1

Introduction

Up to 15% of children undergo general anaesthesia before the age of 3 years, primarily for otorhinolaryngologic surgery [ ]. With the advances in foetal medicine, surgery may now be recommended before birth with nowadays an ever-increasing range of foetal anomalies eligible for antenatal surgical intervention. These foetal interventions vary, from minimally invasive procedures (like intrauterine blood transfusions) performed under local or neuraxial maternal anaesthesia, to major open surgeries (such as spina bifida repair) requiring maternal general anaesthesia. This review outlines the indications for foetal surgery, surgical techniques, management of maternal anaesthesia, the necessity of administering drugs to the foetus, and the neurodevelopmental impact of foetal anaesthesia exposure.

2

Foetal surgery

Due to the widespread implementation of ultrasound screening programs, the detection of congenital anomalies occurs more frequently and at earlier stages [ ]. The impact of these anomalies, particularly those with a dismal prognosis, stimulated research in prenatal interventions likely to improve the health of the affected children. While medical prenatal interventions exist, this section will focus on surgical interventions during the foetal stage, commonly referred to as foetal surgery. Today, foetal surgery is a possibility for numerous conditions aiming to improve foetal prognosis, either through anatomical correction of the malformation or by halting the progression of the disease with a more definitive repair until after birth. Broadly, these interventions can be divided into four categories.

The first category involves needle-based interventions such as the commonly practiced intra-uterine transfusion, used for treating anaemic foetuses [ ]. In this method, following local anaesthesia of the mother, a thin needle is guided by sonographic imaging either into the umbilical vein (at the placental insertion or intra-hepatically) or intra-peritoneally for blood transfusion. Intrauterine transfusion remains a crucial treatment for severe foetal anaemia resulting from e.g. red cell allo-immunization, human parvovirus B19 infection, foeto-maternal haemorrhage, twin-twin transfusion syndrome, and placental or foetal tumors. Other procedures in this category include amniodrainage, radiofrequency ablation, balloon valvoplasty, and shunt placements [ ]. Additionally, in-utero stem cell and gene therapies are being explored as potential interventions [ ].

The second category includes foetoscopic procedures using trocars for direct surgical visualization. A widely performed procedure is laser coagulation of placental vascular anastomoses for twin-twin transfusion syndrome. This syndrome results from an unequal blood flow between monochorionic twins due to shared placental vessels [ ]. During this procedure, a trocar and a foetoscope are percutaneously introduced under sonographic guidance, typically with local or regional anaesthesia in the mother. A laser fibre is advanced via the foetoscope to coagulate connecting vessels, resulting in a 76% survival rate of at least one foetus of which 52% have no major neurologic issues at six months of age. Another procedure in this category involves foetoscopic endoluminal tracheal occlusion (FETO) which has been compared to expectant care in a randomized controlled trial (TOTAL-trial) [ , ]. FETO implicates placing a balloon into the foetal trachea to improve the prognosis of foetuses with congenital diaphragmatic hernia (CDH) ( Fig. 1 ). This occlusion leads to the accumulation of lung fluid, causing lung stretch and subsequent proliferation and growth of the airways and pulmonary vessels [ ]. The TOTAL-trial found a significant increase in survival for severe left-sided CDH (FETO: 16/40 [40%] vs control: 6/40 [15%]; RR 2.67; 95% CI 1.22–6.11; P = 0.009). However, 47% of women experienced preterm rupture of membranes and 75% delivered preterm. For moderate hypoplasia, the survival was 63% vs. 50% (FETO: 62/98 [63%] vs control: 49/98 [50%]; RR 1.27; 95% CI: 0.99–1.63; P = 0.06). A pooled analysis indicated a significant treatment effect for the entire population of left-sided CDH, with the increase in survival to discharge being higher the earlier the balloon had been inserted.

Foetoscopic endoluminal tracheal occlusion (FETO)

FETO involves placing an occlusive balloon into the foetal trachea leading to the accumulation of lung fluid. Used with permission from UZ Leuven.

Foetoscopic procedures are also used for other interventions such as for the exploration of the lower urinary tract [ ], or for foetal spina bifida aperta repair [ , ].

Third, some foetal interventions need an “open” approach , typically involving a laparotomy with hysterotomy to access the foetus. Various techniques such as staples, haemostatic coagulation, and/or sutures, are employed to prevent and manage bleeding during hysterotomy. Historically, this approach was only used for the treatment of congenital pulmonary airway malformations (e.g., in cases of a tracheal web to safeguard the airway) around the time of delivery. This is usually referred to as an Ex Utero Intrapartum Treatment (EXIT) procedure, a modified caesarean delivery technique that allows surgical interventions to be performed while the neonate remains on placental circulation [ ]. Later, the “open” approach was also used for resection of sacrococcygeal teratomas and for foetal spina bifida repair. The Management of Myelomeningocele Study (MOMS trial) compared prenatal open foetal spina bifida repair with standard postnatal neurosurgical repair [ ]. In the experimental group, surgery took place between 19 + 0 and 25 + 6 weeks of pregnancy and involved general anaesthesia, laparotomy, uterus exposure, and hysterotomy. Thereafter, the spinal defect was then closed in layers, similar to the normal postnatal repair. Prenatal repair reduced the need for ventriculoperitoneal shunts at 12 months of age (31/78 [40%] prenatal repair vs 66/80 [82%] postnatal repair; RR 0.48; 95% CI 0.36–0.64; P < 0.001) and was associated with less cerebellar herniation (45/70 [64%] prenatal repair vs 66/69 [96%] postnatal repair; RR 0.67; 95% CI 0.56–0.81; P < 0.001). Additionally, after prenatal repair twice as many children walked by 30 months versus postnatal repair (26/62 [42%] vs 14/67 [21%]; RR 2.01; 95% CI 1.16–3.48; P = 0.01). Later studies confirmed the improvement, with children undergoing prenatal surgery functioning two levels higher than anticipated anatomically [ , ]. However, this technique comes at the cost of a 46% rate of preterm membrane rupture, preterm deliveries averaging 34.1 ± 3.1 weeks of gestation, and complications related to the fundal uterine scar (caesaerean sections for al future deliveries).

Last, the fourth category comprises “hybrid” foetal interventions , combining aspects of open and foetoscopic surgery. These methods are primarily explored for the treatment of spina bifida aperta, aiming to address complications associated with prematurity and uterine scar. In one method, after administering general anaesthesia and performing a laparotomy, the uterus is exteriorized. Then, the membranes are secured, and two or three trocars are inserted for a layered spina bifida repair ( Fig. 2 ). This approach seems to reduce prematurity and preterm membrane rupture rates compared to the open method without affecting neurosurgical outcomes [ , , ]. Another technique involves a 3 cm “mini-laparotomy” [ ]. Once done, the uterine wall and membranes are sutured, a camera port is added, and two more ports are inserted percutaneously, guided by ultrasound using the Seldinger technique. After the spinal defect is repaired, the uterine wall and membranes are secured through the mini-laparotomy [ ].

Three-port foetoscopic SBA repair by laparotomy and exposed uterus

a) First port in the exposed uterus filled with humidified CO2; (b) Placement of three ports (c) Visualization of the exposed lesion (d) After foetal skin closure. Used with permission from UZ Leuven and the patient.

As foetal surgery becomes more common, techniques are evolving to be less invasive and new indications are emerging. This rapidly advancing field is set to innovate treatments for numerous conditions.

3

Maternal anaesthesia conduct

As previously described, foetal surgery refers to highly specialized medical procedures performed on the foetal-maternal unit during pregnancy. Anaesthesia for foetal surgery is a crucial and complex component, vital for ensuring the well-being of both the mother and the unborn child. This involves not only alleviating pain in the mother and/or the foetus but also optimizing surgical conditions by ensuring effective uterine relaxation ( Fig. 3 ). The diverse nature of painful stimulation, the need for profound uterine relaxation, and the intricacy of open procedures contribute to the highly procedure-specific nature of foetal surgery anaesthesia.

Aims of the anaesthesia technique in foetal surgery

Anaesthesia aims to avoid maternal and foetal pain, to provide immobilization and/or to optimize surgical conditions by ensuring uterine relaxation.

The aforementioned groups (minimally invasive, open mid-gestational, ex-utero intra-partum treatment, and hybrid procedures) each require distinct anaesthetic management. For each procedure, three key questions need to be addressed: How can maternal immobility, safety, and comfort be guaranteed? How to ensure foetal immobility (e.g. foetal movement during laser photocoagulation of placental vasculature can harm the foetus), foetal safety, and comfort (minimizing pain, discomfort, stress, and haemodynamic response)? And to what extent is uterine relaxation necessary?

4

Direct administration of drugs to the foetus

The debate surrounding the foetus’s ability to perceive pain remains contentious. In the 1980s, the medical consensus leaned towards the belief that foetuses and even newborns lacked the capacity to experience pain due to the underdeveloped cerebral cortex [ ]. This resulted in surgeries on infants being performed using only neuromuscular blocking drugs and minimal, if any, anaesthesia or pain relief.

The assessment of foetal pain is challenging due to the inability to objectively measure it. Alternative evidence has been sought to understand foetal pain perception [ ]. In the 1990s, it was noted that nociceptive stimuli in the foetus triggered the hypothalamo-pituitary-adrenal axis and the sympathetic nervous system [ , , ]. However, using this stress response as a sole indicator for foetal pain had limitations: it did not involve the cortex, and responses like exercise, hypoxia, or haemorrhage could elicit similar reactions [ ]. Later it was suggested that foetal pain perception required the transmission of pain stimuli from peripheral nociceptors to the somatosensory cortex [ , ]. This pathway, functional only from 24 weeks of gestation, implied that pain perception might only begin at this stage [ , ]. Responses before 24 weeks were viewed as reflexes rather than indicative of pain perception [ ]. However, the latest evidence suggests that thalamic connections to the subplate, forming as early as 12 weeks of gestation, might be functionally equivalent to thalamocortical connections [ , , ]. The subplate, a transient layer in the developing cerebral cortex, acts as an active precursor to the cortex [ , ]. Consequently, the emerging view is that foetal pain perception might occur as early as 12 weeks of gestation [ , , ].

The exposure of foetuses to early pain and stress may impact their neurodevelopment, leading to both short- and long-term consequences [ , ]. Direct administration of fentanyl to the foetus has been observed to blunt the foetal stress response [ , ], a strategy that can be essential in immobilizing the foetus during foetal surgeries [ ]. While volatile anaesthetics and propofol, critical components of foetal anaesthesia, cross the placenta, their foetal concentrations, even after prolonged administration, reach only around 70% and 50% of maternal concentrations, respectively [ , ]. This level may not suffice for foetal surgery [ , , ]. Hence, in procedures involving innervated tissue or necessitating foetal immobilization, a combination of opioids (e.g., 10-20-50 μg/kg fentanyl), a neuromuscular blocker (e.g., 1.2 mg/kg rocuronium, 0.4 mg/kg cisatracurium, or 0.2 mg/kg vecuronium), and an anticholinergic drug (e.g., 10–20 μg/kg atropine) is administered [ , , ]. These drugs are commonly combined in a single syringe, with dosing based on the most recent estimated foetal weight. Procedures involving the umbilical cord, placenta, and foetal membranes, lacking nociceptors, usually do not require direct drug administration to the foetus [ ]. Instead, they are typically performed using local anaesthesia infiltration or regional anaesthesia for the mother. However, foetal movements during these procedures can potentially cause trauma or extend the surgery duration. Administering remifentanil to the mother (at 0.1 μg/kg/min) can offer maternal sedation, foetal immobilization, and foetal analgesia, thereby improving operating conditions [ , ].

Foetal drug administration can occur through various routes, each with its own advantages and limitations [ ]. Intravascular administration offers a rapid onset of action and can be achieved through the umbilical vein, large foetal veins, or intracardiacally, depending on the procedure [ ]. However, this method presents potential disadvantages such as vessel thrombosis, vascular spasm, bleeding, and a possible compromise of the surgical view [ ]. Accessing these routes can also be more challenging compared to other methods [ ]. On the other hand, the intramuscular route is commonly employed during open surgery and some minimally invasive procedures, particularly those guided by ultrasound [ ]. It offers easier administration with a lower risk of bleeding compared to intravenous methods [ ]. However, drug absorption can be variable, making it more challenging to predict the time course of drug action accurately [ ]. Each route presents its own set of advantages and challenges, often influencing the choice of administration method based on the specific requirements and procedure involved [ ].

5

Neurodevelopmental effects of foetal anaesthesia exposure

The groundbreaking work by Ikonomidou et al., in 1999 demonstrated the widespread neuronal apoptosis in foetal and neonatal rats by blocking NMDA receptors using dizocilpine [ ]. This study laid the foundation for the hypothesis that exposure to general anaesthesia during brain development might significantly impact neurodevelopmental outcomes [ ]. In 2003, Jevtovic-Todorovic et al. confirmed this hypothesis using anaesthetics commonly used in clinical practice [ ]. They exposed early neonatal rats to midazolam, nitrous oxide, and isoflurane for 6 h, resulting in extensive neuronal apoptosis and subsequent learning and memory impairments [ ]. In subsequent years, a plethora of investigations conducted across diverse animal species consistently illuminated the potential for compromised neurodevelopment following exposure to commonly utilized general anaesthetics, both prenatally and postnatally [ ]. This body of research, spanning from 2003 to recent years, highlighted the detrimental effects of these anaesthetics on brain development. In response to this growing body of evidence, in 2016, the Food and Drug Administration (FDA) issued a warning emphasizing that repeated or prolonged exposure to general anaesthesia during the third trimester of pregnancy might have adverse effects on neurodevelopment [ ].

A recent meta-analysis, pooling findings from 65 animal studies, underscores the potential for foetal exposure to general anaesthesia to cause neuronal injury and impair subsequent learning and memory in animals post-birth [ ]. This outcome was consistent across various animal species, anaesthetic drugs, and pregnancy trimesters. However, the study highlighted limitations that hinder the direct translation of these findings to clinical practice. One notable limitation lies in the discrepancies between the duration, frequency, and doses of anaesthesia used in animal studies compared to typical clinical scenarios. Neurological impairments were observed with repeated exposures, doses surpassing one MAC (minimum alveolar concentration), or durations extending beyond 3 h. A single exposure at or below 1 MAC for up to 3 h did not yield significant impairments. It is crucial to contextualize the duration of exposure concerning the animal’s pregnancy duration: for instance, exposure to 3 h of anaesthesia in rats, mice, non-human primates, sheep, rabbits, and guinea pigs would roughly equate to 37, 43, 5, 6, 27, and 12 h of anaesthesia in humans, respectively. Another limitation was the absence of surgical stimulation during most animal exposures to anaesthesia, which does not mirror clinical procedures where anaesthesia is often coupled with surgical interventions. Finally, the animal studies generally fell short in terms of monitoring and maintaining physiological stability, not reaching the clinical standards expected in human settings. These limitations collectively caution against directly applying the findings of animal studies to human clinical practices.

Clinical data on the effects of prenatal anaesthesia exposure and surgery are available only for maternal non-obstetric surgery (but not for foetal surgery). An ambidirectional cohort study investigated executive function, behavioural problems and psychiatric diagnoses in 129 children which had been prenatally exposed to general anaesthesia for maternal surgery [ ]. These children were compared with 453 unexposed children [ ]. Ninety percent of exposed children underwent a single exposure with an average duration 91 min [ ]. Bias by confounding, (e.g. for the education and income of the parents) was reduced by using propensity score weighting [ ]. This study found no evidence for an association between prenatal exposure to anaesthesia and impairment of neurodevelopmental outcomes in childhood [ ]. An additional systematic review comparing the neurodevelopmental effects of prenatal anaesthesia exposure for maternal surgery with those of other risk factors and conditions, concluded that the effect size of prenatal anaesthesia exposure found in the bidirectional cohort study was comparable with that of factors frequently encountered during pregnancy and childhood (e.g. the impact of a child spending ≥2 h per day in front of a television screen or the use of >15 episodes of paracetamol during pregnancy.) [ ]. Nevertheless, a constraint of this bidirectional cohort study was the inclusion of a wide range of procedures, with 14% of patients only being subjected to locoregional anaesthesia [ ]. Furthermore, the absence of a discernible difference in the average outcomes among all these patients does not exclude the potential for more substantial impairments in a subset of more vulnerable patients [ ]. Indeed, notable impairments were identified in specific subgroups of children whose mothers underwent general anaesthesia, intra-abdominal surgery, prolonged anaesthesia (>1 h), and laparoscopic surgery [ ]. The impact of general anaesthesia and foetal surgery on foetal brain development remains a topic for prospective investigations.

6

Conclusion

Foetal surgery offers the potential for improved outcomes in various congenital defects. The procedures span from minimally invasive ones under local anaesthesia with maternal sedation to more complex open mid-gestational surgeries that require general anaesthesia. The key goals of anaesthesia in these scenarios are to prevent both maternal and foetal pain, ensure foetal immobilization, and optimize surgical conditions through uterine relaxation. A critical consideration is the potential for the foetus to experience pain, which might occur as early as 12 weeks’ gestation. Consequently, procedures involving innervated tissues or requiring foetal immobilization often necessitate administration of specific medications—such as opioids, neuromuscular blockers, and anticholinergic drugs—directly to the foetus. The administration routes commonly used for these anaesthetic drugs are intramuscular or intravenous, depending on the specifics of the procedure. For foetal procedures necessitating a laparotomy, the avoidance of general anaesthesia is impossible. The administration of anaesthesia is crucial to ensuring the safety of both the mother and the baby, as well as to mitigate any pain experienced during the procedure and to provide optimal uterine relaxation. Although preclinical studies have suggested a potential impact of prenatal anaesthesia exposure on foetal brain development, translating these findings to clinical scenarios faces limitations and complexities. Further research is essential to better understand the nuances and implications of anaesthesia on the developing foetal brain within the clinical context.

Practice points

• •
  

  Advances in ultrasound increase early detection of congenital anomalies, allowing prenatal interventions.

  
  
• •
  

  Evolving techniques in foetal surgery, ranging from needle-based procedures to foetoscopic and “open” methods, are aimed at achieving safer and less invasive treatments.

  
  
• •
  

  Minimally invasive procedures are performed under local anaesthesia (with maternal sedation), while open mid-gestational procedures require general anaesthesia.

  
  
• •
  

  The objective of anaesthesia is to prevent maternal and foetal pain, facilitate immobilization, and/or enhance surgical conditions through the assurance of uterine relaxation.

  
  
• •
  

  In all procedures involving innervated tissue or requiring foetal immobilization, an opioid, neuromuscular blocker and anticholinergic drug needs to be administered directly to the foetus.

  
  
• •
  

  Animal studies showed that prenatal anaesthesia exposure might affect the foetal brain development, but the translation of these results to the clinical setting is limited by several factors.

Research agenda

• •
  

  Explore improvements of current techniques to reduce complications such as preterm delivery and uterine scarring.Further research is warranted to investigate from which gestational age the foetus can experience pain as a conscious and emotional feeling.

  
  
• •
  

  Studies are required to investigate the effects of general and locoregional anaesthesia during foetal surgery on foetal brain development.

Summary

Nowadays, ultrasound screening in unborn children detect congenital anomalies earlier and more frequently, prompting research into foetal surgery that can offer treatment options for various conditions, either by correcting congenital anomalies or halting disease progression until after birth. Minimally invasive procedures are performed under local anaesthesia (with maternal sedation), while open mid-gestational procedures require general anaesthesia. Anaesthesia aims to avoid maternal and foetal pain, to provide immobilization and/or to optimize surgical conditions by ensuring uterine relaxation. As early as the gestational age of 12 weeks, the foetus may experience pain. Therefore, in all procedures involving innervated tissue or requiring foetal immobilization, opioids, neuromuscular blockers and anticholinergic drugs are administered directly to the foetus, most commonly intramuscularly or intravenously. Animal studies showed that prenatal anaesthesia exposure might affect the foetal brain development, but translation of these results to the clinical setting is limited by several factors.

Declaration of artificial intelligence technologies in the writing process

During the preparation of this work the authors used ChatGPT (OpenAI, San Francisco, California, USA, https://chat.openai.com/ ) in order to improve the readability and language. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Funding

• •
  

  TB, SD and SR are funded by the European Society of Anaesthesiology and Intensive Care (ESAIC) Young Investigator Start-up Grant 2021, the Society for Anaesthesia and Resuscitation of Belgium (SARB) research grant 2019 and the Obstetric Anaesthetists’ Association (OAA) International Grant 2019.

  
  
• •
  

  SV is supported by the Flanders Research Foundation (Fonds Wetenschappelijk Onderzoek Vlaanderen T002618N ).

  
  
• •
  

  JD is funded by the Great Ormond Street Hospital Charity Fund , the Welcome Trust ( WT101957 ) and Engineering and Physical Sciences Research Council (ESPRC) ( NS/A000027/1 ).

  
  
• •
  

  None of these funding sources played a role in the study design, data collection, analysis, interpretation or writing of the article.

CRediT authorship contribution statement

Tom Bleeser: Writing – original draft, Writing – review & editing. Arjen Brenders: Writing – original draft, Writing – review & editing. Simen Vergote: Writing – original draft, Writing – review & editing. Jan Deprest: Supervision, Writing – original draft, Writing – review & editing. Steffen Rex: Supervision, Writing – original draft, Writing – review & editing. Sarah Devroe: Supervision, Writing – original draft, Writing – review & editing.