and because of the risk of CO2 embolus.
Anesthetic management for percutaneous minimally invasive fetoscopic surgery of spina bifida aperta: a retrospective, descriptive report of clinical experience.
Animal studies suggest loss or death of brain cells and impaired neurocognitive function following anaesthetic exposure in neonates and late gestation fetuses.
Human studies in this area are limited and currently inconclusive. In 2016, the US Food and Drug Administration issued a warning regarding potential impaired brain development in children following exposure to inhalational agents (isoflurane, sevoflurane, desflurane) and intravenous agents (propofol, midazolam) in the third trimester of pregnancy, especially in procedures lasting >3 h.
The consequences of anaesthetic exposure in the second trimester when fetal surgical procedures are performed is unclear, although there is evidence that brief exposure is not associated with any long-term risk in humans.
The relatively high induction dose of fentanyl also contributed to fetal sedation. Total intravenous anaesthesia alone would have required higher cumulative doses of propofol that might induce fetal acidosis.
Maintenance of maternal normothermia via controlled active warming is essential. A prophylactic antibiotic that crosses the placenta was used as sepsis has been recognised as a cause of post-procedure fetal death.
Uterine tone was restored at removal of trocars by terminating these agents at the end of the procedure. Indomethacin, a cyclo-oxygenase-2 inhibitor, was given to reduce the risk of postoperative contractions and preterm labour, although it has recognised fetal side effects such as renal failure, necrotising enterocolitis, intraventricular haemorrhage and closure of the ductus arteriosus
We also had atosiban and magnesium sulphate available to supplement uterine relaxation. These were required in case 4 to mitigate contractions triggered by an intraperitoneal CO2 leak. A non-randomised cohort study comparing magnesium sulphate versus atosiban during and after open fetal myelomeningocele repair reported similar short-term uterine outcomes without any serious maternal complications.
Fetal oxygenation is dependent on maternal blood flow to the uteroplacental circulation, gas exchange across the placenta and fetal umbilical artery flow. Maternal hypotension, increased intra-uterine pressure, fetal bradycardia and hypovolaemia can cause fetal hypoxia.
We maintained maternal blood pressure with a phenylephrine infusion as its effects on placental blood flow are considered beneficial compared with ephedrine.
Compression of the aorta and inferior vena cava by the gravid uterus when the mother lies supine can reduce maternal blood pressure and hence placental flow, especially in the third trimester. We elected not to use a lateral tilt to assist surgical access, and momentary maternal hypotension was only observed during insertion of the trocars; this was assumed to be a result of aortocaval compression at this time.
Intra-uterine pressure during contractions can reach 60 mmHg but this pressure occurs intermittently, allowing recovery of the circulation between contractions.
Continuous inflation of the uterus to a pressure of 15–30 mmHg for surgery compresses the umbilical cord circumferentially and may impede umbilical arterial blood flow.
Reduced umbilical vein flow will decrease fetal preload and blood pressure. Furthermore, umbilical artery compression may increase umbilical arterial resistance and create back pressure, promoting flow diversion to the pulmonary circulation through the ductus arteriosus, and increase desaturated blood admixture (Fig. 4).
Fig. 4An illustration outlining plausible effects of intra-uterine pressure and transplacental anaesthesia on fetal circulation during fetoscopic surgery
Fetal intracranial pressure is most likely maintained in equilibrium with intra-uterine pressure via the open fontanelle and soft skull. High uterine inflation pressures, coupled with reduced fetal blood pressure, could also reduce cerebral perfusion. To minimise this effect, uterine inflation pressure was maintained just above ‘opening’ pressure, as this is reported to be safe.
and we maintained a relatively low MAC in our cases. Fetal hypovolaemia can be easily overlooked. The fetal blood volume is approximately 120–160 mL/kg in the second trimester
A 20 mL blood loss is a major haemorrhage in a fetus weighing 1.0 kg and with a 120 mL circulating blood volume. In addition, there is an increased bleeding tendency due to an immature coagulation system and high evaporative fluid losses.
During EXIT procedures or spina bifida repair with hysterotomy, direct access to the fetus allows monitoring with pulse oximetry, echocardiography, electrocardiography and umbilical blood gas analysis.
during fetoscopy this is less reliable because the CO2 insufflation contaminates the sample. Venepuncture or cordocentesis risk CO2 embolus into the fetal circulation.
Umbilical artery flow pulsatility index (PI) and fetal heart rate assessed via umbilical artery doppler in cases 2–5 (Fig. 2) seemed superior to pre- and post-procedural ultrasound and intra-operative visual observation of cord pulsation. The images and dopplers obtained were of good quality but required a dedicated person to monitor PI during the procedure.
Monitoring of fetal heart rate and beat-to-beat variability may offer early indications of fetal distress but continuous CTG monitoring is impractical as the gaseous interface prevents beat capture.
Since prolonged decreases in fetal heart rate or significant changes in umbilical artery flow, especially absent or reversed diastolic flow, are linked with increased perinatal morbidity and mortality,
The aetiology of suspected cerebral periventricular leukomalacia in case 1 was largely unexplained but was in keeping with an undetected insult 2–3 weeks before birth, coinciding with the fetoscopic intervention or thereafter. Reduced cerebral and placental blood flow are the leading causes of hypoxic-ischaemic encephalopathy of the new-born,
hence our measures to increase maternal mean blood pressure from 65 to 70 mmHg in cases 1 and 2 to 80 mmHg in cases 3–5 by means of chronotropes (atropine) and vasopressors (phenylephrine). Neonatal sepsis was possible but the C-reactive protein of <0.2 mg/L at birth did not support this. Acute kidney injury in case 1 neonate was unanticipated. Antenatally, there was no renal dysplasia or bladder outflow tract obstruction. Acute kidney injury can result from indomethacin
or hypoperfusion during fetoscopic intervention (Fig. 4). Beta-lactams such as ceftriaxone have been shown to induce interstitial inflammation in animal fetal kidneys,
and possibly due to blood loss at surgery or sepsis, but this was unlikely to have contributed to the kidney and cerebral injury. Babies born with lower haemoglobin levels following twin-to-twin transfusion syndrome have had preserved renal and cerebral function.
Following recovery from general anaesthesia the mothers were monitored in the obstetric high dependency unit for 24–48 h, with vital signs recorded every 4 h and CTG 12 hourly. Pain control was achieved with regular paracetamol and oral morphine as required, in order to minimise fetal and maternal stress-inducing preterm labour.
A single dose of indomethacin 100 mg was administered PR 6 h following surgery and progesterone 200 mg daily on discharge from hospital to reduce the risk of preterm labour. Indomethacin and atosiban were considered the least likely to be associated with serious adverse drug reactions in the mother.
The surgeons visited centres in Brazil and the USA, and the first three operations were attended by experienced fetal obstetric surgeons. There was extensive email correspondence between the UK and Brazilian anaesthetists about anaesthetic techniques. The UK anaesthetic and surgical teams met on several occasions before the procedures to ensure sharing of information and open discussion with the families about surgical and anaesthetic risks to both mother and baby, including image recording before obtaining informed consent.
The simulation with the full operating team on the day of surgery familiarised all personnel with the procedure, patient positioning, and laparoscopic and other equipment requirements.
In summary, percutaneous fetoscopic surgery poses many unique anaesthetic issues. These include inducing profound but reversible uterine relaxation, vigilance for maternal or fetal blood loss, transplacental anaesthesia, fetal immobilisation, fetal monitoring, maintenance of maternal mean blood pressure and a pro-active approach to fetal resuscitation. Postoperative measures to reduce the risk of premature delivery are required. The success of intra-uterine myelomeningocele repair relies on a well-coordinated multidisciplinary approach. Maintaining uterine CO2 inflation pressure close to ‘opening pressure’ seems prudent.
Acknowledgements
The authors thank Drs O. Long, S. Patel, M. Giombini, and M. Brown also lead nursing officers J. Zhu, F. Saud, Z. Kattah and R. Dasalla for their support. The authors also gratefully acknowledge the honorary medical illustration contributions of L. Bouikhsaine.
Declaration of interests
There are no conflicts of interest.
Funding
None.
References
Atta C.A.
Fiest K.M.
Frolkis A.D.
et al.
Global birth prevalence of spina bifida by folic acid fortification status: a systematic review and meta-analysis.
Anesthetic management for percutaneous minimally invasive fetoscopic surgery of spina bifida aperta: a retrospective, descriptive report of clinical experience.
30054-6&id=gr1.jpg){kind=link}
30054-6&id=gr2.jpg){kind=link}
30054-6&id=gr3.jpg){kind=link}
30054-6&id=gr4.jpg){kind=link}